text
stringlengths 649
4.42k
| synonym_substitution
stringlengths 759
4.5k
| butter_fingers
stringlengths 649
4.42k
| random_deletion
stringlengths 453
2.31k
| change_char_case
stringlengths 649
4.42k
| whitespace_perturbation
stringlengths 764
5.02k
| underscore_trick
stringlengths 649
4.42k
|
|---|---|---|---|---|---|---|
}^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$, then $S\subset \overline{S}'$.
ii. If $S'=S$, then $\eta'\in \Xi\left(X\left({}^{\vee}{G}^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$.
We are going to be mostly interested in the case of micro-packets attached to Langlands parameters coming from Arthur parameters. These types of micro-packets, are going to be called Adams-Barbasch-Vogan packets or simply ABV-packets. We continue by recalling the definition of an Arthur parameter.\
An Arthur parameter is a homomorphism $$\begin{aligned}
\psi:W_{\mathbb{R}}\times\textbf{SL}(2,\mathbb{C})\longrightarrow {}^{\vee}{G}^{\Gamma},\end{aligned}$$ satisfying
- The restriction of $\psi$ to $W_{\mathbb{R}}$ is a tempered Langlands parameter (i.e. the closure of $\varphi(W_{\mathbb{R}})$ in the analytic topology is compact).
- The restriction of $\psi$ to $\mathbf{SL}(2,\mathbb{C})$ is holomorphic.
Two such parameters are called equivalent if they are conjugate by the action of ${}^{\vee}{{G}}$. The set of equivalences classes is written $\Psi\left({}^{\vee}{G}^{\Gamma}\right)$ or $\Psi^{z}({G}/\mathbb{R})$, when we want to specify that ${}^{\vee}{G}^{\Gamma}$ is an $E$-group for $G$ with second invariant $z$.
To every Arthur parameter $\psi$, we can associate a Langlands parameter $\varphi_{\psi}$, by the following formula (see Section 4 [@Arthur89]) $$\begin{aligned}
\varphi_{\psi}&:W_{\mathbb{R}}\longrightarrow {}^{\vee}G^{\Gamma},\\
\varphi_{\psi}(w)&=\psi\left(w,\left(\begin{array}{cc}
|w|^{1/2}& 0\\
0& |w|^{-1/2}
\end{array}\right)\right).\end{aligned}$$ Now, to $\varphi_{\psi}$ correspond an orbit $S_{\varphi_{\psi}}$ of ${}^{\vee}{G}$ on $X\left({}^{\vee}{G}^{\Gamma}\right)$. We define $$\ | } ^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$, then $ S\subset \overline{S}'$.
ii. If $ S'=S$, then $ \eta'\in \Xi\left(X\left({}^{\vee}{G}^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$.
We are going to be mostly interested in the case of micro - packet attach to Langlands parameters coming from Arthur parameters. These character of micro - packets, are going to be call Adams - Barbasch - Vogan mailboat or simply ABV - package. We continue by recalling the definition of an Arthur parameter.\
An Arthur argument is a homomorphism $ $ \begin{aligned }
\psi: W_{\mathbb{R}}\times\textbf{SL}(2,\mathbb{C})\longrightarrow { } ^{\vee}{G}^{\Gamma},\end{aligned}$$ satisfying
- The restriction of $ \psi$ to $ W_{\mathbb{R}}$ is a tempered Langlands parameter (i.e. the blockage of $ \varphi(W_{\mathbb{R}})$ in the analytic topology is compact).
- The restriction of $ \psi$ to $ \mathbf{SL}(2,\mathbb{C})$ is holomorphic.
Two such parameters are called equivalent if they are conjugate by the military action of $ { } ^{\vee}{{G}}$. The set of equivalences classes is write $ \Psi\left({}^{\vee}{G}^{\Gamma}\right)$ or $ \Psi^{z}({G}/\mathbb{R})$, when we want to specify that $ { } ^{\vee}{G}^{\Gamma}$ is an $ E$-group for $ G$ with second invariant $ z$.
To every Arthur parameter $ \psi$, we can consort a Langlands parameter $ \varphi_{\psi}$, by the following formula (see Section 4 [ @Arthur89 ]) $ $ \begin{aligned }
\varphi_{\psi}&:W_{\mathbb{R}}\longrightarrow { } ^{\vee}G^{\Gamma},\\
\varphi_{\psi}(w)&=\psi\left(w,\left(\begin{array}{cc }
|w|^{1/2 } & 0\\
0 & |w|^{-1/2 }
\end{array}\right)\right).\end{aligned}$$ nowadays, to $ \varphi_{\psi}$ correspond an orbit $ S_{\varphi_{\psi}}$ of $ { } ^{\vee}{G}$ on $ X\left({}^{\vee}{G}^{\Gamma}\right)$. We define $ $ \ | }^{\Gamla}\right),{}^{\vee}{G}\right)_{S}^{mic}$, thek $S\subset \overliuw{S}'$.
ii. Ih $S'=S$, thsn $\eta'\in \Xi\left(X\left({}^{\vee}{G}^{\Gamma}\right),{}^{\vex}{G}\ritht)_{S}^{muc}$.
We are going to be mustly intvrested ib tht case of micro-pariets atbcched bo Laugoands parametets coming frmm Arthur parakegexs. These types of micro-packets, are gjing to bf called Adams-Farbsfch-Vkgan packets or simply ABV-packets. Se contpnue by recalling the definition of an Arthkr pwrameter.\
An Arthur oarameter iw a rimomorphism $$\cegin{aligntd}
\'si:W_{\mathbb{R}}\jimes\textbf{SL}(2,\mathbb{C})\longrightarruw {}^{\vez}{G}^{\Gamma},\end{aoitnef}$$ satisfying
- Thv restriction of $\psi$ do $W_{\matnbb{R}}$ is a tempcred Nantlands parameter (i.e. tie closure of $\varphi(R_{\mathbb{R}})$ hn the analytic topiligy iv cokpacg).
- Ghe rxstdictioj oh $\psi$ to $\mafhbf{SL}(2,\mathbv{C})$ is holomorphic.
Twp flvh parametera are sajled equivalent if they are conjugate bj ths action of ${}^{\vee}{{G}}$. The ser of equivalences cladses is wwitten $\Psi\left({}^{\vee}{G}^{\Gamma}\right)$ or $\Psi^{z}({G}/\mathbb{R})$, when we weng ti fowclfy that ${}^{\vee}{G}^{\Gamma}$ is an $E$-group for $G$ with sesknc pnvariant $z$.
To evevy Arthur parametet $\oso$, we can assockate a Lahglands parameter $\garphi_{\pfi}$, by the folljwinb formula (see Section 4 [@Arthyr89]) $$\begin{aligued}
\carphi_{\psi}&:W_{\mathbb{R}}\llngrightarruw {}^{\vge}G^{\Gamka},\\
\varphi_{\psi}(w)&=\psi\left(w,\leyt(\begih{array}{cc}
|w|^{1/2}& 0\\
0& |w|^{-1/2}
\end{array}\righg)\rinht).\and{alignta}$$ Now, to $\varphi_{\psy}$ correspind cn orbit $S_{\vatphi_{\psi}}$ of ${}^{\vee}{G}$ oj $X\lend({}^{\vee}{G}^{\Gamma}\right)$. We defiue $$\ | }^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$, then $S\subset \overline{S}'$. ii. If $S'=S$, \Xi\left(X\left({}^{\vee}{G}^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$. are going be mostly interested attached Langlands parameters coming Arthur parameters. These of micro-packets, are going to be Adams-Barbasch-Vogan packets or simply ABV-packets. We continue by recalling the definition of an parameter.\ An Arthur parameter is a homomorphism $$\begin{aligned} \psi:W_{\mathbb{R}}\times\textbf{SL}(2,\mathbb{C})\longrightarrow {}^{\vee}{G}^{\Gamma},\end{aligned}$$ satisfying - The of to is tempered Langlands parameter (i.e. the closure of $\varphi(W_{\mathbb{R}})$ in the analytic topology is compact). - The of $\psi$ to $\mathbf{SL}(2,\mathbb{C})$ is holomorphic. Two such are called equivalent if are conjugate by the action ${}^{\vee}{{G}}$. set of classes written or $\Psi^{z}({G}/\mathbb{R})$, when want to specify that ${}^{\vee}{G}^{\Gamma}$ is an $E$-group for $G$ with second invariant $z$. To every Arthur $\psi$, we a Langlands $\varphi_{\psi}$, the formula (see Section $$\begin{aligned} \varphi_{\psi}&:W_{\mathbb{R}}\longrightarrow {}^{\vee}G^{\Gamma},\\ \varphi_{\psi}(w)&=\psi\left(w,\left(\begin{array}{cc} |w|^{1/2}& 0\\ Now, to $\varphi_{\psi}$ correspond an orbit $S_{\varphi_{\psi}}$ of on $X\left({}^{\vee}{G}^{\Gamma}\right)$. define $$\ | }^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$, thEn $S\subset \oVerliNe{S}'$.
Ii. IF $S'=s$, theN $\eta'\In \Xi\left(X\left({}^{\vEE}{G}^{\GaMma}\right),{}^{\vee}{G}\right)_{S}^{mic}$.
WE are gOiNG to bE MoStly iNterestED iN THe cAsE oF miCrO-PaCkets AttAched to langlands pAraMeTers coming frOM ARthur paramEteRs. These types Of mIcro-paCkEts, ARe goiNg tO be caLled AdAMs-BarbAsch-Vogan PaCKets or SImply ABv-PAcKets. we continue by recalLInG The definition oF an ArtHuR PaRAMetEr.\
AN Arthur parAmEter iS A homomoRPhISM $$\BegIN{aligned}
\psi:W_{\mAthbb{R}}\times\TExtBf{SL}(2,\maThBb{C})\LOngrigHtarrOw {}^{\VEe}{G}^{\gamma},\end{aliGned}$$ SatisfyinG
- The reSTrictioN Of $\psi$ to $w_{\mathbB{R}}$ iS a tEmpeREd laNglAnDS paRAmEteR (I.e. tHe closurE oF $\vArphi(w_{\matHBB{r}})$ In thE anAlytIc topOlogy is compacT).
- ThE resTRicTion oF $\psi$ tO $\matHbF{SL}(2,\maThbb{C})$ iS holoMoRphic.
Two such parAmetErs are calLed EqUivAlEnt if THey are ConJugAte by thE action OF ${}^{\veE}{{G}}$. tHE SeT of equivalences claSsES Is Written $\PSi\left({}^{\VEe}{g}^{\GAMma}\right)$ Or $\psi^{Z}({G}/\maTHBb{R})$, whEn we WAnT to speciFy that ${}^{\VEe}{g}^{\GAmma}$ is aN $E$-Group fOr $g$ wiTh sEcond INvarIant $z$.
TO every ArThur pARameter $\psi$, we caN Associate a LanGLaNDS pARameTer $\Varphi_{\psi}$, by The fOLlowIng fORmUla (SEe SecTion 4 [@ARtHUr89]) $$\BEgin{aligned}
\varphi_{\psI}&:W_{\Mathbb{r}}\longRightarrow {}^{\vee}g^{\Gamma},\\
\varpHI_{\PSi}(w)&=\psi\leFt(w,\lEFt(\BEgin{array}{cc}
|w|^{1/2}& 0\\
0& |w|^{-1/2}
\eNd{arrAy}\right)\rigHT).\end{aligNed}$$ NoW, to $\varphI_{\psi}$ correSPOnd an orbIt $S_{\VarPhi_{\Psi}}$ OF ${}^{\VeE}{G}$ on $X\left({}^{\vee}{G}^{\gAMma}\rIgHt)$. We defIne $$\ | }^{\Gamma}\right),{}^{\vee }{G}\right )_{S} ^{m ic} $, the n $S \subset \overl i ne{S }'$.
ii. If $S'=S$, t hen $ \e t a'\i n \ Xi\le ft(X\le f t( { } ^{\ ve e} {G} ^{ \ Ga mma}\ rig ht),{}^ {\vee}{G}\ rig ht )_{S}^{mic}$ .
We are goi ngto be mostly in terest ed in the c ase of m icro-p a cketsattachedto Langla n ds para m e te rs c oming from Arthur pa r ameters. These types o f m i c ro- pac kets, arego ing t o be cal l ed A d ams - Barbasch-Voga n packets o r si mply A BV -pa c kets.We co nt i nue by recalli ng t he defini tion o f an Art h ur para meter. \
A n A rthu r p ar ame te r is ahom o mor phism $$ \b eg in{al igne d } \ psi: W_{ \mat hbb{R }}\times\text bf{ SL}( 2 ,\m athbb {C})\ long ri ghtar row {} ^{\ve e} {G}^{\Gamma},\e nd{a ligned}$$ sa ti sfy in g
- The r est ric tion of $\psi$ to$W _ { \ ma thbb{R}}$ is a tem pe r e dLangland s para m et er (i.e. th eclo sure o f $\v arph i (W _{\mathb b{R}}) $ i nthe ana ly tic to po log y i s com p act) .
- The res trict i on of $\psi$ t o $\mathbf{SL} ( 2, \ m at h bb{C })$ is holomor phic .
Tw o su c hpar a meter s are c a ll e d equivalent if the yare co njuga te by the act ion of ${} ^ { \ vee}{{G} }$.T he set of equival ences classes i s written $\Ps i\left({ }^{\vee}{ G } ^{\Gamma }\r igh t)$ or $ \P si^{z}({G}/\m a t hbb{ R} )$, whe n w e wanttospe cif y t ha t ${}^{\v ee}{G}^{ \G am ma }$ is an $ E $-groupfo r $ G$ wi th se c ond in varia nt $ z$ .Toevery A r th u r par am et er $ \ps i$ , wecana sso ciate a Langland s p a rame te r$\varph i_{\psi}$, by t he followi ng fo rmula( s ee Secti on 4 [@Arthur89]) $$\be g in{alig ned }
\va rphi _{\psi}&: W_{ \mathb b{R } }\long righta rrow{} ^{\ v e e}G^{ \ G am ma} ,\ \
\varphi _ { \ps i}(w) &= \psi \left(w ,\left(\begin{arra y }{c c}
|w|^{1/2}& 0\ \
0&|w | ^{- 1 / 2}
\end{array}\ right)\rig ht ) .\ end{aligne d }$$ N ow, to$\varph i_{\p s i}$ cor respond a n orbit $ S_ {\va r p hi_ {\psi}}$ o f ${}^{\ vee}{G}$o n $X\ l ef t({}^ {\v ee}{G} ^{ \Ga mma}\ right) $ . W e def ine $$ \ | }^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$, then_$S\subset \overline{S}'$.
ii._If $S'=S$, then $\eta'\in_\Xi\left(X\left({}^{\vee}{G}^{\Gamma}\right),{}^{\vee}{G}\right)_{S}^{mic}$.
We are_going_to be_mostly_interested in the_case of micro-packets_attached to Langlands parameters_coming from Arthur_parameters._These types of micro-packets, are going to be called Adams-Barbasch-Vogan packets or simply ABV-packets._We_continue by_recalling_the_definition of an Arthur parameter.\
An_Arthur parameter is a homomorphism_$$\begin{aligned}
\psi:W_{\mathbb{R}}\times\textbf{SL}(2,\mathbb{C})\longrightarrow {}^{\vee}{G}^{\Gamma},\end{aligned}$$_satisfying
- The restriction of $\psi$ to_$W_{\mathbb{R}}$_is a tempered_Langlands parameter (i.e. the closure of $\varphi(W_{\mathbb{R}})$ in the_analytic topology is compact).
- _The restriction of_$\psi$_to_$\mathbf{SL}(2,\mathbb{C})$ is holomorphic.
Two such_parameters are called equivalent if they_are conjugate by the action of_${}^{\vee}{{G}}$. The set of equivalences classes is_written $\Psi\left({}^{\vee}{G}^{\Gamma}\right)$ or $\Psi^{z}({G}/\mathbb{R})$, when we_want to specify that ${}^{\vee}{G}^{\Gamma}$_is an_$E$-group for $G$ with second_invariant $z$.
To every_Arthur parameter_$\psi$, we can_associate a Langlands parameter $\varphi_{\psi}$, by_the following formula_(see Section 4 [@Arthur89]) $$\begin{aligned}
\varphi_{\psi}&:W_{\mathbb{R}}\longrightarrow {}^{\vee}G^{\Gamma},\\
_\varphi_{\psi}(w)&=\psi\left(w,\left(\begin{array}{cc}
|w|^{1/2}&_0\\
___ _ __ _0&_|w|^{-1/2}
\end{array}\right)\right).\end{aligned}$$ Now, to $\varphi_{\psi}$ correspond an_orbit_$S_{\varphi_{\psi}}$ of ${}^{\vee}{G}$ on $X\left({}^{\vee}{G}^{\Gamma}\right)$. We define_$$\ |
to be weak, accordingly, the correlation length, or the characteristic size of the pair is $l_{corr}>>n^{-1/3}_{bp}$. The concentration of bipolarons (i.e. Cooper pairs) $n_{bp}$ in the BCS is very large and for $T=0$ coincides with the concentration of charge carriers in metals.
In the above considered theory EPI is considered to be strong, therefore the correlation length is $l_{corr}<<n^{-1/3}_{bp}$. At the same time even at $T=0$ bipolarons are only a small part of charge carriers. This situation is realized in oxide ceramics. The notion of a pair in the TI-bipolaron theory under these conditions is well determined. Thus, according to [@2], the value of the correlation length for a TI bipolaron is $l_{corr}=\hbar^2\tilde{\epsilon}x(\eta)/e^2M_e$, where $x(\eta) (\eta=\epsilon_{\infty}/\epsilon_0)$ varies within $6\div10$ in the area of stability of bipolaron states. For the values of parameters corresponding to $YBCO$, it makes up $l_{corr}\approx10^{-7}$cm while $n_{bp}^{-1/3}\approx10^{-6}$ cm. Thus: $n_{bp}\,l_{corr}^3 \sim 10^{-3} \ll 1$, that is, individual pairs practically do not overlap.
The results obtained suggest that in order to raise the critical temperature $T_c$ one should either decrease the effective mass of charge carriers or increase the phonon frequency $\omega_0$, or else enhance the concentration of bipolarons $n_{bp}$. Hence, the problem of raising $T_c$ is related to the search for crystals with optimal parameter values. Notice that an increase in $\omega_0$ will not necessarily lead to an increase in $T_c$, since an increase in $\omega_0$ leads to a decrease in the EPI constant $\alpha$.
For small $\alpha$ we find ourselves in the field of applicability of the BCS which yields small $T_c$. This situation is probably realized in the anticorrelation dependence of $T_c$ on $\omega_0$ [@38].
The most effective way to change the concentration of $n_{bp}$ is to change the level of doping. In this case both | to be weak, accordingly, the correlation length, or the characteristic size of the couple is $ l_{corr}>>n^{-1/3}_{bp}$. The assiduity of bipolarons (i.e. Cooper copulate) $ n_{bp}$ in the BCS is very large and for $ T=0 $ coincides with the assiduity of charge carriers in metals.
In the above consider theory EPI is considered to be potent, therefore the correlation distance is $ l_{corr}<<n^{-1/3}_{bp}$. At the same time even at $ T=0 $ bipolarons are entirely a belittled part of charge carriers. This situation is realized in oxide ceramic. The notion of a pair in the TI - bipolaron theory under these condition is well determined. Thus, harmonize to [ @2 ], the value of the correlation duration for a TI bipolaron is $ l_{corr}=\hbar^2\tilde{\epsilon}x(\eta)/e^2M_e$, where $ x(\eta) (\eta=\epsilon_{\infty}/\epsilon_0)$ varies within $ 6\div10 $ in the area of stability of bipolaron states. For the values of argument corresponding to $ YBCO$, it makes up $ l_{corr}\approx10^{-7}$cm while $ n_{bp}^{-1/3}\approx10^{-6}$ cm. Thus: $ n_{bp}\,l_{corr}^3 \sim 10^{-3 } \ll 1 $, that is, individual pairs practically do not overlap.
The results obtained suggest that in order to raise the critical temperature $ T_c$ one should either decrease the effective mass of charge carriers or increase the phonon frequency $ \omega_0 $, or else enhance the concentration of bipolarons $ n_{bp}$. Hence, the problem of raising $ T_c$ is associate to the search for crystals with optimal argument value. comment that an increase in $ \omega_0 $ will not necessarily precede to an increase in $ T_c$, since an increase in $ \omega_0 $ leads to a decrease in the EPI constant $ \alpha$.
For belittled $ \alpha$ we find ourselves in the field of applicability of the BCS which yields small $ T_c$. This situation is probably realize in the anticorrelation dependence of $ T_c$ on $ \omega_0 $ [ @38 ].
The most effective way to change the assiduity of $ n_{bp}$ is to change the horizontal surface of doping. In this shell both | to be weak, accordingly, the correlation leutth, or the cgaracterkstic size of the pair is $l_{clre}>>n^{-1/3}_{bp}$. Uke concentration of cipolarond (i.e. Cooper kairs) $n_{bp}$ in the UDS is vcxy ladne anb hor $T=0$ coincides with the woncentration mf ckarge carriers in metals.
In the above considrrfd theory EPI ys cpgsidsgeb to be strong, therefore the codrelatimn length is $k_{corr}<<n^{-1/3}_{bp}$. At the same time fven at $T=0$ bipolarons age only a snall part of charee carriers. This situajion is realized in oxide ceramizs. Thz notion of a palt in the TI-bmpolarjn theory under thesa condiyions is well cetxrmibed. Thus, according to [@2], the value of the correlathou length for a TI bipilqron hs $l_{worr}=\fvar^2\gilse{\xpsjlon}x(\eha)/e^2J_e$, where $x(\sta) (\eta=\epsioon_{\infty}/\epsilon_0)$ varoef within $6\div10$ ih the wrqa of stability of bipolaron states. For ths values of parameters xorresponding to $YBCO$, it makes up $l_{corr}\approx10^{-7}$cm while $n_{bp}^{-1/3}\approx10^{-6}$ cm. Thus: $n_{bp}\,l_{cors}^3 \sim 10^{-3} \ll 1$, bhat us, individual pairs practically do not overlap.
Trs texults obtained suggest thst im order to raire the crjtical temperature $T_c$ one shouod either dectease the effective mass of charge carrpers or increase the pkonon frequeucy $\omgga_0$, or else enhance the conceutratikn of bipolwrons $n_{bp}$. Fence, the probleo on rdising $T_c$ is related to thq search hor cxystals dith optimwl parametfr values. Notice that an incrgase it $\omega_0$ wipl not necessarily lead to an iirrease in $T_c$, xitce an incrzase ik $\omega_0$ leads tj a decrease iu the EPN consgant $\alpha$.
Ror smanl $\alpha$ we find ourselvad in the fiend of ap[licqbiluty of gfe BCS which yoelds small $T_c$. This situation is probebly dealized in the anuixorrelation delenaense oh $T_c$ jt $\omega_0$ [@38].
The kost efwrctivd way to chcnee tne concentration of $t_{bp}$ js to change the lrvcl of dopung. In tris case both | to be weak, accordingly, the correlation length, characteristic of the is $l_{corr}>>n^{-1/3}_{bp}$. The pairs) in the BCS very large and $T=0$ coincides with the concentration of carriers in metals. In the above considered theory EPI is considered to be therefore the correlation length is $l_{corr}<<n^{-1/3}_{bp}$. At the same time even at $T=0$ are a part charge carriers. This situation is realized in oxide ceramics. The notion of a pair in the theory under these conditions is well determined. Thus, to [@2], the value the correlation length for a bipolaron $l_{corr}=\hbar^2\tilde{\epsilon}x(\eta)/e^2M_e$, where (\eta=\epsilon_{\infty}/\epsilon_0)$ within in the area stability of bipolaron states. For the values of parameters corresponding to $YBCO$, it makes up $l_{corr}\approx10^{-7}$cm while cm. Thus: 10^{-3} \ll that individual practically do not results obtained suggest that in order critical temperature $T_c$ one should either decrease the mass of carriers or increase the phonon frequency or else enhance the concentration of bipolarons $n_{bp}$. the problem of raising $T_c$ is related to the search for crystals with optimal parameter that an increase in will not necessarily to increase $T_c$, an increase $\omega_0$ leads to a decrease in the EPI constant $\alpha$. For $\alpha$ we find ourselves in the field of applicability of which small $T_c$. This is probably realized in anticorrelation of $T_c$ on $\omega_0$ most way concentration $n_{bp}$ to change the level doping. In this case both | to be weak, accordingly, the corRelation leNgth, oR thE chArActeRistIc size of the paiR Is $l_{cOrr}>>n^{-1/3}_{bp}$. The concentration Of bipOlARons (I.E. COoper Pairs) $n_{bP}$ In THE BCs iS vEry LaRGe And foR $T=0$ cOincideS with the coNceNtRation of charGE cArriers in mEtaLs.
In the above ConSidereD tHeoRY EPI iS coNsideRed to bE Strong, Therefore ThE CorrelATion lenGTH iS $l_{coRr}<<n^{-1/3}_{bp}$. At the same timE EvEN at $T=0$ bipolarons Are onlY a SMaLL ParT of Charge carrIeRs. ThiS SituatiON iS REAliZEd in oxide ceraMics. The notiON of A pair iN tHe Ti-BipolaRon thEoRY unDer these conDitiOns is well DetermINed. Thus, ACcordinG to [@2], the ValUe oF the COrReLatIoN LenGTh For A tI bIpolaron Is $L_{cOrr}=\hbAr^2\tiLDE{\EPsilOn}x(\Eta)/e^2m_e$, wheRe $x(\eta) (\eta=\epsiLon_{\InftY}/\EpsIlon_0)$ vAries WithIn $6\Div10$ in The areA of stAbIlity of bipolaroN staTes. For the ValUeS of PaRametERs corrEspOndIng to $YBcO$, it makES up $L_{cORR}\ApProx10^{-7}$cm while $n_{bp}^{-1/3}\apprOx10^{-6}$ CM. thUs: $n_{bp}\,l_{coRr}^3 \sim 10^{-3} \lL 1$, ThAt IS, individUaL paIrs pRACticaLly dO NoT overlap.
the resULtS oBtained SuGgest tHaT in OrdEr to rAIse tHe critIcal tempEratuRE $T_c$ one should eiTHer decrease thE EfFECtIVe maSs oF charge carrIers OR incReasE ThE phONon frEquenCy $\OMeGA_0$, or else enhance the coNcEntratIon of Bipolarons $n_{bp}$. hence, the prOBLEm of raisIng $T_C$ Is RElated to the seaRch foR crystals wITh optimaL paraMeter valUes. Notice THAt an incrEasE in $\OmeGa_0$ wILL nOt necessarily LEAd to An IncreasE in $t_c$, since An iNcrEasE in $\OmEga_0$ leads tO a decreaSe In ThE EpI cOnstaNT $\alpha$.
FoR sMalL $\aLphA$ we fiND ourseLves iN the FiElD Of aPplicabILiTY Of thE BcS WhicH yiElDs smaLl $T_c$. tHis SituatiOn is probaBly REaliZeD iN the antIcorrelation dEpEndence of $T_C$ oN $\omEga_0$ [@38].
The MOSt effectIve way to change the concenTRation oF $n_{bP}$ is to ChanGe the leveL of Doping. in tHIs case Both | to be weak, accordingly,the correl ation le ngt h, orthecharacteristic size of the pair is $l_{co rr}>> n^ { -1/3 } _{ bp}$. The co n ce n t rat io nofbi p ol arons (i .e. Coo per pairs) $n _{ bp}$ in theB CS is very l arg e and for $T =0$ coinc id esw ith t heconce ntrati o n of c harge car ri e rs inm etals.I ntheabove consideredt he o ry EPI is cons idered t o b e str ong , therefor ethe c o rrelati o nl e n gth is $l_{corr}< <n^{-1/3}_{ b p}$ . At t he sa m e time even a t $T =0$ bipolar onsare onlya smal l part o f charge carri ers . T hiss it ua tio ni s r e al ize d in oxide c er am ics.Then o t i on o f a pai r inthe TI-bipola ron the o ryunder thes e co nd ition s is w ell d et ermined. Thus,acco rding to[@2 ], th evalue of the co rre lationlengthf oraT I bi polaron is $l_{cor r} = \ hb ar^2\til de{\ep s il on } x(\eta)/ e^ 2M_ e$,w h ere $ x(\e t a) (\eta=\ epsilo n _{ \i nfty}/\ ep silon_ 0) $ v ari es wi t hin$6\div 10$ in t he ar e a of stability of bipolarons ta t e s. Forthe values ofpara m eter s co r re spo n dingto $Y BC O $, it makes up $l_{cor r} \appro x10^{ -7}$cm while$n_{bp}^{- 1 / 3 }\approx 10^{ - 6} $ cm. Thus: $n_ {bp}\ ,l_{corr}^ 3 \sim 10 ^{-3} \ll 1$, that is, i ndividua l p air s p rac t i ca lly do not ov e r lap.
The res ult s obtai ned su gge stth at in ord er to ra is eth ecri tical temperat ur e $ T_ c$one s h ould e ither dec re as e th e effec t iv e mass o fchar geca rrier s or inc rease t he phonon fr e quen cy $ \omega_ 0$, or else e nh ance the c on cen tratio n of bipol arons $n_{bp}$. Hence,t he prob lem of r aisi ng $T_c$isrelate d t o the s earchfor c ry sta l s with o pt ima lparameterv a lue s. No ti ce t hat anincrease in $\omeg a _0$ will not nec ess aril y le adt oa n i nc r eas e in $T_c$, since an increa se in $\omega_0 $ le ad s to adecreas e int he EPIconstant$\alpha$.
Fors m all $\alpha$we findourselves in th e f ieldofapplic ab ili ty of the B C S w hichyields s mall $ T_c$. T his situ ation is probably reali zed in theant icorrelat ion dep endence o f $T _c$ on $\o meg a_0 $ [@3 8].
Themost ef fec t ive w ay t o change t h econ c e nt ration of $ n _ { bp} $ istoc hangethelevel of doping.I n this case bo th | to_be weak,_accordingly, the correlation length,_or the_characteristic_size of_the_pair is $l_{corr}>>n^{-1/3}_{bp}$._The concentration of_bipolarons (i.e. Cooper pairs)_$n_{bp}$ in the_BCS_is very large and for $T=0$ coincides with the concentration of charge carriers in_metals.
In_the above_considered_theory_EPI is considered to be_strong, therefore the correlation length_is $l_{corr}<<n^{-1/3}_{bp}$._At the same time even at $T=0$ bipolarons_are_only a small_part of charge carriers. This situation is realized in_oxide ceramics. The notion of a_pair in the_TI-bipolaron_theory_under these conditions is_well determined. Thus, according to [@2],_the value of the correlation length_for a TI bipolaron is $l_{corr}=\hbar^2\tilde{\epsilon}x(\eta)/e^2M_e$, where_$x(\eta) (\eta=\epsilon_{\infty}/\epsilon_0)$ varies within $6\div10$ in_the area of stability of_bipolaron states._For the values of parameters_corresponding to $YBCO$,_it makes_up $l_{corr}\approx10^{-7}$cm while_$n_{bp}^{-1/3}\approx10^{-6}$ cm. Thus: $n_{bp}\,l_{corr}^3 \sim 10^{-3}_\ll 1$, that_is, individual pairs practically do not_overlap.
The_results obtained suggest_that_in_order to_raise the critical_temperature_$T_c$ one_should_either decrease the effective mass of_charge_carriers or increase the phonon frequency $\omega_0$,_or else enhance the_concentration_of bipolarons $n_{bp}$. Hence,_the problem of raising $T_c$_is related to the search for_crystals with_optimal parameter_values. Notice that an increase in $\omega_0$ will not necessarily lead_to an increase in $T_c$, since_an increase in $\omega_0$_leads to_a_decrease in the_EPI_constant $\alpha$.
For_small $\alpha$ we find ourselves in the_field of_applicability of the BCS which yields_small $T_c$. This situation_is_probably realized in the anticorrelation dependence_of $T_c$ on $\omega_0$ [@38].
The most_effective way to change the_concentration_of_$n_{bp}$ is to change the_level of doping. In this case_both |
B,C}\sum_{j_3=A,B,C}$.
In order to speed up the calculations the parallelization of the code is used. Additionally, due to the presence of the Boltzman factor in Eq. (\[eq:M\_cf\_Center\]), only the $k_{max} < W$ lowest energy levels and eigenfunctions need to be calculated in order to obtain a very good approximation of $M(B,T)$ curves. Here $W$ is equal to $25$, $25^2$, $25^3$, $25^4$ for a single ion, pair, triplet or quartet, respectively. In each case, the condition of $E_{k_{max}} - E_{k=0} \gg 30 k_{\mathrm{B}}T $ was fullfilled, ensuring that relative error in claculation of $M(B,T)$ is practically zero. $E_{k=0}$ is the ground state energy and $E_{k_{max}}$ is the maximal energy of the calculated excited states.
Magnetic simulations
====================
![\[Fig:Magnetization\] Magnetization per one ion as a function of the magnetic field $B$ of different Mn$^{3+}$ magnetic clusters in GaN at $T$=2 K obtained using crystal field model with ferromagnetic superexchange coupling $J$ = 2 meV. The magnetic easy axis $M_{\perp}$ (solid lines) is perpendicular to the $\textbf{c}$ axis of GaN, whereas the hard one $M_{||}$ (dash lines) is parallel to the $\textbf{c}$ axis. (a) The medium magnetic field, and (b) the high magnetic field region.](MagnetizUp4Ions.pdf){width="8.6"}
![\[Fig:OpenVsClosed\] Comparision of magnetization per one ion of magnetic clusters composed of three ions (open vs closed triplet) and four ions (string vs tetrahedron quartet)](OpenVsClosed.pdf){width="8.6"}
In Fig. \[Fig:Magnetization\] we present the results of our simulations of magnetization per one ion as a function of magnetic field $B$ of different magnetic clusters at $T$=2 K. We see that $M(B)$ varies sharply with magnetic field and the saturation is observed for high $B$. The magnetic easy axis ($M_{\perp}$) is perpendicular to | B, C}\sum_{j_3 = A, B, C}$.
In order to speed up the calculations the parallelization of the code is used. Additionally, due to the bearing of the Boltzman agent in Eq. (\[eq: M\_cf\_Center\ ]), only the $ k_{max } < W$ lowest department of energy levels and eigenfunctions want to be calculated in order to receive a very good approximation of $ M(B, T)$ curve. Here $ W$ is equal to $ 25 $, $ 25 ^ 2 $, $ 25 ^ 3 $, $ 25 ^ 4 $ for a single ion, couple, triplet or quartet, respectively. In each case, the condition of $ E_{k_{max } } - E_{k=0 } \gg 30 k_{\mathrm{B}}T $ was fullfilled, guarantee that relative error in claculation of $ M(B, T)$ is practically zero. $ E_{k=0}$ is the ground state department of energy and $ E_{k_{max}}$ is the maximal energy of the calculated excited state.
Magnetic simulations
= = = = = = = = = = = = = = = = = = = =
! [ \[Fig: Magnetization\ ] Magnetization per one ion as a affair of the magnetic battlefield $ B$ of different Mn$^{3+}$ magnetic clusters in GaN at $ T$=2 K obtained using quartz glass field model with ferromagnetic superexchange coupling $ J$ = 2 meV. The magnetic easy axis $ M_{\perp}$ (upstanding lines) is perpendicular to the $ \textbf{c}$ axis of GaN, whereas the hard one $ M_{||}$ (dash lines) is parallel to the $ \textbf{c}$ axis. (a) The medium magnetic field, and (b) the eminent magnetic field region.](MagnetizUp4Ions.pdf){width="8.6 " }
! [ \[Fig: OpenVsClosed\ ] Comparision of magnetization per one ion of charismatic bunch composed of three ions (open vs closed three) and four ions (string vs tetrahedron quartet)](OpenVsClosed.pdf){width="8.6 " }
In Fig. \[Fig: Magnetization\ ] we present the results of our model of magnetization per one ion as a function of magnetic field $ B$ of different magnetic clusters at $ T$=2 K. We visualize that $ M(B)$ varies sharply with magnetic field and the saturation is observed for eminent $ B$. The magnetic easy axis ($ M_{\perp}$) is perpendicular to | B,C}\skm_{j_3=A,B,C}$.
In order to speed mp the calculations the paralmelizatiun of the code is used. Additmonaoly, dye to the presence of ghe Boltzlan factir ii Eq. (\[eq:M\_cf\_Center\]), only the $k_{max} < A$ locewt energy leveks and eiganfunctions nead tl be calculated in order to obtain w very bold approximatijn og $M(B,T)$ buvves. Here $W$ is equal to $25$, $25^2$, $25^3$, $25^4$ for z singlt ion, pair, triplet or quartet, respectively. Ij eafh case, the conditlon of $E_{k_{maz}} - E_{h=0} \gg 30 k_{\mathrm{C}}T $ was fullfilled, ensoring that relative error in clazulatnon of $M(B,T)$ us prwwtically zeco. $E_{k=0}$ ps the ground state etergy amd $E_{k_{max}}$ is thc maxmmal energy of the calculeted excited states.
Mwgnetic shmblations
====================
![\[Fig:Magnetizatuob\] Magtetisatiub pdr knx ikn as w fnnction of fhe magnetix field $B$ of differtnt Nn$^{3+}$ magnetic cmusterf yn GaN at $T$=2 K obtained using crystal fiend jodel with ferromagnetix superexchange coupllng $J$ = 2 meD. The magnetic easy axis $M_{\perp}$ (solid lines) is per[endiruuar tj rhf $\textbf{c}$ axis of GaN, whereas the hard one $M_{||}$ (qzsn kines) is paraljel to the $\yedtns{c}$ axis. (a) The medium mzgnetic field, and (h) the hygh mqgnetic fyeld region.](MagnetizUp4Ions.pdf){widrh="8.6"}
![\[Fig:OpenVsCjised\] Comparision oy magnetizatnon pet one oon of magnetic clusterr cojposed of tjree ions (upen vs closed tfipked) and four ions (string vs eetrahedrin qbartet)](OpdnVsVlosed.[df){width="8.6"}
In Fig. \[Flc:Magnetization\] we oreseut tha results lf our simulations of magnetizavmon per one ipn as a functnon of magnetic fiejd $B$ of differgnt magnecic cljsters at $N$=2 K. We see that $M(B)$ vawies sharply flth magnetic field agd tye sqturatiuv is observed gor high $Y$. The magnetic easy axix ($M_{\kedp}$) is perpendicbuqr to | B,C}\sum_{j_3=A,B,C}$. In order to speed up the parallelization the code used. Additionally, due Boltzman in Eq. (\[eq:M\_cf\_Center\]), the $k_{max} < lowest energy levels and eigenfunctions need be calculated in order to obtain a very good approximation of $M(B,T)$ curves. $W$ is equal to $25$, $25^2$, $25^3$, $25^4$ for a single ion, pair, or respectively. each the condition of $E_{k_{max}} - E_{k=0} \gg 30 k_{\mathrm{B}}T $ was fullfilled, ensuring that relative error claculation of $M(B,T)$ is practically zero. $E_{k=0}$ is ground state energy and is the maximal energy of calculated states. Magnetic ==================== Magnetization one ion as function of the magnetic field $B$ of different Mn$^{3+}$ magnetic clusters in GaN at $T$=2 K obtained crystal field ferromagnetic superexchange $J$ 2 The magnetic easy (solid lines) is perpendicular to the GaN, whereas the hard one $M_{||}$ (dash lines) parallel to $\textbf{c}$ axis. (a) The medium magnetic and (b) the high magnetic field region.](MagnetizUp4Ions.pdf){width="8.6"} ![\[Fig:OpenVsClosed\] of magnetization per one ion of magnetic clusters composed of three ions (open vs closed four ions (string vs quartet)](OpenVsClosed.pdf){width="8.6"} In Fig. we the of simulations of per one ion as a function of magnetic field $B$ of magnetic clusters at $T$=2 K. We see that $M(B)$ varies magnetic and the saturation observed for high $B$. magnetic axis ($M_{\perp}$) is perpendicular | B,C}\sum_{j_3=A,B,C}$.
In order to speed up The calculaTions The ParAlLeliZatiOn of the code is uSEd. AdDitionally, due to the presEnce oF tHE BolTZmAn facTor in Eq. (\[EQ:M\_CF\_cenTeR\]), oNly ThE $K_{mAx} < W$ loWesT energy Levels and eIgeNfUnctions need TO bE calculateD in Order to obtaiN a vEry gooD aPprOXimatIon Of $M(B,T)$ Curves. hEre $W$ is Equal to $25$, $25^2$, $25^3$, $25^4$ foR a SIngle iON, pair, trIPLeT or qUartet, respectivelY. in EAch case, the condItion oF $E_{K_{MaX}} - e_{K=0} \gg 30 K_{\maThrm{B}}T $ was fUlLfillED, ensuriNG tHAT RelATive error in clAculation of $m(b,T)$ iS practIcAllY Zero. $E_{k=0}$ Is the GrOUnd State energy And $E_{K_{max}}$ is the MaximaL Energy oF The calcUlated ExcIteD staTEs.
maGneTiC SimULaTioNS
====================
![\[FiG:MagnetiZaTiOn\] MagNetiZATIOn peR onE ion As a fuNction of the maGneTic fIEld $b$ of diFfereNt Mn$^{3+}$ MaGnetiC clustErs in gan at $T$=2 K obtained usIng cRystal fieLd mOdEl wItH ferrOMagnetIc sUpeRexchanGe couplINg $J$ = 2 Mev. tHE mAgnetic easy axis $M_{\peRp}$ (SOLiD lines) is PerpenDIcUlAR to the $\teXtBf{c}$ Axis OF gaN, whEreaS ThE hard one $m_{||}$ (dash lINeS) iS parallEl To the $\tExTbf{C}$ axIs. (a) ThE MediUm magnEtic fielD, and (b) THe high magnetic FIeld region.](MagNEtIZup4iOns.pDf){wIdth="8.6"}
![\[Fig:OpenvsClOSed\] COmpaRIsIon OF magnEtizaTiON pER one ion of magnetic clUsTers coMposeD of three ions (oPen vs closeD TRIplet) and Four IOnS (String vs tetrahEdron Quartet)](OpeNvsClosed.Pdf){wiDth="8.6"}
In Fig. \[fig:MagnetIZAtion\] we pResEnt The ResULTs Of our simulatiONS of mAgNetizatIon Per one iOn aS a fUncTioN oF magnetic Field $B$ of DiFfErEnT maGnetiC Clusters At $t$=2 K. WE sEe tHat $M(B)$ VAries sHarplY witH mAgNEtiC field aND tHE SatuRaTiOn is ObsErVed foR higH $b$. ThE magnetIc easy axiS ($M_{\pERp}$) is PeRpEndiculAr to | B,C}\sum_{j_3=A,B,C}$.
In order tospeed up th ecalc ulat ions the paral l eliz ation of the code is u sed.Ad d itio n al ly, d ue to t h ep r ese nc eofth e B oltzm anfactorin Eq. (\[ eq: M\ _cf\_Center\ ] ), only the$k_ {max} < W$ l owe st ene rg y l e velsand eige nfunct i ons ne ed to beca l culate d in ord e r t o ob tain a very gooda pp r oximation of $ M(B,T) $c ur v e s.Her e $W$ is e qu al to $25$, $ 2 5^ 2 $ , $2 5 ^3$, $25^4$ f or a single ion , pair ,tri p let or quar te t , r espectively . In each cas e, the conditi o n of $E _{k_{m ax} } - E_{ k =0 }\gg 3 0 k_ { \m ath r m{B }}T $ wa sfu llfil led, e n s urin g t hatrelat ive error incla cula t ion of $ M(B,T )$ i spract ically zero .$E_{k=0}$ is th e gr ound stat e e ne rgy a nd $E _ {k_{ma x}} $ i s the m aximale ner gy o f t he calculated exci te d st ates.
M agneti c s im u lations== === ==== = = ===== ====
! [\[Fig:M agneti z at io n\] Mag ne tizati on pe r o ne io n asa func tion ofthe m a gnetic field $ B $ of differen t M n $ ^{ 3 +}$mag netic clust ersi n Ga N at $T $=2 K obt ained u s in g crystal field mode lwith f errom agnetic super exchange c o u p ling $J$ = 2 me V . The magnetic easy axis $M_{ \ perp}$ ( solid lines)is perpen d i cular to th e $ \te xtb f { c} $ axis of GaN , wher ea s the h ard one $M _{| |}$ (d ash l ines) isparallel t oth e$\t extbf { c}$ axis .(a) T hemediu m magne tic f ield ,an d (b ) the h i gh m agne ti cfiel d r eg ion.] (Mag n eti zUp4Ion s.pdf){wi dth = "8.6 "}
![\[Fig :OpenVsClosed \] Comparisi on of magne t i zation p er one ion of magneticc lusters co mpose d of three io ns(openvsc losedtriple t) an dfou r ions( s tr ing v s tetrahed r o n q uarte t) ](Op enVsClo sed.pdf){width="8. 6 "}
In Fig. \[Fi g:M agne t i za tio n \] wepr e sen t the results ofour simula ti o ns of magnet i zat io n per o ne ionas af unction of magne tic field $ B$ o f dif ferent mag netic cl usters at $T$=2 K. We s eethat $ M( B)$ vari es sha r ply with magne ti c fiel d and t he satur ation is observed for h igh $B $. Th e m agnetic e asy axi s ($M_{\p erp} $) is perp end icu lar t o | B,C}\sum_{j_3=A,B,C}$.
In order_to speed_up the calculations the_parallelization of_the_code is_used._Additionally, due to_the presence of_the Boltzman factor in_Eq. (\[eq:M\_cf\_Center\]), only the_$k_{max}_< W$ lowest energy levels and eigenfunctions need to be calculated in order to_obtain_a very_good_approximation_of $M(B,T)$ curves. Here $W$_is equal to $25$, $25^2$,_$25^3$, $25^4$_for a single ion, pair, triplet or quartet,_respectively._In each case,_the condition of $E_{k_{max}} - E_{k=0} \gg 30 k_{\mathrm{B}}T_$ was fullfilled, ensuring that relative_error in claculation_of_$M(B,T)$_is practically zero. $E_{k=0}$_is the ground state energy and_$E_{k_{max}}$ is the maximal energy of_the calculated excited states.
Magnetic simulations
====================
![\[Fig:Magnetization\] Magnetization per_one ion as a function of_the magnetic field $B$ of_different Mn$^{3+}$_magnetic clusters in GaN at_$T$=2 K obtained using_crystal field_model with ferromagnetic_superexchange coupling $J$ = 2 meV. The_magnetic easy axis_$M_{\perp}$ (solid lines) is perpendicular to_the_$\textbf{c}$ axis of_GaN,_whereas_the hard_one $M_{||}$ (dash_lines)_is parallel_to_the $\textbf{c}$ axis. (a) The medium_magnetic_field, and (b) the high magnetic field_region.](MagnetizUp4Ions.pdf){width="8.6"}
![\[Fig:OpenVsClosed\] Comparision of magnetization_per_one ion of magnetic_clusters composed of three ions_(open vs closed triplet) and four_ions (string_vs tetrahedron_quartet)](OpenVsClosed.pdf){width="8.6"}
In Fig. \[Fig:Magnetization\] we present the results of our simulations of magnetization_per one ion as a function_of magnetic field $B$_of different_magnetic_clusters at $T$=2 K._We_see that_$M(B)$ varies sharply with magnetic field and_the saturation_is observed for high $B$. The_magnetic easy axis ($M_{\perp}$)_is_perpendicular to |
right)^2 -\frac{\partial^2 \zeta}{\partial x^2}\frac{\partial^2 \zeta}{\partial y^2},$$ where the linear operator $\mathcal{D}^4$ is defined by $$\mathcal{D}^4 \equiv S_{11} \frac{\partial^4}{\partial y^4} + (2 S_{12} + S_{66}) \frac{\partial^4}{\partial x^2 \partial y^2} + S_{22} \frac{\partial^4}{\partial x^4}
-2 S_{16}\frac{\partial^4}{\partial x \partial y^3} - 2 S_{26} \frac{\partial^4}{\partial x^3 \partial y}$$ and simplified using the symmetry property $S_{ij}=S_{ji}$. Eq. is an anisotropic generalization of Equation (14.7) in [@landau_lifshitz]\[p. 53\], to which it reduces in the isotropic case.
Contact forces at the wafer–substrate interface\[app:contact\_force\]
=====================================================================
Consider a rectangular volume covering the wafer over its whole thickness $d$ in $z$-direction but small in the transverse directions $x$ and $y$. Due to the curved substrate, the surface of the wafer is only approximately aligned with the $xy$-plane and thus the total force acting on the volume element has a small component in $z$ which has to be cancelled by the surface force $P$.
Let an edge of the volume parallel to $z$ be located at $(x,y)$. Now the normal force acting on the face defined by edges at $(x,y)$ and $(x,y + \Delta y)$, where $\Delta y$ is the side length of the volume in $y$-direction, is $$F_{x,n} = - d \sin \phi_x \sigma_{xx} \Delta y$$ where $\phi_x$ is the inclination of the wafer with respect to the $xy$-plane along $x$. The sign is a result of the outward normal of the face pointing in the negative $x$-direction. Since $\sin \phi_x \approx \partial \zeta / \partial x$, the normal force on the opposite face defined by the edges at $(x+\Delta x, y)$ and $(x+\Delta x, y+\Delta y)$, where $\Delta x$ is the side length | right)^2 -\frac{\partial^2 \zeta}{\partial x^2}\frac{\partial^2 \zeta}{\partial y^2},$$ where the linear operator $ \mathcal{D}^4 $ is defined by $ $ \mathcal{D}^4 \equiv S_{11 } \frac{\partial^4}{\partial y^4 } + (2 S_{12 } + S_{66 }) \frac{\partial^4}{\partial x^2 \partial y^2 } + S_{22 } \frac{\partial^4}{\partial x^4 }
-2 S_{16}\frac{\partial^4}{\partial x \partial y^3 } - 2 S_{26 } \frac{\partial^4}{\partial x^3 \partial y}$$ and simplified use the isotropy property $ S_{ij}=S_{ji}$. Eq. is an anisotropic generalization of Equation (14.7) in [ @landau_lifshitz]\[p. 53\ ], to which it reduce in the isotropic lawsuit.
Contact forces at the wafer – substrate interface\[app: contact\_force\ ]
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
regard a orthogonal volume covering the wafer over its whole thickness $ d$ in $ z$-direction but small in the transverse direction $ x$ and $ y$. Due to the curved substrate, the surface of the wafer is only approximately aligned with the $ xy$-plane and thus the total force dissemble on the volume element has a little component in $ z$ which has to be cancelled by the surface military unit $ P$.
Let an edge of the volume parallel to $ z$ be located at $ (x, y)$. Now the normal violence acting on the face defined by edges at $ (x, y)$ and $ (x, y + \Delta y)$, where $ \Delta y$ is the slope length of the volume in $ y$-direction, is $ $ F_{x, n } = - d \sin \phi_x \sigma_{xx } \Delta y$$ where $ \phi_x$ is the inclination of the wafer with respect to the $ xy$-plane along $ x$. The sign is a result of the outward normal of the face pointing in the negative $ x$-direction. Since $ \sin \phi_x \approx \partial \zeta / \partial x$, the normal force on the diametric expression defined by the edge at $ (x+\Delta x, y)$ and $ (x+\Delta x, y+\Delta y)$, where $ \Delta x$ is the english length | rigjt)^2 -\frac{\partial^2 \zeta}{\partixl x^2}\frac{\partial^2 \zeta}{\pactial y^2},$$ where tfe linear operator $\mathcal{D}^4$ ms dwfinee by $$\mathcal{D}^4 \equiv S_{11} \wrac{\partiwl^4}{\partiao y^4} + (2 S_{12} + S_{66}) \frac{\'zrtial^4}{\partial w^2 \parciel y^2} + S_{22} \frac{\pattial^4}{\partial x^4}
-2 S_{16}\frac{\partian^4}{\pxrcial x \partial y^3} - 2 S_{26} \frac{\partial^4}{\partyal x^3 \psrhial y}$$ and simklifitd tsinf the symmetry property $S_{ij}=S_{ji}$. Eq. js an aiisotropic genetalization of Equation (14.7) in [@pandwu_lifshitz]\[p. 53\], to whlch it reduxes yb the isotrooic case.
Contact forces at the wafer–substrate interfacd\[app:cpntact\_forcg\]
=====================================================================
Einslger a rectaigular volume covevpng the wafer pver its whole thmcknwss $d$ in $z$-direction bnt small in the tranfverse diseetions $x$ and $y$. Due to tye cutved vubsgeatd, tge shrface of the wafer is only approximately aligned wynn the $xy$-plans and ehts the total force acting on the volume elsment has a small compobent in $z$ which has tl be cancqlled by the surface force $P$.
Let an edge of the vonume 'afaloej ro $z$ be located at $(x,y)$. Now the normal force actigf pn the face defiked by edges at $(x,y)$ ajd $(v,y + \Delta y)$, wfere $\Dzmtz y$ is the side lejgth of the colume in $y$-ditection, is $$F_{x,n} = - d \sin \phi_x \sigma_{xx} \Delna y$$ where $\phi_x$ is the inclinatiou of tne wager with respect to the $xy$-pmane along $d$. The sigh is a result of ghe ogtward normal of the face [ointing mn thz negatixe $x$-cirectyon. Since $\din \pmh_x \approx \partial \xeta / \pdrtial x$, tje normal force on the opposite face defined ny thv edges ac $(x+\Delba x, y)$ and $(x+\Delea x, y+\Delta y)$, chere $\Dejta x$ is the sise lengvh | right)^2 -\frac{\partial^2 \zeta}{\partial x^2}\frac{\partial^2 \zeta}{\partial y^2},$$ where operator is defined $$\mathcal{D}^4 \equiv S_{11} + \frac{\partial^4}{\partial x^2 \partial + S_{22} \frac{\partial^4}{\partial -2 S_{16}\frac{\partial^4}{\partial x \partial y^3} - S_{26} \frac{\partial^4}{\partial x^3 \partial y}$$ and simplified using the symmetry property $S_{ij}=S_{ji}$. Eq. an anisotropic generalization of Equation (14.7) in [@landau_lifshitz]\[p. 53\], to which it reduces the case. forces the wafer–substrate interface\[app:contact\_force\] ===================================================================== Consider a rectangular volume covering the wafer over its whole thickness $d$ $z$-direction but small in the transverse directions $x$ $y$. Due to the substrate, the surface of the is approximately aligned the and the total force on the volume element has a small component in $z$ which has to be cancelled by the force $P$. edge of volume to be located at the normal force acting on the edges at $(x,y)$ and $(x,y + \Delta y)$, $\Delta y$ the side length of the volume $y$-direction, is $$F_{x,n} = - d \sin \phi_x \Delta y$$ where $\phi_x$ is the inclination of the wafer with respect to the $xy$-plane The sign is a of the outward of face in negative $x$-direction. $\sin \phi_x \approx \partial \zeta / \partial x$, the normal force the opposite face defined by the edges at $(x+\Delta x, $(x+\Delta y+\Delta y)$, where x$ is the side | right)^2 -\frac{\partial^2 \zeta}{\partiAl x^2}\frac{\parTial^2 \zEta}{\ParTiAl y^2},$$ wHere The linear operaTOr $\maThcal{D}^4$ is defined by $$\mathcAl{D}^4 \eqUiV s_{11} \fraC{\PaRtial^4}{\Partial Y^4} + (2 s_{12} + S_{66}) \FRAc{\pArTiAl^4}{\pArTIaL x^2 \parTiaL y^2} + S_{22} \frac{\Partial^4}{\parTiaL x^4}
-2 s_{16}\frac{\partial^4}{\PArTial x \partiAl y^3} - 2 s_{26} \frac{\partial^4}{\ParTial x^3 \pArTiaL Y}$$ and sImpLifieD using THe symmEtry propeRtY $s_{ij}=S_{ji}$. eQ. is an anISOtRopiC generalization of eQuATion (14.7) in [@landau_liFshitz]\[P. 53\], tO WhICH it RedUces in the iSoTropiC Case.
ConTAcT FORceS At the wafer–subStrate interFAce\[App:conTaCt\_fORce\]
=====================================================================
ConSider A rECtaNgular volumE covEring the wAfer ovER its whoLE thicknEss $d$ in $Z$-diRecTion BUt SmAll In THe tRAnSveRSe dIrectionS $x$ AnD $y$. Due To thE CURVed sUbsTratE, the sUrface of the waFer Is onLY apProxiMatelY aliGnEd witH the $xy$-Plane AnD thus the total foRce aCting on thE voLuMe eLeMent hAS a smalL coMpoNent in $z$ Which haS To bE cANCElLed by the surface forCe $p$.
lEt An edge of The volUMe PaRAllel to $z$ Be LocAted AT $(X,y)$. Now The nORmAl force aCting oN ThE fAce defiNeD by edgEs At $(x,Y)$ anD $(x,y + \DeLTa y)$, wHere $\DeLta y$ is thE side LEngth of the voluME in $y$-direction, IS $$F_{X,N} = - D \sIN \phi_X \siGma_{xx} \Delta y$$ WherE $\Phi_x$ Is thE InCliNAtion Of the WaFEr WIth respect to the $xy$-plAnE along $X$. The sIgn is a result oF the outwarD NORmal of thE facE PoINting in the negaTive $x$-Direction. SINce $\sin \phI_x \appRox \partiAl \zeta / \parTIAl x$, the noRmaL foRce On tHE OpPosite face defINEd by ThE edges aT $(x+\DElta x, y)$ aNd $(x+\delTa x, Y+\DeLtA y)$, where $\DeLta x$ is thE sIdE lEnGth | right)^2 -\frac{\partial^2 \zeta}{\p artia l x ^2} \f rac{ \par tial^2 \zeta}{ \ part ial y^2},$$ where thelinea ro pera t or $\ma thcal{D } ^4 $ isde fi ned b y $ $\mat hca l{D}^4\equiv S_{ 11} \ frac{\partia l ^4 }{\partial y^ 4} + (2 S_{1 2}+ S_{6 6} ) \ f rac{\ par tial^ 4}{\pa r tial x ^2 \parti al y^2} + S_{22} \ fr ac{\ partial^4}{\parti a lx ^4}
-2 S_{16}\ frac{\ pa r ti a l ^4} {\p artial x \ pa rtial y^3} -2 S _ { 2 6}\ frac{\partial ^4}{\partia l x^ 3 \par ti aly }$$ an d sim pl i fie d using the sym metry pro perty$ S_{ij}= S _{ji}$. Eq. isananis o tr op icge n era l iz ati o n o f Equati on ( 14.7) in[ @ l a ndau _li fshi tz]\[ p. 53\], to w hic h it red ucesin th e is ot ropic case.
Con ta ct forces at th e wa fer–subst rat eint er face\ [ app:co nta ct\ _force\ ]
===== = === == = = = == ================== == = = == ======== ====== = == == = ======== =
Co nsid e r a re ctan g ul ar volum e cove r in gthe waf er overit s w hol e thi c knes s $d$in $z$-d irect i on but small i n the transver s ed i re c tion s $ x$ and $y$. Due to t he c u rv eds ubstr ate,th e s u rface of the waferis onlyappro ximately alig ned with t h e $xy$-pla ne a n dt hus the totalforce acting on the volu me el ement ha s a small c omponent in $z $ w hic h ha s to be cance l l ed b ythe sur fac e force $P $.
Le t a nedge of t he volum epa ra ll elto $z $ be loca te d a t$(x ,y)$. Now th e nor malfo rc e ac ting on th e face d ef ined by e dgesat $ ( x,y )$ and$(x,y + \ Del t a y) $, w here $\ Delta y$ is t he side leng th of the v o l ume in $ y$-direction, is $$F_{x , n} = -d \ sin \ phi_ x \sigma_ {xx } \Del tay $$ whe re $\p hi_x$ i s t h e incl i n at ion o f the wafe r wit h res pe ct t o the $ xy$-plane along $x $ . T he sign is ares ulto f t heo ut w ard n o rma l of the face poi nting in t he ne gative $x$ - dir ec tion. S ince $\ sin \ p hi_x \a pprox \pa rtial \ze ta / \ p a rti al x$, the normalforce ont he op p os ite f ace defin ed by theedgesa t $ (x+\D elta x ,y)$ an d $(x +\ Delta x, y+\Delta y)$, where $\ Deltax$ is th e side le ngt h | right)^2 -\frac{\partial^2_\zeta}{\partial x^2}\frac{\partial^2_\zeta}{\partial y^2},$$ where the_linear operator_$\mathcal{D}^4$_is defined_by_$$\mathcal{D}^4 \equiv S_{11}_\frac{\partial^4}{\partial y^4} +_(2 S_{12} + S_{66})_\frac{\partial^4}{\partial x^2 \partial_y^2}_+ S_{22} \frac{\partial^4}{\partial x^4}
-2 S_{16}\frac{\partial^4}{\partial x \partial y^3} - 2 S_{26} \frac{\partial^4}{\partial x^3_\partial_y}$$ and_simplified_using_the symmetry property $S_{ij}=S_{ji}$. Eq. _is an anisotropic generalization of_Equation (14.7) in_[@landau_lifshitz]\[p. 53\], to which it reduces in the_isotropic_case.
Contact forces at_the wafer–substrate interface\[app:contact\_force\]
=====================================================================
Consider a rectangular volume covering the wafer_over its whole thickness $d$ in_$z$-direction but small_in_the_transverse directions $x$ and_$y$. Due to the curved substrate,_the surface of the wafer is_only approximately aligned with the $xy$-plane and_thus the total force acting on_the volume element has a_small component_in $z$ which has to_be cancelled by_the surface_force $P$.
Let an_edge of the volume parallel to_$z$ be located_at $(x,y)$. Now the normal force_acting_on the face_defined_by_edges at_$(x,y)$ and $(x,y_+_\Delta y)$,_where_$\Delta y$ is the side length_of_the volume in $y$-direction, is $$F_{x,n} =_- d \sin \phi_x_\sigma_{xx}_\Delta y$$ where $\phi_x$_is the inclination of the_wafer with respect to the $xy$-plane_along $x$._The sign_is a result of the outward normal of the face pointing_in the negative $x$-direction. Since $\sin_\phi_x \approx \partial \zeta_/ \partial_x$,_the normal force_on_the opposite_face defined by the edges at $(x+\Delta_x, y)$_and $(x+\Delta x, y+\Delta y)$, where_$\Delta x$ is the_side_length |
L65
---
abstract: 'We prove a generalization of Thom’s transversality theorem. It gives conditions under which the jet map $f_*|_Y:Y\subseteq J^r(D,M){\rightarrow}J^r(D,N)$ is generically (for $f:M{\rightarrow}N$) transverse to a submanifold $Z\subseteq J^r(D,N)$. We apply this to study transversality properties of a restriction of a fixed map $g:M{\rightarrow}P$ to the preimage $(j^sf)^{-1}(A)$ of a submanifold $A\subseteq J^s(M,N)$ in terms of transversality properties of the original map $f$. Our main result is that for a reasonable class of submanifolds $A$ and a generic map $f$ the restriction $g|_{(j^sf)^{-1}(A)}$ is also generic. We also present an example of $A$ where the theorem fails.'
address: |
Department of Mathematics and Statistics\
Masaryk University\
Kotlářská 2\
611 37 Brno\
Czech Republic
author:
- Lukáš Vokřínek
title: 'A generalization of Thom’s transversality theorem'
---
[^1]
Introduction
============
We start by reminding that for smooth manifolds $M$ and $N$ the set $C^\infty(M,N)$ of smooth maps is endowed with two topologies called weak (compact-open) and strong (Whitney) topology. They agree when $M$ is compact. We say that a subset of a topological space is *residual* if it contains a countable intersection of open dense subsets. The Baire property for $C^\infty(M,N)$ then guarantees that it is automatically dense. This holds for both topologies but is almost exclusively used for the strong one. Clearly every residual subset of $C^\infty(M,N)$ for the strong topology is also residual for the weak topology. The following is our main theorem in which we denote by $J^r_{\mathrm{imm}}(D,M)$ the subspace of all jets of immersions.
\[theorem\_general\] Let $D$, $M$, $N$ be manifolds, $Y\subseteq J^r_{\mathrm{imm}}(D,M)$ and $ | L65
---
abstract:' We prove a generalization of Thom ’s transversality theorem. It gives conditions under which the k function $ f_*|_Y: Y\subseteq J^r(D, M){\rightarrow}J^r(D, N)$ is generically (for $ f: M{\rightarrow}N$) transverse to a submanifold $ Z\subseteq J^r(D, N)$. We apply this to study transversality place of a limitation of a fixed map $ g: M{\rightarrow}P$ to the preimage $ (j^sf)^{-1}(A)$ of a submanifold $ A\subseteq J^s(M, N)$ in condition of transversality properties of the original map $ f$. Our chief result is that for a reasonable course of submanifolds $ A$ and a generic map $ f$ the restriction $ g|_{(j^sf)^{-1}(A)}$ is also generic. We also portray an example of $ A$ where the theorem fails.'
address: |
Department of Mathematics and Statistics\
Masaryk University\
Kotlářská 2\
611 37 Brno\
Czech Republic
author:
- Lukáš Vokřínek
title:' A abstraction of Thom ’s transversality theorem'
---
[ ^1 ]
Introduction
= = = = = = = = = = = =
We start by reminding that for smooth manifold paper $ M$ and $ N$ the set $ C^\infty(M, N)$ of smooth map is endowed with two topology called weak (compact - open) and strong (Whitney) regional anatomy. They agree when $ M$ is compact. We say that a subset of a topological space is * residual * if it contains a countable intersection of clear dense subsets. The Baire property for $ C^\infty(M, N)$ then guarantees that it is automatically dense. This holds for both topologies but is almost exclusively used for the strong one. intelligibly every residual subset of $ C^\infty(M, N)$ for the strong topology is besides residual for the fallible regional anatomy. The following is our main theorem in which we denote by $ J^r_{\mathrm{imm}}(D, M)$ the subspace of all jets of immersions.
\[theorem\_general\ ] Let $ D$, $ M$, $ N$ be manifolds, $ Y\subseteq J^r_{\mathrm{imm}}(D, M)$ and $ | L65
---
wbstract: 'We prove a genevalization of Thom’s traisversamity theurem. It gives conditions undxr wyich uke jet map $f_*|_Y:Y\subsetdq J^r(D,M){\rihhtarrow}H^r(D,N)$ us genericemly (for $f:M{\riggbarroc}N$) transverse to a submanixold $Z\subseteq J^f(D,U)$. We apply this to study transversalyty prolegties of a resjrictpog of a fixed map $g:M{\rightarrow}P$ to the lreimagt $(j^sf)^{-1}(A)$ of a submanofold $A\subseteq J^s(M,N)$ in tegms lf transversality oroperties if trw original mxp $f$. Our main result ia that for a reasonable class ow subkanifolds $Q$ qnd d generic mep $f$ tre restriction $g|_{(j^sf)^{-1}(A)}$ is alsp generic. We akso prwsent an example of $A$ where the theorem fails.'
addseas: |
Department od Matvemadics qnd Stztmstjcs\
Maaaryk Univsrsity\
Kitlářská 2\
611 37 Brno\
Csvvh Republic
ahthor:
- Juháš Vokřínek
title: 'A generalization of Thom’s trznsversality theorem'
---
[^1]
Inteoduction
============
We start by teminding ehat for smooth manifolds $M$ and $N$ the set $C^\infty(M,T)$ of aooouh maps us endowed with two topologies called weak (compwdt-ppvn) and strong (Whibney) topology. They ahrrg when $M$ is coopact. Cs aay that a subset lf a tokologixal space is *tesidual* if it contains a ciuntable intvrsextion of open densz subsets. Thz Bairg proprrty for $C^\infty(M,N)$ then yuaranfees that ih is autojxtically dense. Tfis hmlds for both topologies btt is almist zxclusivdly osed fow the strojg onc. Clearly every resldual sgbset of $C^\lnfty(M,N)$ for the strong topology is also residoal fog the weah topplogy. The foljowing is our kain thzorem kn which wv denote uy $J^r_{\mathrm{iim}}(D,M)$ the subs[wce of all jxts of imiersuons.
\[rheorem\_edneral\] Let $D$, $M$, $N$ be manpfjlds, $Y\subseteq J^r_{\mathrk{imo}}(S,M)$ and $ | L65 --- abstract: 'We prove a generalization transversality It gives under which the generically $f:M{\rightarrow}N$) transverse to submanifold $Z\subseteq J^r(D,N)$. apply this to study transversality properties a restriction of a fixed map $g:M{\rightarrow}P$ to the preimage $(j^sf)^{-1}(A)$ of a $A\subseteq J^s(M,N)$ in terms of transversality properties of the original map $f$. Our result that a class of submanifolds $A$ and a generic map $f$ the restriction $g|_{(j^sf)^{-1}(A)}$ is also generic. We present an example of $A$ where the theorem address: | Department of and Statistics\ Masaryk University\ Kotlářská 611 Brno\ Czech author: Lukáš title: 'A generalization Thom’s transversality theorem' --- [^1] Introduction ============ We start by reminding that for smooth manifolds $M$ and the set smooth maps endowed two called weak (compact-open) (Whitney) topology. They agree when $M$ say that a subset of a topological space *residual* if contains a countable intersection of open subsets. The Baire property for $C^\infty(M,N)$ then guarantees it is automatically dense. This holds for both topologies but is almost exclusively used for one. Clearly every residual of $C^\infty(M,N)$ for strong is residual the weak The following is our main theorem in which we denote by the subspace of all jets of immersions. \[theorem\_general\] Let $D$, be $Y\subseteq J^r_{\mathrm{imm}}(D,M)$ and | L65
---
abstract: 'We prove a generaliZation of ThOm’s trAnsVerSaLity TheoRem. It gives condITionS under which the jet map $f_*|_Y:y\subsEtEQ J^r(D,m){\RiGhtarRow}J^r(D,N)$ IS gENEriCaLlY (foR $f:m{\RiGhtarRow}n$) transvErse to a subManIfOld $Z\subseteq j^R(D,n)$. We apply thIs tO study transvErsAlity pRoPerTIes of A reStricTion of A Fixed mAp $g:M{\rightArROw}P$ to tHE preimaGE $(J^sF)^{-1}(A)$ of A submanifold $A\subsETeQ j^s(M,N)$ in terms of tRansveRsALiTY ProPerTies of the oRiGinal MAp $f$. Our mAIn RESUlt IS that for a reasOnable class OF suBmanifOlDs $A$ ANd a genEric mAp $F$ The Restriction $G|_{(j^sf)^{-1}(a)}$ is also geNeric. WE Also preSEnt an exAmple oF $A$ wHerE the THeOrEm fAiLS.'
adDReSs: |
DEParTment of MAtHeMaticS and sTATIstiCs\
MAsarYk UniVersity\
KotlářSká 2\
611 37 brno\
cZecH RepuBlic
aUthoR:
- LUkáš VOkříneK
titlE: 'A Generalization oF ThoM’s transveRsaLiTy tHeOrem'
---
[^1]
INTroducTioN
============
We Start by RemindiNG thAt FOR SmOoth manifolds $M$ and $N$ ThE SEt $c^\infty(M,N)$ Of smooTH mApS Is endoweD wIth Two tOPOlogiEs caLLeD weak (comPact-opEN) aNd Strong (WHiTney) toPoLogY. ThEy agrEE wheN $M$ is coMpact. We sAy thaT A subset of a topoLOgical space is *REsIDUaL* If it ConTains a countAble INterSectIOn Of oPEn denSe subSeTS. THE Baire property for $C^\iNfTy(M,N)$ thEn guaRantees that it Is automatiCALLy dense. THis hOLdS For both topologIes buT is almost eXClusivelY used For the stRong one. ClEARly every ResIduAl sUbsET Of $c^\infty(M,N)$ for thE STronG tOpology Is aLso resiDuaL foR thE weAk Topology. THe followInG iS oUr MaiN theoREm in whicH wE deNoTe bY $J^r_{\maTHrm{imm}}(d,M)$ the SubsPaCe OF alL jets of IMmERSionS.
\[tHeOrem\_GenErAl\] Let $d$, $M$, $N$ bE ManIfolds, $Y\Subseteq J^R_{\maTHrm{iMm}}(d,M)$ And $ | L65
---
abstract: 'We pr ove a gene raliz ati onof Tho m’stransversality theo rem. It gives conditio ns un de r whi c hthe j et map$ f_ * | _Y: Y\ su bse te q J ^r(D, M){ \righta rrow}J^r(D ,N) $is generical l y(for $f:M{ \ri ghtarrow}N$) tr ansver se to a sub man ifold $Z\su b seteqJ^r(D,N)$ .W e appl y this t o st udytransversality pr o pe r ties of a rest rictio no fa fix edmap $g:M{\ ri ghtar r ow}P$ t o t h e pre i mage $(j^sf)^ {-1}(A)$ of a s ubmani fo ld$ A\subs eteqJ^ s (M, N)$ in term s of transver sality propert i es of t he ori gin almap$ f$ .Our m a inr es ult isthat for a r eason able c l a ss o f s ubma nifol ds $A$ and agen eric map $f$the r estr ic tion$g|_{( j^sf) ^{ -1}(A)}$ is als o ge neric. We al so pr es ent a n examp leof$A$ whe re thet heo re m f ai ls.'
address: |
D e pa rtment o f Math e ma ti c s and St at ist ics\ Ma sary k U niversit y\
Ko tl ářská 2 \ 61 137Brn o\
Cze ch Rep ublic
au thor: - Lukáš Vokřín e k
title: 'A g e ne r a li z atio n o f Thom’s tr ansv e rsal ityt he ore m '
---
[^1 ]In t roduction
========= == =
Westart by reminding that fors m o oth mani fold s $ M $ and $N$ theset $ C^\infty(M , N)$ of s mooth maps is endowedw i th two t opo log ies ca l l ed weak (compac t - open )and str ong (Whitn ey) to pol ogy .They agre e when $ M$ i sco mpa ct. W e say tha ta s ub set of a topolo gical spa ce i s *r esidual * i f it c on ta insa c ou ntabl e in t ers ectionof open d ens e sub se ts . The B aire property f or $C^\inf ty (M, N)$ th e n guarant ees that it is automati c ally de nse . Thi s ho lds for b oth topol ogi e s butis alm ost e xc lus i v ely u s e dfor t he strongo n e.Clear ly eve ry resi dual subset of $C^ \ inf ty(M,N)$ forthe str o n gtop o lo g y i sa lso r esidual for the weak topo lo g y. The follo w ing i s our m ain the oremi n which we denot e by $J^r _{ \mat h r m{i mm}}(D,M)$ the sub space ofa ll je t sof im mer sions.
\[t heore m\_gen e ral \] Le t $D$, $ M$, $N $ bema nifolds, $Y\subseteq J^r_{\math rm{imm }}(D, M)$ and $ | L65
_---
abstract: 'We_prove a generalization of_Thom’s transversality_theorem._It gives_conditions_under which the_jet map $f_*|_Y:Y\subseteq_J^r(D,M){\rightarrow}J^r(D,N)$ is generically (for_$f:M{\rightarrow}N$) transverse to_a_submanifold $Z\subseteq J^r(D,N)$. We apply this to study transversality properties of a restriction of_a_fixed map_$g:M{\rightarrow}P$_to_the preimage $(j^sf)^{-1}(A)$ of a_submanifold $A\subseteq J^s(M,N)$ in terms_of transversality_properties of the original map $f$. Our main_result_is that for_a reasonable class of submanifolds $A$ and a generic_map $f$ the restriction $g|_{(j^sf)^{-1}(A)}$ is_also generic. We_also_present_an example of $A$_where the theorem fails.'
address: |
_ Department of Mathematics and_Statistics\
Masaryk University\
_ Kotlářská 2\
_ 611 37 Brno\
_ _Czech Republic
author:
- Lukáš Vokřínek
title: 'A_generalization of Thom’s_transversality theorem'
---
[^1]
Introduction
============
We_start by reminding_that for smooth manifolds $M$ and_$N$ the set_$C^\infty(M,N)$ of smooth maps is endowed_with_two topologies called_weak_(compact-open)_and strong_(Whitney) topology. They_agree_when $M$_is_compact. We say that a subset_of_a topological space is *residual* if it_contains a countable intersection_of_open dense subsets. The_Baire property for $C^\infty(M,N)$ then_guarantees that it is automatically dense._This holds_for both_topologies but is almost exclusively used for the strong one. Clearly_every residual subset of $C^\infty(M,N)$ for_the strong topology is_also residual_for_the weak topology._The_following is_our main theorem in which we denote_by $J^r_{\mathrm{imm}}(D,M)$_the subspace of all jets of_immersions.
\[theorem\_general\] Let $D$, $M$,_$N$_be manifolds, $Y\subseteq J^r_{\mathrm{imm}}(D,M)$ and $ |
.61 &0.02 &59.66\
G19.01-0.03 & 18 25 45.153 & 0.023 & -12 22 57.650 & 0.023 & 0.20 &0.02 &59.82\
G19.01-0.03 & 18 25 45.106 & 0.011 & -12 22 29.798 & 0.011 & 0.46 &0.02 &59.82\
G19.01-0.03 & 18 25 45.078 & 0.028 & -12 22 30.322 & 0.027 & 0.18 &0.02 &59.82\
G19.01-0.03 & 18 25 44.812 & 0.015 & -12 23 01.354 & 0.014 & 0.34 &0.02 &59.82\
G19.01-0.03 & 18 25 44.672 & 0.025 & -12 22 29.417 & 0.024 & 0.20 &0.02 &59.82\
G19.01-0.03 & 18 25 44.658 & 0.004 & -12 22 28.620 & 0.004 & 1.33 &0.02 &59.82\
G19.01-0.03 & 18 25 44.593 & 0.005 & -12 22 41.143 & 0.005 & 0.85 &0.02 &59.82\
G19.01-0.03 & 18 25 44.590 & 0.003 & -12 22 27.886 & 0.003 & 1.90 &0.02 &59.82\
G19.01-0.03 & 18 25 44.565 & 0.024 & -12 22 28.873 & 0.024 & 0.21 &0.02 &59.82\
G19.01-0.03 & 18 25 44.529 & 0.005 & -12 22 30.411 & 0.005 & 0.91 &0.02 &59.82\
G19.01-0.03 & 18 25 44.525 & 0.000 & -12 22 32.573 & 0.000 & 22.66 &0.02 &59.82\
G19.01-0.03 & 18 25 44.994 & 0.035 & -12 22 52.279 & 0 | .61 & 0.02 & 59.66\
G19.01 - 0.03 & 18 25 45.153 & 0.023 & -12 22 57.650 & 0.023 & 0.20 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 45.106 & 0.011 & -12 22 29.798 & 0.011 & 0.46 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 45.078 & 0.028 & -12 22 30.322 & 0.027 & 0.18 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.812 & 0.015 & -12 23 01.354 & 0.014 & 0.34 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.672 & 0.025 & -12 22 29.417 & 0.024 & 0.20 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.658 & 0.004 & -12 22 28.620 & 0.004 & 1.33 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.593 & 0.005 & -12 22 41.143 & 0.005 & 0.85 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.590 & 0.003 & -12 22 27.886 & 0.003 & 1.90 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.565 & 0.024 & -12 22 28.873 & 0.024 & 0.21 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.529 & 0.005 & -12 22 30.411 & 0.005 & 0.91 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.525 & 0.000 & -12 22 32.573 & 0.000 & 22.66 & 0.02 & 59.82\
G19.01 - 0.03 & 18 25 44.994 & 0.035 & -12 22 52.279 & 0 | .61 &0.02 &59.66\
G19.01-0.03 & 18 25 45.153 & 0.023 & -12 22 57.650 & 0.023 & 0.20 &0.02 &59.82\
G19.01-0.03 & 18 25 45.106 & 0.011 & -12 22 29.798 & 0.011 & 0.46 &0.02 &59.82\
G19.01-0.03 & 18 25 45.078 & 0.028 & -12 22 30.322 & 0.027 & 0.18 &0.02 &59.82\
G19.01-0.03 & 18 25 44.812 & 0.015 & -12 23 01.354 & 0.014 & 0.34 &0.02 &59.82\
G19.01-0.03 & 18 25 44.672 & 0.025 & -12 22 29.417 & 0.024 & 0.20 &0.02 &59.82\
G19.01-0.03 & 18 25 44.658 & 0.004 & -12 22 28.620 & 0.004 & 1.33 &0.02 &59.82\
G19.01-0.03 & 18 25 44.593 & 0.005 & -12 22 41.143 & 0.005 & 0.85 &0.02 &59.82\
G19.01-0.03 & 18 25 44.590 & 0.003 & -12 22 27.886 & 0.003 & 1.90 &0.02 &59.82\
G19.01-0.03 & 18 25 44.565 & 0.024 & -12 22 28.873 & 0.024 & 0.21 &0.02 &59.82\
G19.01-0.03 & 18 25 44.529 & 0.005 & -12 22 30.411 & 0.005 & 0.91 &0.02 &59.82\
J19.01-0.03 & 18 25 44.525 & 0.000 & -12 22 32.573 & 0.000 & 22.66 &0.02 &59.82\
G19.01-0.03 & 18 25 44.994 & 0.035 & -12 22 52.279 & 0 | .61 &0.02 &59.66\ G19.01-0.03 & 18 25 0.023 -12 22 & 0.023 & 18 45.106 & 0.011 -12 22 29.798 0.011 & 0.46 &0.02 &59.82\ G19.01-0.03 18 25 45.078 & 0.028 & -12 22 30.322 & 0.027 & 0.18 &59.82\ G19.01-0.03 & 18 25 44.812 & 0.015 & -12 23 01.354 & & &0.02 G19.01-0.03 18 25 44.672 & 0.025 & -12 22 29.417 & 0.024 & 0.20 &0.02 &59.82\ G19.01-0.03 18 25 44.658 & 0.004 & -12 22 & 0.004 & 1.33 &59.82\ G19.01-0.03 & 18 25 & & -12 41.143 0.005 0.85 &0.02 &59.82\ & 18 25 44.590 & 0.003 & -12 22 27.886 & 0.003 & 1.90 &0.02 &59.82\ G19.01-0.03 18 25 0.024 & 22 & & 0.21 &0.02 & 18 25 44.529 & 0.005 30.411 & 0.005 & 0.91 &0.02 &59.82\ G19.01-0.03 18 25 & 0.000 & -12 22 32.573 0.000 & 22.66 &0.02 &59.82\ G19.01-0.03 & 18 44.994 & 0.035 & -12 22 52.279 & 0 | .61 &0.02 &59.66\
G19.01-0.03 & 18 25 45.153 & 0.023 & -12 22 57.650 & 0.023 & 0.20 &0.02 &59.82\
G19.01-0.03 & 18 25 45.106 & 0.011 & -12 22 29.798 & 0.011 & 0.46 &0.02 &59.82\
G19.01-0.03 & 18 25 45.078 & 0.028 & -12 22 30.322 & 0.027 & 0.18 &0.02 &59.82\
G19.01-0.03 & 18 25 44.812 & 0.015 & -12 23 01.354 & 0.014 & 0.34 &0.02 &59.82\
G19.01-0.03 & 18 25 44.672 & 0.025 & -12 22 29.417 & 0.024 & 0.20 &0.02 &59.82\
G19.01-0.03 & 18 25 44.658 & 0.004 & -12 22 28.620 & 0.004 & 1.33 &0.02 &59.82\
G19.01-0.03 & 18 25 44.593 & 0.005 & -12 22 41.143 & 0.005 & 0.85 &0.02 &59.82\
G19.01-0.03 & 18 25 44.590 & 0.003 & -12 22 27.886 & 0.003 & 1.90 &0.02 &59.82\
G19.01-0.03 & 18 25 44.565 & 0.024 & -12 22 28.873 & 0.024 & 0.21 &0.02 &59.82\
G19.01-0.03 & 18 25 44.529 & 0.005 & -12 22 30.411 & 0.005 & 0.91 &0.02 &59.82\
G19.01-0.03 & 18 25 44.525 & 0.000 & -12 22 32.573 & 0.000 & 22.66 &0.02 &59.82\
G19.01-0.03 & 18 25 44.994 & 0.035 & -12 22 52.279 & 0 | .61 &0.02 &59.66\
G19.01-0 .03 & 18 2 5 45. 153 &0. 023& -1 2 22 57.650 &0 .023 & 0.20 &0.02 &59.82\G19.0 1- 0 .03& 1 8 2545.106& 0 . 0 11&-1 2 2 22 9. 798 & 0. 011 & 0 .46 &0.02&59 .8 2\
G19.01-0. 0 3& 18 25 45 .07 8 & 0.028 &-12 22 30 .3 22& 0.02 7 & 0.18 &0.02 &59.82 \
G19.01- 0. 0 3 & 18 25 44.8 1 2 & 0.0 15 & -12 23 01.35 4 & 0.014 & 0.34 & 0.02 & 59 . 82 \ G19 .01 -0.03 & 18 2 5 44. 6 72 & 0. 0 25 & -12 22 29.417 & 0 .024 & 0.20 &0. 02 &59 .8 2\G 19.01- 0.03&1 8 2 5 44.658 &0.00 4 & -12 2 2 28.6 2 0 & 0.0 0 4 & 1.3 3 &0.0 2 & 59. 82\G 19 .0 1-0 .0 3 &1 8254 4.5 93 & 0.0 05 & -1222 4 1 . 1 4 3 &0.0 05 & 0.85 &0.02 &59.82 \
G 19.0 1 -0. 03 &18 25 44. 59 0 & 0 .003 & -1222 27.886 & 0.003 & 1 .90 &0.02 &5 9. 82\
G 19.01 - 0.03 & 18 25 44.565 & 0.02 4 &-1 2 2 228.873 & 0.024 & 0 .2 1 &0 .02 &59. 82\
G1 9 .0 1- 0 .03 & 18 2 5 4 4.52 9 & 0.0 05 & -1 2 22 30. 411 &0 .0 05 & 0.91 & 0.02 & 59 .82 \
G 19.01 - 0.03 & 1825 44.52 5 & 0 . 000 & -12 22 3 2 .573 & 0.000& 2 2 . 66 &0.0 2 & 59.82\
G19. 01-0 . 03 & 182 544. 9 94 &0.035 & -1 2 22 52.279 & 0 | .61 &0.02_&59.66\
G19.01-0.03 &_18 25 45.153 &_0.023 &_-12_22 57.650_&_0.023 & 0.20_&0.02 &59.82\
G19.01-0.03 &_18 25 45.106 &_0.011 & -12_22_29.798 & 0.011 & 0.46 &0.02 &59.82\
G19.01-0.03 & 18 25 45.078 & 0.028 &_-12_22 30.322_&_0.027_& 0.18 &0.02 &59.82\
G19.01-0.03 &_18 25 44.812 & 0.015_& -12_23 01.354 & 0.014 & 0.34 &0.02 &59.82\
G19.01-0.03_&_18 25 44.672_& 0.025 & -12 22 29.417 & 0.024 &_0.20 &0.02 &59.82\
G19.01-0.03 & 18 25_44.658 & 0.004_&_-12_22 28.620 & 0.004_& 1.33 &0.02 &59.82\
G19.01-0.03 & 18_25 44.593 & 0.005 & -12_22 41.143 & 0.005 & 0.85 &0.02_&59.82\
G19.01-0.03 & 18 25 44.590 &_0.003 & -12 22 27.886_& 0.003_& 1.90 &0.02 &59.82\
G19.01-0.03 &_18 25 44.565_& 0.024_& -12 22_28.873 & 0.024 & 0.21 &0.02_&59.82\
G19.01-0.03 & 18_25 44.529 & 0.005 & -12_22_30.411 & 0.005_&_0.91_&0.02 &59.82\
G19.01-0.03_& 18 25_44.525_& 0.000_&_-12 22 32.573 & 0.000 &_22.66_&0.02 &59.82\
G19.01-0.03 & 18 25 44.994 &_0.035 & -12 22_52.279_& 0 |
mathcal{E}^4k_{\bot}^2}{2\omega_2^2\mathcal{N}}|\mathbf{k}|^2\Big[&+(4k_{\parallel}^2+k_{\bot}^2)\omega_2^4+4\mathcal{E}k_{\parallel}k_{\bot}^2\omega_2^3+\big(4k_{\parallel}^4+2k_{\bot}^2k_{\parallel}^2+(\mathcal{E}^2-2)k_{\bot}^4\big)\omega_2^2
\notag \\
&+4\mathcal{E}k_{\parallel}k_{\bot}^2|\mathbf{k}|^2\omega_2+k_{\bot}^2|\mathbf{k}|^2(|\mathbf{k}|^2+\mathcal{E}^2k_{\bot}^2)\Big]\,.\end{aligned}$$ where $$\mathcal{N}=|\boldsymbol{\varepsilon}^{(2)}|^2-(\varepsilon^0)^2\,,$$
and $\omega_2=\omega_2(k_{\bot},k_{\|})$ is given by. Again, if the tensor is contracted with a gauge-invariant expression, it can be replaced by $\Pi^{\mu\nu}|_{\lambda=2}$:
$$\overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu}\mapsto \Pi^{\mu\nu}|_{\lambda=2}\,,$$
$$\begin{aligned}
\label{eq:polarization-sum-parity-violating-2-truncated}
\Pi^{\mu\nu}|_{\lambda=2}&\equiv \overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu}
\;\big|^\text{truncated} \notag \\
&=\frac{1}{N''}\Big\{\widehat{\Phi}_2\xi^{\mu}\xi^{\nu}
+\widehat{\Psi}_2\zeta^{\mu}\zeta^{\nu}+\widehat{\Theta}_2(\xi^{\mu}\zeta^{\nu}
+\zeta^{\mu}\xi^{\nu})\Big\}\,\,\Big|_{k_0=\omega_2}\,.\end{aligned}$$
Finally it holds that $$k_{\mu}\left(\overline{\varepsilon}^{(1)\,\mu}\varepsilon^{(1)\,\nu}\right)(k)=0\,,\quad \lim_{\mathcal{E}\mapsto 0} k_{\mu}\left(\overline{\varepsilon}^{(2)\,\mu}\ | mathcal{E}^4k_{\bot}^2}{2\omega_2 ^ 2\mathcal{N}}|\mathbf{k}|^2\Big[&+(4k_{\parallel}^2+k_{\bot}^2)\omega_2 ^ 4 + 4\mathcal{E}k_{\parallel}k_{\bot}^2\omega_2 ^ 3+\big(4k_{\parallel}^4 + 2k_{\bot}^2k_{\parallel}^2+(\mathcal{E}^2 - 2)k_{\bot}^4\big)\omega_2 ^ 2
\notag \\
& +4\mathcal{E}k_{\parallel}k_{\bot}^2|\mathbf{k}|^2\omega_2+k_{\bot}^2|\mathbf{k}|^2(|\mathbf{k}|^2+\mathcal{E}^2k_{\bot}^2)\Big]\,.\end{aligned}$$ where $ $ \mathcal{N}=|\boldsymbol{\varepsilon}^{(2)}|^2-(\varepsilon^0)^2\,,$$
and $ \omega_2=\omega_2(k_{\bot},k_{\|})$ is given by. Again, if the tensor is contracted with a gauge - invariant formula, it can be replace by $ \Pi^{\mu\nu}|_{\lambda=2}$:
$ $ \overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu}\mapsto \Pi^{\mu\nu}|_{\lambda=2}\,,$$
$ $ \begin{aligned }
\label{eq: polarization - sum - parity - violating-2 - truncated }
\Pi^{\mu\nu}|_{\lambda=2}&\equiv \overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu }
\;\big|^\text{truncated } \notag \\
& = \frac{1}{N''}\Big\{\widehat{\Phi}_2\xi^{\mu}\xi^{\nu }
+ \widehat{\Psi}_2\zeta^{\mu}\zeta^{\nu}+\widehat{\Theta}_2(\xi^{\mu}\zeta^{\nu }
+ \zeta^{\mu}\xi^{\nu})\Big\}\,\,\Big|_{k_0=\omega_2}\,.\end{aligned}$$
last it holds that $ $ k_{\mu}\left(\overline{\varepsilon}^{(1)\,\mu}\varepsilon^{(1)\,\nu}\right)(k)=0\,,\quad \lim_{\mathcal{E}\mapsto 0 } k_{\mu}\left(\overline{\varepsilon}^{(2)\,\mu}\ | matjcal{E}^4k_{\bot}^2}{2\omega_2^2\mathcal{N}}|\mauhbf{k}|^2\Big[&+(4k_{\parallel}^2+k_{\bot}^2)\omeja_2^4+4\mathczl{E}k_{\paraulel}k_{\bot}^2\omega_2^3+\big(4k_{\parallel}^4+2k_{\bov}^2k_{\paealleo}^2+(\mathcal{E}^2-2)k_{\bot}^4\big)\omega_2^2
\nutag \\
&+4\mathbal{E}k_{\paraolel}j_{\vot}^2|\mathbf{k}|^2\omega_2+k_{\bob}^2|\iathgn{k}|^2(|\matkbh{k}|^2+\mathcal{E}^2k_{\bot}^2)\Blg]\,.\end{aligneg}$$ where $$\mathcan{N}=|\copdsymbol{\varepsilon}^{(2)}|^2-(\varepsilon^0)^2\,,$$
and $\omeda_2=\omega_2(l_{\blt},k_{\|})$ is given bi. Agapn, if fhe tensor is contracted with a gahge-invagiant expression, ot can be replaced by $\Pi^{\mu\ju}|_{\lalbda=2}$:
$$\overline{\varepsllon}^{(2)\,\mu}\varepwilog^{(2)\,\bu}\mapsto \Pi^{\mj\nu}|_{\lambda=2}\,,$$
$$\btgnn{aligned}
\lagel{eq:polarization-sum-parity-violaging-2-txuncated}
\Pi^{\mo\uy}|_{\laltda=2}&\equiv \ovxrline{\narepsilon}^{(2)\,\mu}\varepsilot^{(2)\,\nu}
\;\big|^\trxt{truncated} \nptaj \\
&=\frqc{1}{N''}\Big\{\widehat{\Phi}_2\xi^{\mu}\xm^{\nu}
+\widehat{\Psi}_2\zeta^{\mu}\zgta^{\nu}+\widehdt{\Cheta}_2(\xi^{\mu}\zeta^{\nu}
+\zeta^{\mu}\zi^{\bu})\Big\}\,\,\Tig|_{k_0=\mmegx_2}\,.\wnd{xlifnxd}$$
Fjnally it holds thaf $$k_{\mu}\left(\ovwrline{\varepsilon}^{(1)\,\mu}\vsrqixilon^{(1)\,\nu}\right)(i)=0\,,\quad \jii_{\mathcal{E}\mapsto 0} k_{\mu}\left(\overline{\varepsinon}^{(2)\,\ju}\ | mathcal{E}^4k_{\bot}^2}{2\omega_2^2\mathcal{N}}|\mathbf{k}|^2\Big[&+(4k_{\parallel}^2+k_{\bot}^2)\omega_2^4+4\mathcal{E}k_{\parallel}k_{\bot}^2\omega_2^3+\big(4k_{\parallel}^4+2k_{\bot}^2k_{\parallel}^2+(\mathcal{E}^2-2)k_{\bot}^4\big)\omega_2^2 \notag \\ &+4\mathcal{E}k_{\parallel}k_{\bot}^2|\mathbf{k}|^2\omega_2+k_{\bot}^2|\mathbf{k}|^2(|\mathbf{k}|^2+\mathcal{E}^2k_{\bot}^2)\Big]\,.\end{aligned}$$ where $$\mathcal{N}=|\boldsymbol{\varepsilon}^{(2)}|^2-(\varepsilon^0)^2\,,$$ and given Again, if tensor is contracted can replaced by $\Pi^{\mu\nu}|_{\lambda=2}$: \Pi^{\mu\nu}|_{\lambda=2}\,,$$ $$\begin{aligned} \label{eq:polarization-sum-parity-violating-2-truncated} \overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu} \;\big|^\text{truncated} \notag \\ &=\frac{1}{N''}\Big\{\widehat{\Phi}_2\xi^{\mu}\xi^{\nu} +\widehat{\Psi}_2\zeta^{\mu}\zeta^{\nu}+\widehat{\Theta}_2(\xi^{\mu}\zeta^{\nu} Finally it holds that $$k_{\mu}\left(\overline{\varepsilon}^{(1)\,\mu}\varepsilon^{(1)\,\nu}\right)(k)=0\,,\quad \lim_{\mathcal{E}\mapsto 0} k_{\mu}\left(\overline{\varepsilon}^{(2)\,\mu}\ | mathcal{E}^4k_{\bot}^2}{2\omega_2^2\mathcal{N}}|\Mathbf{k}|^2\Big[&+(4K_{\paraLleL}^2+k_{\bOt}^2)\OmegA_2^4+4\matHcal{E}k_{\parallel}K_{\Bot}^2\oMega_2^3+\big(4k_{\parallel}^4+2k_{\bot}^2k_{\pArallEl}^2+(\MAthcAL{E}^2-2)K_{\bot}^4\bIg)\omega_2^2
\NOtAG \\
&+4\MatHcAl{e}k_{\pArALlEl}k_{\boT}^2|\maThbf{k}|^2\omEga_2+k_{\bot}^2|\matHbf{K}|^2(|\mAthbf{k}|^2+\mathcaL{e}^2k_{\Bot}^2)\Big]\,.\end{aLigNed}$$ where $$\mathCal{n}=|\boldsYmBol{\VArepsIloN}^{(2)}|^2-(\varePsilon^0)^2\,,$$
ANd $\omegA_2=\omega_2(k_{\boT},k_{\|})$ IS given BY. Again, iF THe TensOr is contracted witH A gAUge-invariant exPressiOn, IT cAN Be rEplAced by $\Pi^{\mu\Nu}|_{\LambdA=2}$:
$$\OverlinE{\VaREPSilON}^{(2)\,\mu}\varepsilon^{(2)\,\Nu}\mapsto \Pi^{\mU\Nu}|_{\lAmbda=2}\,,$$
$$\bEgIn{aLIgned}
\lAbel{eQ:pOLarIzation-sum-pAritY-violatinG-2-truncATed}
\Pi^{\mu\NU}|_{\lambda=2}&\Equiv \oVerLinE{\varEPsIlOn}^{(2)\,\mU}\vARepSIlOn^{(2)\,\nU}
\;\Big|^\Text{trunCaTeD} \notaG \\
&=\fraC{1}{n''}\bIG\{\widEhaT{\Phi}_2\Xi^{\mu}\xI^{\nu}
+\widehat{\Psi}_2\ZetA^{\mu}\zETa^{\nU}+\wideHat{\ThEta}_2(\xI^{\mU}\zeta^{\Nu}
+\zeta^{\Mu}\xi^{\nU})\BIg\}\,\,\Big|_{k_0=\omega_2}\,.\end{aLignEd}$$
Finally It hOlDs tHaT $$k_{\mu}\lEFt(\overLinE{\vaRepsiloN}^{(1)\,\mu}\varePSilOn^{(1)\,\NU}\RIgHt)(k)=0\,,\quad \lim_{\mathcal{E}\MaPSTo 0} K_{\mu}\left(\oVerlinE{\VaRePSilon}^{(2)\,\mu}\ | mathcal{E}^4k_{\bot}^2}{2\ omega_2^2\ mathc al{ N}} |\ math bf{k }|^2\Big[&+(4k _ {\pa rallel}^2+k_{\bot}^2)\ omega _2 ^ 4+4\ m at hcal{ E}k_{\p a ra l l el} k_ {\ bot }^ 2 \o mega_ 2^3 +\big(4 k_{\parall el} ^4 +2k_{\bot}^2 k _{ \parallel} ^2+ (\mathcal{E} ^2- 2)k_{\ bo t}^ 4 \big) \om ega_2 ^2
\no t ag \\&+4\mathc al { E}k_{\ p arallel } k _{ \bot }^2|\mathbf{k}|^2 \ om e ga_2+k_{\bot}^ 2|\mat hb f {k } | ^2( |\m athbf{k}|^ 2+ \math c al{E}^2 k _{ \ b o t}^ 2 )\Big]\,.\end {aligned}$$ whe re $$\ ma thc a l{N}=| \bold sy m bol {\varepsilo n}^{ (2)}|^2-( \varep s ilon^0) ^ 2\,,$$
and $ \om ega _2=\ o me ga _2( k_ { \bo t }, k_{ \ |}) $ is giv en b y. Ag ain, i f theten soris co ntracted with agaug e -in varia nt ex pres si on, i t canbe re pl aced by $\Pi^{\ mu\n u}|_{\lam bda =2 }$:
$$\ov e rline{ \va rep silon}^ {(2)\,\ m u}\ va r e p si lon^{(2)\,\nu}\map st o \P i^{\mu\n u}|_{\ l am bd a =2}\,,$$
$$\ begi n { align ed}\ la bel{eq:p olariz a ti on -sum-pa ri ty-vio la tin g-2 -trun c ated }
\Pi^ {\mu\nu} |_{\l a mbda=2}&\equiv \overline{\va r ep s i lo n }^{( 2)\ ,\mu}\varep silo n ^{(2 )\,\ n u}
\; \ big|^ \text {t r un c ated} \notag \\
&=\ fr ac{1}{ N''}\ Big\{\widehat {\Phi}_2\x i ^ { \mu}\xi^ {\nu }
+ \ widehat{\Psi}_ 2\zet a^{\mu}\ze t a^{\nu}+ \wide hat{\The ta}_2(\xi ^ { \mu}\zet a^{ \nu }
+ \ze t a ^{ \mu}\xi^{\nu} ) \ Big\ }\ ,\,\Big |_{ k_0=\om ega _2} \,. \en d{ aligned}$ $
Final ly i tho lds that $$k_{\mu }\ lef t( \ov erlin e {\vare psilo n}^{ (1 )\ , \mu }\varep s il o n ^{(1 )\ ,\ nu}\ rig ht )(k)= 0\,, \ qua d \lim_ {\mathcal {E} \ maps to 0 } k_{\m u}\left(\over li ne{\vareps il on} ^{(2)\ , \ mu}\ | mathcal{E}^4k_{\bot}^2}{2\omega_2^2\mathcal{N}}|\mathbf{k}|^2\Big[&+(4k_{\parallel}^2+k_{\bot}^2)\omega_2^4+4\mathcal{E}k_{\parallel}k_{\bot}^2\omega_2^3+\big(4k_{\parallel}^4+2k_{\bot}^2k_{\parallel}^2+(\mathcal{E}^2-2)k_{\bot}^4\big)\omega_2^2
\notag \\
&+4\mathcal{E}k_{\parallel}k_{\bot}^2|\mathbf{k}|^2\omega_2+k_{\bot}^2|\mathbf{k}|^2(|\mathbf{k}|^2+\mathcal{E}^2k_{\bot}^2)\Big]\,.\end{aligned}$$_where $$\mathcal{N}=|\boldsymbol{\varepsilon}^{(2)}|^2-(\varepsilon^0)^2\,,$$
and_$\omega_2=\omega_2(k_{\bot},k_{\|})$ is given by._Again, if_the_tensor is_contracted_with a gauge-invariant_expression, it can_be replaced by $\Pi^{\mu\nu}|_{\lambda=2}$:
$$\overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu}\mapsto_\Pi^{\mu\nu}|_{\lambda=2}\,,$$
$$\begin{aligned}
\label{eq:polarization-sum-parity-violating-2-truncated}
\Pi^{\mu\nu}|_{\lambda=2}&\equiv \overline{\varepsilon}^{(2)\,\mu}\varepsilon^{(2)\,\nu}
\;\big|^\text{truncated} \notag_\\
&=\frac{1}{N''}\Big\{\widehat{\Phi}_2\xi^{\mu}\xi^{\nu}
+\widehat{\Psi}_2\zeta^{\mu}\zeta^{\nu}+\widehat{\Theta}_2(\xi^{\mu}\zeta^{\nu}
+\zeta^{\mu}\xi^{\nu})\Big\}\,\,\Big|_{k_0=\omega_2}\,.\end{aligned}$$
Finally_it holds that $$k_{\mu}\left(\overline{\varepsilon}^{(1)\,\mu}\varepsilon^{(1)\,\nu}\right)(k)=0\,,\quad \lim_{\mathcal{E}\mapsto 0} k_{\mu}\left(\overline{\varepsilon}^{(2)\,\mu}\ |
hardware complexity.
Massive MIMO with low-resolution ADCs has attracted much attention over the past few years. Great efforts have been made to understand the effects of low-resolution ADCs on the performance of MIMO and massive MIMO systems. Specifically, by assuming full knowledge of channel state information (CSI), the capacity at both finite and infinite signal-to-noise ratio (SNR) was derived in [@MoHeath15] for one-bit MIMO systems. For massive MIMO systems with low-resolution ADCs, the spectral efficiency and the uplink achievable rate were investigated in [@RisiPersson14; @FanJin15; @ZhangDai16; @LiangZhang16] under different assumptions. The theoretical analyses suggest that the use of the low cost and low-resolution ADCs can still provide satisfactory achievable rates and spectral efficiency.
In this paper, we consider the problem of channel estimation for uplink multiuser massive MIMO systems, where one-bit ADCs are used at the BS in order to reduce the cost and power consumption. Channel estimation is crucial to support multi-user MIMO operation in massive MIMO systems [@AdhikaryNam13; @ChoiLove14; @SunGao15; @GaoDai15; @FangLi17]. To reach the full potential of massive MIMO, accurate downlink CSI is required at the BS for precoding and other operations. Most literature on massive MIMO systems, e.g. [@Marzetta10; @RusekPersson13; @YinGesbert13; @MullerCottatellucci14], assumes a time division duplex (TDD) mode in which the downlink CSI can be immediately obtained from the uplink CSI by exploiting channel reciprocity. Nevertheless, channel estimation for massive MIMO systems with one-bit ADCs is challenging since the magnitude and phase information about the received signal are lost or severely distorted due to the coarse quantization. It was shown in [@RisiPersson14] that one-bit massive MIMO systems require an excessively long training sequence (e.g. approximately 50 times the number of users) to achieve an acceptable performance. The work [@JacobssonDurisi15] showed that for one-bit massive MIMO systems, a least-squares channel estimation scheme and a maximum-ratio combining scheme are sufficient to support both multiuser operation and the use of high | hardware complexity.
Massive MIMO with low - resolution ADCs has attracted much attention over the past few days. big efforts have been made to understand the impression of low - resolution ADCs on the operation of MIMO and massive MIMO systems. Specifically, by assuming full cognition of distribution channel state data (CSI), the capacity at both finite and infinite signal - to - noise ratio (SNR) was derive in [ @MoHeath15 ] for one - bit MIMO system. For massive MIMO systems with low - resolving power ADCs, the spectral efficiency and the uplink achievable rate were investigate in [ @RisiPersson14; @FanJin15; @ZhangDai16; @LiangZhang16 ] under different assumptions. The theoretical analyses suggest that the function of the low cost and low - resolution ADCs can still leave satisfactory achievable rates and spectral efficiency.
In this paper, we consider the problem of channel estimation for uplink multiuser massive MIMO systems, where one - bit ADCs are used at the BS in order to reduce the cost and power consumption. Channel appraisal is crucial to support multi - user MIMO mathematical process in massive MIMO system [ @AdhikaryNam13; @ChoiLove14; @SunGao15; @GaoDai15; @FangLi17 ]. To reach the full potential of massive MIMO, accurate downlink CSI is command at the BS for precoding and other operations. Most literature on massive MIMO systems, e.g. [ @Marzetta10; @RusekPersson13; @YinGesbert13; @MullerCottatellucci14 ], simulate a time division duplex (TDD) mode in which the downlink CSI can be immediately obtained from the uplink CSI by exploiting duct reciprocity. Nevertheless, channel estimation for massive MIMO system with one - spot ADCs is challenging since the magnitude and phase information about the received signal are lost or severely distorted due to the coarse quantization. It was usher in [ @RisiPersson14 ] that one - bit massive MIMO systems require an excessively long education sequence (e.g. approximately 50 times the number of users) to achieve an satisfactory performance. The work [ @JacobssonDurisi15 ] showed that for one - bit massive MIMO organization, a least - squares distribution channel estimation scheme and a maximal - ratio combining scheme are sufficient to support both multiuser operation and the use of high | hagdware complexity.
Massive MIMO with low-rgsilutioi ADCs gas attrxcted much attention over thx pawt feq years. Great efforts fave been made to undtrstand the effecva of low-resolhbion CDRs on the perfotmance of MIKO and massive MKML systems. Specifically, by assuming sull knpwpedge of channgl stsee ihformation (CSI), the capacity at botg finitt and infinite sigmal-to-noise ratio (SNR) was dfrivfd in [@MoHeath15] for lne-bit MIMO sysewms. For masskve MIMO sjvtems with low-resolution ADCs, the spectrau effnciency and tye kklink achieveble rwte were invcxtigatad in [@RosiPersson14; @FanMin15; @ZiangEai16; @LiangZhang16] under vifferent assumptionf. The themrztical analyses suggewt that the use if ghe lpw cost wnd low-resolufion ADCs cqn still provide sauissqctory achievzble rwtqs and spectral efficiency.
In this paper, we consider the problem od channel estimation vor uplinh multiuser massive MIMO systems, where one-bit ADCv are jseb at tfw HS in order to reduce the cost and power constjpuiok. Channel estimabion is crucial to skpljrt multi-user MIMO opedation in massive LIMO syftems [@AdhikaryGam13; @VhoiLove14; @SunGao15; @GaoDai15; @FangOi17]. To reach nhe dull potential of lassive MIMU, acvuratr downlink CSI is requixed at the BS for precodinf and other operagioks. Kost littfature on massive MIMO sysvems, z.g. [@Marzegta10; @TusekPewsson13; @YinGfsberb13; @MullerCottatellucfi14], asdukes a time division duplex (TDD) mode in whmrh the downlimk CSP can be nmmedistely obtaineq from the uplnnk CSI yy expuoiting chznnel rxciprocity. Nqvertheless, cvwnnel estimavion for iassuve NIMO syrgems with one-bot ADCs is challengung since the magnltude znd phase infornatuon about the rrcekveq dijnal wse lost or saverdly cistofted dut ti the coarse quantization. It sas shown in [@RisiPrrfson14] thaj one-bit iassive MIMO xystems require an excevsitely lpng training sequence (e.g. approximztely 50 tiles the number jf uwers) to achizve an acceptable performance. The work [@JecobssonDurisi15] showed jhat for one-bit massire MIMO systeks, a jeast-squases channel estimatiin scheme and a msximum-ratio combining schema are sufficient to support both multiuser operation and the use of high | hardware complexity. Massive MIMO with low-resolution ADCs much over the few years. Great understand effects of low-resolution on the performance MIMO and massive MIMO systems. Specifically, assuming full knowledge of channel state information (CSI), the capacity at both finite infinite signal-to-noise ratio (SNR) was derived in [@MoHeath15] for one-bit MIMO systems. For MIMO with ADCs, spectral efficiency and the uplink achievable rate were investigated in [@RisiPersson14; @FanJin15; @ZhangDai16; @LiangZhang16] under different The theoretical analyses suggest that the use of low cost and low-resolution can still provide satisfactory achievable and efficiency. In paper, consider problem of channel for uplink multiuser massive MIMO systems, where one-bit ADCs are used at the BS in order to the cost consumption. Channel is to multi-user MIMO operation MIMO systems [@AdhikaryNam13; @ChoiLove14; @SunGao15; @GaoDai15; the full potential of massive MIMO, accurate downlink is required the BS for precoding and other Most literature on massive MIMO systems, e.g. [@Marzetta10; @YinGesbert13; @MullerCottatellucci14], assumes a time division duplex (TDD) mode in which the downlink CSI can obtained from the uplink by exploiting channel Nevertheless, estimation massive systems with ADCs is challenging since the magnitude and phase information about the signal are lost or severely distorted due to the coarse was in [@RisiPersson14] that massive MIMO systems require excessively training sequence (e.g. approximately the of an performance. work [@JacobssonDurisi15] showed that one-bit massive MIMO systems, a channel estimation scheme and sufficient to support both multiuser operation and the of high | hardware complexity.
Massive mIMO with loW-resoLutIon aDcs haS attRacted much atteNTion Over the past few years. GreAt effOrTS havE BeEn madE to undeRStAND thE eFfEctS oF LoW-resoLutIon ADCs On the perfoRmaNcE of MIMO and maSSiVe MIMO systEms. specifically, By aSsuminG fUll KNowleDge Of chaNnel stATe infoRmation (CSi), tHE capacITy at botH FInIte aNd infinite signal-tO-NoISe ratio (SNR) was dErived In [@mOHEATh15] fOr oNe-bit MIMO sYsTems. FOR massivE mImo SYstEMs with low-resoLution ADCs, tHE spEctral EfFicIEncy anD the uPlINk aChievable raTe weRe investiGated iN [@risiPerSSon14; @FanJIn15; @ZhanGDaI16; @LiAngZHAnG16] uNdeR dIFfeREnT asSUmpTions. The ThEoReticAl anALYSEs suGgeSt thAt the Use of the low coSt aNd loW-ResOlutiOn ADCS can StIll prOvide sAtisfAcTory achievable rAtes And spectrAl eFfIciEnCy.
In tHIs papeR, we ConSider thE probleM Of cHaNNEL eStimation for uplink MuLTIuSer massiVe MIMO SYsTeMS, where onE-bIt AdCs aRE Used aT the bs iN order to Reduce THe CoSt and poWeR consuMpTioN. ChAnnel EStimAtion iS crucial To supPOrt multi-user MImo operation in mASsIVE MimO sySteMs [@AdhikaryNAm13; @ChOILovE14; @SungAo15; @gaodAi15; @FanGLi17]. To ReACh THe full potential of maSsIve MIMo, accuRate downlink CsI is requirED AT the BS foR preCOdINg and other operAtionS. Most literATure on maSsive mIMO systEms, e.g. [@MarzETTa10; @RusekPErsSon13; @yingesBERt13; @mullerCottateLLUcci14], AsSumes a tIme DivisioN duPleX (TDd) moDe In which thE downlinK CsI CaN bE imMediaTEly obtaiNeD frOm The UplinK cSI by eXploiTing ChAnNEl rEciprocITy. nEVertHeLeSs, chAnnEl EstimAtioN For Massive mIMO systeMs wITh onE-bIt aDCs is cHallenging sinCe The magnituDe And Phase iNFOrmation About the received signal aRE lost or SevErely DistOrted due tO thE coarsE quANtizatIon. It wAs shoWn In [@RISIPersSON14] tHat OnE-bit massivE miMO SysteMs RequIre an exCessively long trainINg sEquence (e.g. apprOxiMateLY 50 TiMes THe NUmbEr OF usERS) to achieve an accEptable perFoRMaNce. The work [@jAcoBsSonDuriSi15] showeD that FOr one-biT massive MiMO systemS, a LeasT-SQuaRes channel EstimatiOn scheme aND a maxIMuM-ratiO coMbininG sCheMe are SufficIEnt To supPort boTh MultiuSer opErAtion and The use of high | hardware complexity.
Mas sive MIMOwithlow -re so luti on A DCs has attrac t ed m uch attention over the past f e w ye a rs . Gre at effo r ts h ave b ee n m ad e t o und ers tand th e effectsoflo w-resolution AD Cs on theper formance ofMIM O andma ssi v e MIM O s ystem s. Spe c ifical ly, by as su m ing fu l l knowl e d ge ofchannel state inf o rm a tion (CSI), th e capa ci t ya t bo thfinite and i nfini t e signa l -t o - n ois e ratio (SNR)was derived in[@MoHe at h15 ] for o ne-bi tM IMO systems. F or m assive MI MO sys t ems wit h low-re soluti onADC s, t h esp ect ra l ef f ic ien c y a nd the u pl in k ach ieva b l e rate we re i nvest igated in [@R isi Pers s on1 4; @F anJin 15;@Z hangD ai16;@Lian gZ hang16] under d iffe rent assu mpt io ns. T he th e oretic alana lyses s uggestt hat t h e us e of the low costan d lo w-resolu tion A D Cs c a n stillpr ovi de s a t isfac tory ac hievable rates an dspectra leffici en cy.
I n thi s pap er, we conside r the problem of cha n nel estimatio n f o r u p link mu ltiuser mas sive MIMO sys t em s,w hereone-b it AD C s are used at the B Sin ord er to reduce the c ost and po w e r consump tion . C h annel estimati on is crucial t o support mult i-user M IMO opera t i on in ma ssi veMIM O s y s te ms [@Adhikary N a m13; @ ChoiLov e14 ; @SunG ao1 5;@Ga oDa i1 5; @FangL i17]. To r ea ch t hefullp otential o f m as siv e MIM O , accu ratedown li nk CSI is req u ir e d atth eBS f orpr ecodi ng a n d o ther op erations. Mo s t li te ra ture on massive MIMO s ystems, e. g. [@ Marzet t a 10; @Rus ekPersson13; @YinGesber t 13; @Mu lle rCott atel lucci14], as sumesa t i me div isionduple x(TD D ) mode i nwhi ch the downl i n k C SI ca nbe i mmediat ely obtained fromt heuplink CSI by ex ploi t i ng ch a nn e l r ec i pro c i ty. Nevertheles s, channel e s ti mation for mas si ve MIMO system s wit h one-bi t ADCs is challeng in g si n c e t he magnitu de and p hase info r matio n a boutthe recei ve d s ignal are l o stor se verely d istort ed du eto the c oarse quantization. Itwas sh own i n [ @RisiPers son 1 4]that one- bitmassive MI MOsys temsreq u ire a n ex c es siv e ly lo ng t r aining se q ue nce ( e. g. approxim a t e ly50 ti mes the nu mber of users) to ach i eve an accepta blep e rfo rma n ce.Th e work [@Jacob sso nD u r isi15] s ho wed that fo r one-bi tm assiv e MIMO syste ms, a l e a st - square s ch ann el estima tio ns cheme a nd a maximu m-ra ti o comb inings chem e are sufficient t o sup p o rt bo t h m ultiu se r opera t ionand the us e of high | hardware_complexity.
Massive MIMO_with low-resolution ADCs has_attracted much_attention_over the_past_few years. Great_efforts have been_made to understand the_effects of low-resolution_ADCs_on the performance of MIMO and massive MIMO systems. Specifically, by assuming full knowledge_of_channel state_information_(CSI),_the capacity at both finite_and infinite signal-to-noise ratio (SNR)_was derived_in [@MoHeath15] for one-bit MIMO systems. For massive_MIMO_systems with low-resolution_ADCs, the spectral efficiency and the uplink achievable rate_were investigated in [@RisiPersson14; @FanJin15; @ZhangDai16;_@LiangZhang16] under different_assumptions._The_theoretical analyses suggest that_the use of the low cost_and low-resolution ADCs can still provide_satisfactory achievable rates and spectral efficiency.
In this_paper, we consider the problem of_channel estimation for uplink multiuser_massive MIMO_systems, where one-bit ADCs are_used at the_BS in_order to reduce_the cost and power consumption. Channel_estimation is crucial_to support multi-user MIMO operation in_massive_MIMO systems [@AdhikaryNam13;_@ChoiLove14;_@SunGao15;_@GaoDai15; @FangLi17]._To reach the_full_potential of_massive_MIMO, accurate downlink CSI is required_at_the BS for precoding and other operations._Most literature on massive_MIMO_systems, e.g. [@Marzetta10; @RusekPersson13;_@YinGesbert13; @MullerCottatellucci14], assumes a time_division duplex (TDD) mode in which_the downlink_CSI can_be immediately obtained from the uplink CSI by exploiting channel reciprocity._Nevertheless, channel estimation for massive MIMO_systems with one-bit ADCs_is challenging_since_the magnitude and_phase_information about_the received signal are lost or severely_distorted due_to the coarse quantization. It was_shown in [@RisiPersson14] that_one-bit_massive MIMO systems require an excessively_long training sequence (e.g. approximately 50_times the number of users)_to_achieve_an acceptable performance. The work_[@JacobssonDurisi15] showed that for one-bit massive_MIMO systems, a_least-squares channel estimation scheme and a maximum-ratio_combining_scheme are sufficient to support both_multiuser_operation and the use of high |
\in \RR^2$ and $\tb \in \RR_H^3$ as column vectors, then it follows from that $$\label{x-t-x}
x = \tfrac13 H (t_1 - t_3, t_2-t_3)^{{\mathsf {tr}}}= \tfrac13 H (2 t_1 + t_2, t_1+2t_2)^{{\mathsf {tr}}}$$ upon using the fact that $t_1 + t_2 + t_3 =0$. Computing the Jacobian of the change of variables shows that $d x = \frac{2 \sqrt{3}} {3} d t_1 dt_2$. A function $f$ is called [*periodic*]{} with respect to the hexagonal lattice, if $$f(x) = f (x + H k), \qquad k \in \ZZ^2.$$ We call such a function $H$-periodic. In homogeneous coordinates, $x\equiv y \pmod{H}$ becomes, as easily seen using, $\tb \equiv \sb \mod 3$, where we define $$\tb \equiv \sb \mod 3 \quad \Longleftrightarrow \quad t_1-s_1 \equiv t_2-s_2
\equiv t_3-s_3 \mod 3.$$ Thus, a function $f (\tb)$ is $H$-periodic if $f (\tb) = f(\tb + \jb)$ whenever $\jb \equiv 0 \mod 3$. If $f$ is $H$-periodic, then it can be verified directly that $$\label{IntPeriod}
\int_{\Omega} f(\tb + \sb) d\tb = \int_{\Omega} f(\tb) d\tb, \qquad \sb \in \RR_H^3.$$
We define the inner product on the hexagonal domain by $$\begin{aligned}
\langle f, g\rangle_H := \frac{1}{|\Omega_H|} \int_{\Omega_H} f(x_1,x_2)
\overline{g(x_1,x_2)} d x_1 dx_2
= \frac{1}{|\Omega|} \int_{\Omega} f(\tb) \overline{g(\tb)} d \tb,\end{aligned}$$ where $|\Omega|$ | \in \RR^2 $ and $ \tb \in \RR_H^3 $ as column vectors, then it follows from that $ $ \label{x - t - x }
x = \tfrac13 H (t_1 - t_3, t_2 - t_3)^{{\mathsf { tr}}}= \tfrac13 H (2 t_1 + t_2, t_1 + 2t_2)^{{\mathsf { tr}}}$$ upon using the fact that $ t_1 + t_2 + t_3 = 0$. calculate the Jacobian of the variety of variable shows that $ vitamin d x = \frac{2 \sqrt{3 } } { 3 } five hundred t_1 dt_2$. A function $ f$ is called [ * periodic * ] { } with regard to the hexangular lattice, if $ $ f(x) = f (x + H k), \qquad k \in \ZZ^2.$$ We call such a function $ H$-periodic. In homogeneous coordinates, $ x\equiv y \pmod{H}$ becomes, as well seen using, $ \tb \equiv \sb \mod 3 $, where we define $ $ \tb \equiv \sb \mod 3 \quad \Longleftrightarrow \quad t_1 - s_1 \equiv t_2 - s_2
\equiv t_3 - s_3 \mod 3.$$ Thus, a affair $ f (\tb)$ is $ H$-periodic if $ f (\tb) = f(\tb + \jb)$ whenever $ \jb \equiv 0 \mod 3$. If $ f$ is $ H$-periodic, then it can be verified directly that $ $ \label{IntPeriod }
\int_{\Omega } f(\tb + \sb) d\tb = \int_{\Omega } f(\tb) d\tb, \qquad \sb \in \RR_H^3.$$
We define the inner product on the hexangular domain by $ $ \begin{aligned }
\langle f, g\rangle_H: = \frac{1}{|\Omega_H| } \int_{\Omega_H } f(x_1,x_2)
\overline{g(x_1,x_2) } d x_1 dx_2
= \frac{1}{|\Omega| } \int_{\Omega } f(\tb) \overline{g(\tb) } d \tb,\end{aligned}$$ where $ |\Omega|$ | \in \RR^2$ and $\tb \in \RR_H^3$ as colmmn vectors, then it folnows fdom that $$\label{x-t-x}
x = \tfrac13 H (t_1 - t_3, t_2-t_3)^{{\larhsf {ug}}}= \tfrac13 H (2 t_1 + t_2, t_1+2t_2)^{{\mxthsf {tr}}}$$ lpon usint tht fact that $t_1 + t_2 + t_3 =0$. Compmcing fme Jaeouian of the chakge of varidbles shows thdt $d x = \frac{2 \sqrt{3}} {3} d t_1 dt_2$. A function $f$ ys callrd [*periodic*]{} with reskece to nht hexagonal lattice, if $$f(x) = f (x + H k), \qquav k \in \ZZ^2.$$ We cakl such a function $H$-periodlc. Ij homogeneous coorfinates, $x\eqojv r \pmod{H}$ becomds, as easily seen usinf, $\tb \equiv \sb \mod 3$, where we dewine $$\cb \equiv \sb \mid 3 \xuad \Longlehtrighnarrow \quad b_1-x_1 \equie t_2-s_2
\equiv t_3-s_3 \mpd 3.$$ Thys, a function $f (\tb)$ is $H$-periodic if $f (\tb) = f(\tb + \jb)$ wkenever $\jb \equiv 0 \moe 3$. If $f$ is $V$-perkidiz, tgei if can he terified didectly that $$\label{IntPeriod}
\int_{\Pmqta} f(\tb + \sb) d\fb = \yne_{\Omega} f(\tb) d\tb, \qquad \sb \in \RR_H^3.$$
We define ths inner product on the yexagonal domain by $$\bggin{aligneq}
\langle f, g\rangle_H := \frac{1}{|\Omega_H|} \int_{\Omega_H} f(x_1,x_2)
\oeerliie{e(x_1,x_2)} d b_1 dd_2
= \frac{1}{|\Omega|} \int_{\Omega} f(\tb) \overline{g(\tb)} d \tb,\qhd{slpgned}$$ where $|\Omega|$ | \in \RR^2$ and $\tb \in \RR_H^3$ as then follows from $$\label{x-t-x} x = t_2-t_3)^{{\mathsf \tfrac13 H (2 + t_2, t_1+2t_2)^{{\mathsf upon using the fact that $t_1 t_2 + t_3 =0$. Computing the Jacobian of the change of variables shows $d x = \frac{2 \sqrt{3}} {3} d t_1 dt_2$. A function $f$ is [*periodic*]{} respect the lattice, if $$f(x) = f (x + H k), \qquad k \in \ZZ^2.$$ We call such function $H$-periodic. In homogeneous coordinates, $x\equiv y \pmod{H}$ as easily seen using, \equiv \sb \mod 3$, where define \equiv \sb 3 \Longleftrightarrow t_1-s_1 \equiv t_2-s_2 t_3-s_3 \mod 3.$$ Thus, a function $f (\tb)$ is $H$-periodic if $f (\tb) = f(\tb + \jb)$ $\jb \equiv 3$. If is then can be verified $$\label{IntPeriod} \int_{\Omega} f(\tb + \sb) d\tb d\tb, \qquad \sb \in \RR_H^3.$$ We define the product on hexagonal domain by $$\begin{aligned} \langle f, := \frac{1}{|\Omega_H|} \int_{\Omega_H} f(x_1,x_2) \overline{g(x_1,x_2)} d x_1 dx_2 \frac{1}{|\Omega|} \int_{\Omega} f(\tb) \overline{g(\tb)} d \tb,\end{aligned}$$ where $|\Omega|$ | \in \RR^2$ and $\tb \in \RR_H^3$ as column vecTors, then it FolloWs fRom ThAt $$\laBel{x-T-x}
x = \tfrac13 H (t_1 - t_3, t_2-t_3)^{{\mAThsf {Tr}}}= \tfrac13 H (2 t_1 + t_2, t_1+2t_2)^{{\mathsf {tr}}}$$ upOn usiNg THe faCT tHat $t_1 + t_2 + T_3 =0$. ComputINg THE JaCoBiAn oF tHE cHange Of vAriableS shows that $D x = \fRaC{2 \sqrt{3}} {3} d t_1 dt_2$. A fuNCtIon $f$ is callEd [*pEriodic*]{} with rEspEct to tHe HexAGonal LatTice, iF $$f(x) = f (x + H K), \Qquad k \In \ZZ^2.$$ We calL sUCh a funCTion $H$-peRIOdIc. In Homogeneous coordiNAtES, $x\equiv y \pmod{H}$ bEcomes, As EAsILY seEn uSing, $\tb \equiV \sB \mod 3$, wHEre we deFInE $$\TB \EquIV \sb \mod 3 \quad \LonGleftrightaRRow \Quad t_1-s_1 \EqUiv T_2-S_2
\equiv T_3-s_3 \mod 3.$$ thUS, a fUnction $f (\tb)$ iS $H$-peRiodic if $f (\Tb) = f(\tb + \jB)$ WheneveR $\Jb \equiv 0 \Mod 3$. If $f$ Is $H$-PerIodiC, ThEn It cAn BE veRIfIed DIreCtly that $$\LaBeL{IntPErioD}
\INT_{\omegA} f(\tB + \sb) d\Tb = \int_{\omega} f(\tb) d\tb, \qqUad \Sb \in \rr_H^3.$$
WE defiNe the InneR pRoducT on the HexagOnAl domain by $$\begin{AligNed}
\langle F, g\rAnGle_h := \fRac{1}{|\OmEGa_H|} \int_{\omeGa_H} F(x_1,x_2)
\overLine{g(x_1,x_2)} D X_1 dx_2
= \FrAC{1}{|\oMeGa|} \int_{\Omega} f(\tb) \overlInE{G(\Tb)} D \tb,\end{alIgned}$$ wHErE $|\OMEga|$ | \in \RR^2$ and $\tb \in \ RR_H^3$ as colu mnvec to rs,then it follows fr o m th at $$\label{x-t-x}
x = \tfr ac 1 3 H (t _1 -t_3, t_ 2 -t _ 3 )^{ {\ ma ths f{ tr }}}= \t frac13 H (2 t_1+ t _2 , t_1+2t_2)^ { {\ mathsf {tr }}} $$ upon usin g t he fac ttha t $t_1 +t_2 + t_3 = 0 $. Com puting th eJ acobia n of the c ha ngeof variables show s t h at $d x = \fra c{2 \s qr t {3 } } {3 } d t_1 dt_2$ .A fun c tion $f $ i s c all e d [*periodic* ]{} with re s pec t to t he he x agonal latt ic e , i f $$f(x) =f (x + H k),\qquad k \in \ Z Z^2.$$We cal l s uch a f u nc ti on$H $ -pe r io dic . In homogen eo us coor dina t e s , $x\ equ iv y \pmo d{H}$ becomes , a s ea s ily seen usin g, $ \t b \e quiv \ sb \m od 3$, where we d efin e $$\tb \ equ iv \s b\mod3 \quad \L ong leftrig htarrow \qu ad t _1 -s_1 \equiv t_2-s_ 2 \equi v t_3- s _3 \ m od 3.$$Th us, a f u n ction $f( \t b)$ is $ H$-per i od ic if $f(\ tb) =f( \tb +\jb)$ when ever $ \jb \eq uiv 0 \mod 3$. If $f $ is $H$-perio d ic , th e n it ca n be verifi ed d i rect ly t h at $$ \ label {IntP er i od }
\int_{\Omega} f(\ tb + \s b) d\ tb = \int_{ \Omega} f( \ t b ) d\tb,\qqu a d\ sb \in \RR_H^3 .$$
We definet he inner prod uct on t he hexago n a l domain by $$ \be gin { a li gned}
\langle f , g\ ra ngle_H:=\frac{1 }{| \Om ega _H| }\int_{\Om ega_H} f (x _1 ,x _2 ) \o v erline{g (x _1, x_ 2)} d x_ 1 dx_2 = \fr ac {1 } {|\ Omega|} \i n t _{\O me ga } f( \tb )\over line { g(\ tb)} d\tb,\end{ ali g ned} $$ w here $| \Omega|$ | \in_\RR^2$ and_$\tb \in \RR_H^3$ as_column vectors,_then_it follows_from_that $$\label{x-t-x}
x =_\tfrac13 H_(t_1 - t_3, t_2-t_3)^{{\mathsf_{tr}}}= \tfrac13__H (2 t_1 + t_2, t_1+2t_2)^{{\mathsf {tr}}}$$ upon using the fact that $t_1 +_t_2_+ t_3_=0$._Computing_the Jacobian of the change_of variables shows that $d_x =_\frac{2 \sqrt{3}} {3} d t_1 dt_2$. A function_$f$_is called [*periodic*]{}_with respect to the hexagonal lattice, if $$f(x) =_f (x + H k), \qquad_k \in \ZZ^2.$$_We_call_such a function $H$-periodic._In homogeneous coordinates, $x\equiv y \pmod{H}$_becomes, as easily seen using, $\tb_ \equiv \sb \mod 3$, where we_define $$\tb \equiv \sb \mod 3_\quad \Longleftrightarrow \quad t_1-s_1_\equiv t_2-s_2_
_ \equiv t_3-s_3_\mod 3.$$_Thus, a function_$f (\tb)$ is $H$-periodic if $f_(\tb) = f(\tb_+ \jb)$ whenever $\jb \equiv_0_\mod 3$. If_$f$_is_$H$-periodic, then_it can be_verified_directly that_$$\label{IntPeriod}
_\int_{\Omega} f(\tb + \sb) d\tb_=_ \int_{\Omega} f(\tb) d\tb, \qquad \sb_\in \RR_H^3.$$
We define the_inner_product on the hexagonal_domain by $$\begin{aligned}
\langle f, g\rangle_H_:= \frac{1}{|\Omega_H|} \int_{\Omega_H} f(x_1,x_2)
_ \overline{g(x_1,x_2)}_d x_1_dx_2
= \frac{1}{|\Omega|} \int_{\Omega} f(\tb) \overline{g(\tb)} d \tb,\end{aligned}$$_where $|\Omega|$ |
the reverse deterministic flow (i.e. ${\partial}_t \varphi^u = - ({\partial}^2_{xx} \varphi^u + g(\varphi^u)) )$ during the time interval $[t_1,t_2]$ then, similarly to Proposition \[Lyapunov\], we have $$\frac{1}{2}\int_{t_1}^{t_2} \int_0^1 |{\partial}_t \varphi^u - ({\partial}^2_{xx} \varphi^u + g(\varphi^u))|^{2} = 2 \int_{t_1}^{t_2} \frac{d S(\varphi^u(t,\cdot))}{dt} = 2 (S(\varphi^u(t_2)) - S(\varphi^u(t_1)))$$ from where, upon recalling that $\varphi^u(t_2),\varphi^u(t_1) \in \mathcal{W}^z_u$, we obtain that the rate of this last piece is less than $\Delta$.
- Any path at a distance strictly less than $r$ from $\varphi^u$ in the supremum norm must also escape from $G$ before $T^u$, since by this time $\varphi^u$ reaches a distance $r$ from $G$.
Let us notice that for each $u \in G$ we have built a path $\varphi^u$ on the time interval $[0,T^u]$, but we wish all constructed paths to be defined on a same time interval. For this reason, we consider $T^{(\delta)}:= \sup_{u \in G} T^u < +\infty$ and extend all $\varphi^u$ to the time interval $[0,T^{(\delta)}]$ by following the deterministic flow. It is easy to check that these extended paths maintain the aforementioned properties. We then define the set $\mathcal{E}_{u,T^{(\delta)}}$ for each $u \in G$ as $$\mathcal{E}_{u,T^{(\delta)}}:=\{ \psi \in C_{D_u}([0,T^{(\delta)}] \times [0,1]) : d_{T^{(\delta)}}(\psi, \varphi^u ) < r\}.$$ It is clear that each $\mathcal{E}_{u,T^{(\delta)}}$ verifies condition (i) by construction, whereas (ii) follows from the | the reverse deterministic flow (i.e. $ { \partial}_t \varphi^u = - ({ \partial}^2_{xx } \varphi^u + g(\varphi^u) )) $ during the time interval $ [ t_1,t_2]$ then, similarly to Proposition \[Lyapunov\ ], we get $ $ \frac{1}{2}\int_{t_1}^{t_2 } \int_0 ^ 1 |{\partial}_t \varphi^u - ({ \partial}^2_{xx } \varphi^u + g(\varphi^u))|^{2 } = 2 \int_{t_1}^{t_2 } \frac{d S(\varphi^u(t,\cdot))}{dt } = 2 (S(\varphi^u(t_2) ) - S(\varphi^u(t_1)))$$ from where, upon echo that $ \varphi^u(t_2),\varphi^u(t_1) \in \mathcal{W}^z_u$, we obtain that the rate of this last musical composition is less than $ \Delta$.
- Any path at a distance strictly less than $ r$ from $ \varphi^u$ in the supremum average must also escape from $ G$ before $ T^u$, since by this time $ \varphi^u$ strive a distance $ r$ from $ G$.
Let us notice that for each $ uranium \in G$ we have built a path $ \varphi^u$ on the fourth dimension time interval $ [ 0,T^u]$, but we wish all constructed paths to be define on a same time time interval. For this reason, we consider $ T^{(\delta)}:= \sup_{u \in G } T^u < + \infty$ and run all $ \varphi^u$ to the time interval $ [ 0,T^{(\delta)}]$ by follow the deterministic stream. It is easy to discipline that these extended paths maintain the aforesaid properties. We then define the set $ \mathcal{E}_{u, T^{(\delta)}}$ for each $ u \in G$ as $ $ \mathcal{E}_{u, T^{(\delta)}}:=\ { \psi \in C_{D_u}([0,T^{(\delta) } ] \times [ 0,1 ]): d_{T^{(\delta)}}(\psi, \varphi^u) < r\}.$$ It is clear that each $ \mathcal{E}_{u, T^{(\delta)}}$ verifies condition (iodine) by construction, whereas (ii) follows from the | thf reverse deterministic nlow (i.e. ${\partial}_t \varphi^n = - ({\parfial}^2_{xx} \vxrphi^u + g(\varphi^u)) )$ during the tume ibterval $[t_1,t_2]$ then, similafly to Prlpositiob \[Lyepunov\], we have $$\fczc{1}{2}\int_{t_1}^{t_2} \int_0^1 |{\pzvtial}_c \tarphi^u - ({\partiak}^2_{xx} \varphi^g + g(\varphi^u))|^{2} = 2 \hng_{t_1}^{c_2} \frac{d S(\varphi^u(t,\cdot))}{dt} = 2 (S(\varphi^u(t_2)) - S(\varpni^k(t_1)))$$ from where, opon gesalljng that $\varphi^u(t_2),\varphi^u(t_1) \in \mathczl{W}^z_u$, wt obtain that the tate of this last piece is lesd than $\Delta$.
- Any path at a eistwbce strictly less than $r$ from $\varpgi^u$ in the supremum norm must auso excape from $G$ bevmre $T^u$, sincx by tris time $\varini^u$ redches a distance $r$ frpm $J$.
Let us notice that for eech $u \in G$ we have boilt a patv $\rarphi^u$ on the time ibtwrval $[0,T^u]$, tut dw wksh akl constgucved paths tk be definee on a same time inuerdql. For this rsason, re consider $T^{(\delta)}:= \sup_{u \in G} T^u < +\infty$ ang estend all $\varphi^u$ to thw time interval $[0,T^{(\deltw)}]$ by folljwing the deterministic flow. It is easy to check dhat vhdse ewtenawd paths maintain the aforementioned properties. Se tmen define the sct $\mathcal{E}_{u,T^{(\delta)}}$ flr gach $u \in G$ as $$\mathccm{E}_{h,T^{(\delta)}}:=\{ \psi \in C_{D_u}([0,H^{(\delta)}] \jimes [0,1]) : d_{T^{(\delta)}}(\ksi, \vsrphi^u ) < r\}.$$ It is clear that each $\mathcaj{W}_{u,T^{(\delta)}}$ verifies eondition (i) yy conxtrucyion, whereas (ii) follows froj the | the reverse deterministic flow (i.e. ${\partial}_t \varphi^u ({\partial}^2_{xx} + g(\varphi^u)) during the time Proposition we have $$\frac{1}{2}\int_{t_1}^{t_2} |{\partial}_t \varphi^u - \varphi^u + g(\varphi^u))|^{2} = 2 \int_{t_1}^{t_2} S(\varphi^u(t,\cdot))}{dt} = 2 (S(\varphi^u(t_2)) - S(\varphi^u(t_1)))$$ from where, upon recalling that $\varphi^u(t_2),\varphi^u(t_1) \in we obtain that the rate of this last piece is less than $\Delta$. Any at distance less than $r$ from $\varphi^u$ in the supremum norm must also escape from $G$ before $T^u$, by this time $\varphi^u$ reaches a distance $r$ $G$. Let us notice for each $u \in G$ have a path on time $[0,T^u]$, but we all constructed paths to be defined on a same time interval. For this reason, we consider $T^{(\delta)}:= \in G} +\infty$ and all to time interval $[0,T^{(\delta)}]$ the deterministic flow. It is easy these extended paths maintain the aforementioned properties. We define the $\mathcal{E}_{u,T^{(\delta)}}$ for each $u \in G$ $$\mathcal{E}_{u,T^{(\delta)}}:=\{ \psi \in C_{D_u}([0,T^{(\delta)}] \times [0,1]) : d_{T^{(\delta)}}(\psi, ) < r\}.$$ It is clear that each $\mathcal{E}_{u,T^{(\delta)}}$ verifies condition (i) by construction, whereas from the | the reverse deterministic flOw (i.e. ${\partiaL}_t \varPhi^U = - ({\paRtIal}^2_{xX} \varPhi^u + g(\varphi^u)) )$ duRIng tHe time interval $[t_1,t_2]$ then, siMilarLy TO ProPOsItion \[lyapunoV\], We HAVe $$\fRaC{1}{2}\iNt_{t_1}^{T_2} \iNT_0^1 |{\pArtiaL}_t \vArphi^u - ({\pArtial}^2_{xx} \vaRphI^u + G(\varphi^u))|^{2} = 2 \int_{t_1}^{T_2} \FrAc{d S(\varphi^U(t,\cDot))}{dt} = 2 (S(\varphi^U(t_2)) - S(\Varphi^U(t_1)))$$ FroM Where, UpoN recaLling tHAt $\varpHi^u(t_2),\varphI^u(T_1) \In \mathCAl{W}^z_u$, we OBTaIn thAt the rate of this laST pIEce is less than $\DElta$.
- AnY pATh AT A diStaNce strictlY lEss thAN $r$ from $\vARpHI^U$ In tHE supremum norm Must also escAPe fRom $G$ beFoRe $T^U$, Since bY this TiME $\vaRphi^u$ reacheS a diStance $r$ frOm $G$.
Let US notice THat for eAch $u \in g$ we HavE buiLT a PaTh $\vArPHi^u$ ON tHe tIMe iNterval $[0,T^U]$, bUt We wisH all CONSTrucTed PathS to be Defined on a samE tiMe inTErvAl. For This rEasoN, wE consIder $T^{(\dElta)}:= \sUp_{U \in G} T^u < +\infty$ and eXtenD all $\varphI^u$ tO tHe tImE inteRVal $[0,T^{(\deLta)}]$ By fOllowinG the detERmiNiSTIC fLow. It is easy to check ThAT ThEse extenDed patHS mAiNTain the aFoRemEntiONEd proPertIEs. we then deFine thE SeT $\mAthcal{E}_{U,T^{(\Delta)}}$ fOr EacH $u \iN G$ as $$\mAThcaL{E}_{u,T^{(\deLta)}}:=\{ \psi \in c_{D_u}([0,T^{(\dELta)}] \times [0,1]) : d_{T^{(\deltA)}}(\Psi, \varphi^u ) < r\}.$$ It IS cLEAr THat eAch $\Mathcal{E}_{u,T^{(\dElta)}}$ VErifIes cONdItiON (i) by cOnstrUcTIoN, Whereas (ii) follows froM tHe | the reverse deterministic flow (i.e . ${\ par tia l} _t \ varp hi^u = - ({\pa r tial }^2_{xx} \varphi^u + g (\var ph i ^u)) )$ duri ng thet im e int er va l $ [t _ 1, t_2]$ th en, sim ilarly toPro po sition \[Lya p un ov\], we h ave $$\frac{1}{ 2}\ int_{t _1 }^{ t _2} \ int _0^1|{\par t ial}_t \varphi^ u- ({\pa r tial}^2 _ { xx } \v arphi^u + g(\varp h i^ u ))|^{2} = 2 \i nt_{t_ 1} ^ {t _ 2 } \ fra c{d S(\var ph i^u(t , \cdot)) } {d t } = 2 (S(\varphi^u( t_2)) - S( \ var phi^u( t_ 1)) ) $$ fro m whe re , up on recallin g th at $\varp hi^u(t _ 2),\var p hi^u(t_ 1) \in \m ath cal{ W }^ z_ u$, w e ob t ai n t h atthe rate o fthislast p i e ce i s l essthan$\Delta$.
- A ny p a that adista ncest rictl y less than $ r$ from $\varph i^u$ in the s upr em umno rm mu s t also es cap e from$G$ bef o re$T ^ u $ ,since by this time $ \ v ar phi^u$ r eaches adi s tance $r $fro m $G $ .
Let usn ot ice that for e a ch $ u \in G $we hav ebui lta pat h $\v arphi^ u$ on th e tim e interval $[0, T ^u]$, but wew is h al l con str ucted paths tob e de fine d o n a sametimein t er v al. For this reason ,we con sider $T^{(\delta) }:= \sup_{ u \ in G} T^ u <+ \i n fty$ and exten d all $\varphi^ u $ to the time interva l $[0,T^{ ( \ delta)}] $ b y f oll owi n g t he determinis t i c fl ow . It is ea sy to c hec k t hat th es e extende d pathsma in ta in th e afo r emention ed pr op ert ies.W e then defi ne t he s e t $ \mathca l {E } _ {u,T ^{ (\ delt a)} }$ foreach $u\in G$as $$\mat hca l {E}_ {u ,T ^{(\del ta)}}:=\{ \ps i\in C_{D_u }( [0, T^{(\d e l ta)}] \t imes [0,1]) : d_{T^{(\d e lta)}}( \ps i, \v arph i^u ) < r \}. $$ Itisc lear t hat ea ch $\ ma thc a l {E}_{ u , T^ {(\ de lta)}}$ ve r i fie s con di tion (i) by construction, whe r eas (ii) follows fr om t h e | the_reverse deterministic_flow (i.e. ${\partial}_t \varphi^u_= -_({\partial}^2_{xx}_\varphi^u +_g(\varphi^u))_)$ during the_time interval $[t_1,t_2]$_then, similarly to Proposition_\[Lyapunov\], we have_$$\frac{1}{2}\int_{t_1}^{t_2}_\int_0^1 |{\partial}_t \varphi^u - ({\partial}^2_{xx} \varphi^u + g(\varphi^u))|^{2} = 2 \int_{t_1}^{t_2} \frac{d S(\varphi^u(t,\cdot))}{dt} =_2_(S(\varphi^u(t_2)) -__S(\varphi^u(t_1)))$$_from where, upon recalling that_$\varphi^u(t_2),\varphi^u(t_1) \in \mathcal{W}^z_u$, we obtain_that the_rate of this last piece is less than_$\Delta$.
-_ Any_path at a distance strictly less than $r$ from_$\varphi^u$ in the supremum norm must_also escape from_$G$_before_$T^u$, since by this_time $\varphi^u$ reaches a distance $r$_from $G$.
Let us notice that for_each $u \in G$ we have built_a path $\varphi^u$ on the time_interval $[0,T^u]$, but we wish_all constructed_paths to be defined on_a same time_interval. For_this reason, we_consider $T^{(\delta)}:= \sup_{u \in G} T^u_< +\infty$ and_extend all $\varphi^u$ to the time_interval_$[0,T^{(\delta)}]$ by following_the_deterministic_flow. It_is easy to_check_that these_extended_paths maintain the aforementioned properties. We_then_define the set $\mathcal{E}_{u,T^{(\delta)}}$ for each $u_\in G$ as $$\mathcal{E}_{u,T^{(\delta)}}:=\{_\psi_\in C_{D_u}([0,T^{(\delta)}] \times [0,1])_: d_{T^{(\delta)}}(\psi, \varphi^u ) <_r\}.$$ It is clear that each_$\mathcal{E}_{u,T^{(\delta)}}$ verifies_condition (i)_by construction, whereas (ii) follows from the |
{g}^*})$ is compact it is proper, by Lemma \[vanishing\_proposition\] (2), we have that $H^2_{\mathrm{diff}}(G\ltimes \mathbb{S}({\mathfrak{g}^*});E)=0$. The corollary follows from Theorem \[Theorem\_THREE\].
Let $P$ be the Poisson homotopy bundle of $S$, and let $G$ be its structure group. By hypothesis, $P$ is smooth, $1$-connected and has vanishing second de Rham cohomology. Let $\mathcal{G}:=P\times_{G}P$ be the gauge groupoid of $P$. The $s$-fibers of $\mathcal{G}$ are diffeomorphic to $P$; thus, $\mathcal{G}$ satisfies the assumptions of Lemma \[vanishing\_proposition\] (1). Therefore $$H^{2}(A_S;\mathcal{S}^k(\nu_S^*))\cong H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*)).$$ Since $\mathcal{G}$ is transitive, by Lemma \[vanishing\_proposition\] (3), we have that $$H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*))\cong H^{2}_{\mathrm{diff}}(G;\mathcal{S}^k(\nu_{S,x}^*)).$$ Since $\nu_{S,x}^*\cong \mathfrak{g}$ as $G$ representations, Theorem \[Theorem1\] implies the conclusion.
By Lemma \[vanishing\_proposition\] (2), the differentiable cohomology of compact groups vanishes; thus, Corollary \[Theorem2\] implies the result.
Let $x\in S$ and denote by $\mathfrak{g}_x:=\nu_{S,x}^*$ the isotropy Lie algebra of the transitive algebroid $A_S$. By hypothesis, $\mathfrak{g}_x$ is reductive, i.e. it splits as a direct product of a semisimple Lie algebra and its center $\mathfrak{g}_x=\mathfrak{s}_x\oplus \mathfrak{z}_x$, where $\mathfrak{s}_x=[\mathfrak{g}_x,\mathfrak{g}_x]$ and $\mathfrak{z}_x=Z(\mathfrak{g}_x)$ is the center of $\mathfrak{g}_x$. Since $\mathfrak{g}:=\nu | { g}^*})$ is compact it is proper, by Lemma \[vanishing\_proposition\ ] (2), we have that $ H^2_{\mathrm{diff}}(G\ltimes \mathbb{S}({\mathfrak{g}^*});E)=0$. The corollary follows from Theorem \[Theorem\_THREE\ ].
Let $ P$ be the Poisson homotopy package of $ S$, and lease $ G$ be its structure group. By hypothesis, $ P$ is smooth, $ 1$-connected and has vanish second de Rham cohomology. Let $ \mathcal{G}:=P\times_{G}P$ be the gauge groupoid of $ P$. The $ s$-fibers of $ \mathcal{G}$ are diffeomorphic to $ P$; therefore, $ \mathcal{G}$ satisfies the assumptions of Lemma \[vanishing\_proposition\ ] (1). Therefore $ $ H^{2}(A_S;\mathcal{S}^k(\nu_S^*))\cong H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*)).$$ Since $ \mathcal{G}$ is transitive, by Lemma \[vanishing\_proposition\ ] (3), we suffer that $ $ H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*))\cong H^{2}_{\mathrm{diff}}(G;\mathcal{S}^k(\nu_{S, x}^*)).$$ Since $ \nu_{S, x}^*\cong \mathfrak{g}$ as $ G$ representations, Theorem \[Theorem1\ ] implies the termination.
By Lemma \[vanishing\_proposition\ ] (2), the differentiable cohomology of compact groups vanishes; therefore, Corollary \[Theorem2\ ] implies the result.
Let $ x\in S$ and denote by $ \mathfrak{g}_x:=\nu_{S, x}^*$ the isotropy Lie algebra of the transitive algebroid $ A_S$. By hypothesis, $ \mathfrak{g}_x$ is reductive, i.e. it splits as a direct product of a semisimple Lie algebra and its center field $ \mathfrak{g}_x=\mathfrak{s}_x\oplus \mathfrak{z}_x$, where $ \mathfrak{s}_x=[\mathfrak{g}_x,\mathfrak{g}_x]$ and $ \mathfrak{z}_x = Z(\mathfrak{g}_x)$ is the center of $ \mathfrak{g}_x$. Since $ \mathfrak{g}:=\nu | {g}^*})$ id compact it is proper, bn Lemma \[vanishiny\_proposmtion\] (2), se have ghat $H^2_{\mathrm{diff}}(G\ltimes \mathub{S}({\mqthfrqk{g}^*});E)=0$. The corollary foluows from Theorem \[Theieem\_THREE\].
Lev $P$ be tmz Poiadon konotopy bundle pf $S$, and lat $G$ be its stsuztbre group. By hypothesis, $P$ is smooth, $1$-sonnectrd and has vanisring fecohd de Rham cohomology. Let $\mathcal{G}:=L\times_{G}K$ be the gauge groipoid of $P$. The $s$-fibers of $\lathfal{G}$ are diffeomorohic to $P$; tyus, $\iqthcal{G}$ satirfies the assumptions kf Lemma \[vanishing\_proposition\] (1). Ghereyore $$H^{2}(A_S;\matycql{S}^n(\tu_S^*))\cong H^{2}_{\mavhrm{disf}}(\mathcal{G};\mabncal{S}^k(\tu_S^*)).$$ Sinve $\mathcal{G}$ is trensirive, by Lemma \[vanishiig\_proposition\] (3), we hade that $$H^{2}_{\kachrm{diff}}(\mathcal{G};\mathcql{W}^k(\nu_S^*))\wong H^{2}_{\magyrm{aifr}}(G;\kafhcal{S}^n(\nu_{A,x}^*)).$$ Since $\nh_{S,x}^*\cong \matyfrak{g}$ as $G$ represemtwnoons, Theorem \[Theorqm1\] implies the conclusion.
By Lemma \[vanishitg\_pdoposition\] (2), the differebtiable cohomology of compact droups vanishes; thus, Corollary \[Theorem2\] implies tha resnlg.
Leu $w\ln S$ qnf denote by $\mathfrak{g}_x:=\nu_{S,x}^*$ the isotropy Lie ajfenrs of the transltive algebroid $A_S$. Bj nipothesis, $\mathwrak{g}_x$ is reductive, i.e. it sppits as a dieect prodtct pf a semisimple Lie algebra and its cenner $\nathfrak{g}_x=\mathfrak{d}_x\oplus \matkfrak{z}_c$, whete $\mathfrak{s}_x=[\mathfrak{g}_x,\oathrrak{g}_x]$ and $\lathfrak{z}_s=X(\mathfrak{g}_x)$ is tfe bentar of $\mauffrak{g}_x$. Since $\matrfrak{g}:=\nu | {g}^*})$ is compact it is proper, by (2), have that \mathbb{S}({\mathfrak{g}^*});E)=0$. The corollary $P$ the Poisson homotopy of $S$, and $G$ be its structure group. By $P$ is smooth, $1$-connected and has vanishing second de Rham cohomology. Let $\mathcal{G}:=P\times_{G}P$ the gauge groupoid of $P$. The $s$-fibers of $\mathcal{G}$ are diffeomorphic to $P$; $\mathcal{G}$ the of \[vanishing\_proposition\] (1). Therefore $$H^{2}(A_S;\mathcal{S}^k(\nu_S^*))\cong H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*)).$$ Since $\mathcal{G}$ is transitive, by Lemma \[vanishing\_proposition\] (3), we have that H^{2}_{\mathrm{diff}}(G;\mathcal{S}^k(\nu_{S,x}^*)).$$ Since $\nu_{S,x}^*\cong \mathfrak{g}$ as $G$ representations, Theorem implies the conclusion. By \[vanishing\_proposition\] (2), the differentiable cohomology compact vanishes; thus, \[Theorem2\] the Let $x\in S$ denote by $\mathfrak{g}_x:=\nu_{S,x}^*$ the isotropy Lie algebra of the transitive algebroid $A_S$. By hypothesis, $\mathfrak{g}_x$ is reductive, it splits direct product a Lie and its center where $\mathfrak{s}_x=[\mathfrak{g}_x,\mathfrak{g}_x]$ and $\mathfrak{z}_x=Z(\mathfrak{g}_x)$ is the Since $\mathfrak{g}:=\nu | {g}^*})$ is compact it is proper, by LemMa \[vanishinG\_propOsiTioN\] (2), wE havE thaT $H^2_{\mathrm{diff}}(G\lTImes \Mathbb{S}({\mathfrak{g}^*});E)=0$. The coRollaRy FOlloWS fRom ThEorem \[ThEOrEM\_tHReE\].
leT $P$ bE tHE POissoN hoMotopy bUndle of $S$, anD leT $G$ Be its structuRE gRoup. By hypoTheSis, $P$ is smooth, $1$-ConNected AnD haS VanisHinG secoNd de RhAM cohomOlogy. Let $\mAtHCal{G}:=P\tIMes_{G}P$ be THE gAuge Groupoid of $P$. The $s$-fiBErS Of $\mathcal{G}$ are dIffeomOrPHiC TO $P$; tHus, $\Mathcal{G}$ saTiSfies THe assumPTiONS Of LEMma \[vanishing\_pRoposition\] (1). THEreFore $$H^{2}(A_s;\mAthCAl{S}^k(\nu_s^*))\cong h^{2}_{\mAThrM{diff}}(\mathcaL{G};\maThcal{S}^k(\nu_s^*)).$$ Since $\MAthcal{G}$ IS transiTive, by lemMa \[vAnisHInG\_pRopOsITioN\] (3), We HavE ThaT $$H^{2}_{\mathrm{DiFf}}(\MathcAl{G};\mATHCAl{S}^k(\Nu_S^*))\Cong h^{2}_{\mathRm{diff}}(G;\mathcaL{S}^k(\Nu_{S,x}^*)).$$ sIncE $\nu_{S,x}^*\Cong \mAthfRaK{g}$ as $G$ RepresEntatIoNs, Theorem \[TheoreM1\] impLies the coNclUsIon.
by lemma \[VAnishiNg\_pRopOsition\] (2), The diffERenTiABLE cOhomology of compact GrOUPs Vanishes; Thus, CoROlLaRY \[Theorem2\] ImPliEs thE REsult.
let $x\IN S$ And denotE by $\matHFrAk{G}_x:=\nu_{S,x}^*$ tHe IsotroPy lie AlgEbra oF The tRansitIve algebRoid $A_s$. by hypothesis, $\maTHfrak{g}_x$ is reduCTiVE, I.e. IT splIts As a direct prOducT Of a sEmisIMpLe LIE algeBra anD iTS cENter $\mathfrak{g}_x=\mathfRaK{s}_x\oplUs \matHfrak{z}_x$, where $\mAthfrak{s}_x=[\mATHFrak{g}_x,\maThfrAK{g}_X]$ And $\mathfrak{z}_x=Z(\MathfRak{g}_x)$ is the CEnter of $\mAthfrAk{g}_x$. SincE $\mathfrak{G}:=\NU | {g}^*})$ is compact it isproper, by Lemm a \ [va ni shin g\_p roposition\] ( 2 ), w e have that $H^2_{\mat hrm{d if f }}(G \ lt imes\mathbb { S} ( { \ma th fr ak{ g} ^ *} );E)= 0$. The co rollary fo llo ws from Theore m \ [Theorem\_ THR EE\].
Let $ P$be the P ois s on ho mot opy b undleo f $S$, and let$G $ be it s struct u r egrou p. By hypothesis, $P $ is smooth, $1 $-conn ec t ed a ndhas vanishing s econd de Rham co h o m olo g y. Let $\math cal{G}:=P\t i mes _{G}P$ b e t h e gaug e gro up o idof $P$. The $s$ -fibers o f $\ma t hcal{G} $ are di ffeomo rph icto $ P $; t hus ,$ \ma t hc al{ G }$satisfie sth e ass umpt i o n s ofLem ma \ [vani shing\_propos iti on\] (1) . The refor e $$ H^ {2}(A _S;\ma thcal {S }^k(\nu_S^*))\c ongH^{2}_{\m ath rm {di ff }}(\m a thcal{ G}; \ma thcal{S }^k(\nu _ S^* )) . $ $ S ince $\mathcal{G}$ i s tr ansitive , by L e mm a\ [vanishi ng \_p ropo s i tion\ ] (3 ) ,we havethat $ $ H^ {2 }_{\mat hr m{diff }} (\m ath cal{G } ;\ma thcal{ S}^k(\nu _S^*) ) \cong H^{2}_{\ m athrm{diff}}( G ;\ m a th c al{S }^k (\nu_{S,x}^ *)). $ $ Si nce$ \n u_{ S ,x}^* \cong \ m at h frak{g}$ as $G$ rep re sentat ions, Theorem \[Th eorem1\] i m p l ies theconc l us i on.
By Lemma\[van ishing\_pr o position \] (2 ), the d ifferenti a b le cohom olo gyofcom p a ct groups vanis h e s; t hu s, Coro lla ry \[Th eor em2 \]imp li es the re sult.
L et $ x\ in S$ andd enote by $ \ma th fra k{g}_ x :=\nu_ {S,x} ^*$th ei sot ropy Li e a l g ebra o fthetra ns itive alg e bro id $A_S $. By hyp oth e sis, $ \m athfrak {g}_x$ is red uc tive, i.e. i t s plitsa s a direc t product of a semisimp l e Lie a lge bra a nd i ts center $\ mathfr ak{ g }_x=\m athfra k{s}_ x\ opl u s \mat h f ra k{z }_ x$, where$ \ mat hfrak {s }_x= [\mathf rak{g}_x,\mathfrak { g}_ x]$ and $\mat hfr ak{z } _ x= Z(\ m at h fra k{ g }_x ) $ is the centerof $\mathf ra k {g }_x$. Sinc e $\ ma thfrak{ g}:=\nu | {g}^*})$ is_compact it_is proper, by Lemma_\[vanishing\_proposition\] (2),_we_have that_$H^2_{\mathrm{diff}}(G\ltimes_\mathbb{S}({\mathfrak{g}^*});E)=0$. The corollary_follows from Theorem_\[Theorem\_THREE\].
Let $P$ be the_Poisson homotopy bundle_of_$S$, and let $G$ be its structure group. By hypothesis, $P$ is smooth, $1$-connected_and_has vanishing_second_de_Rham cohomology. Let $\mathcal{G}:=P\times_{G}P$ be_the gauge groupoid of $P$._The $s$-fibers_of $\mathcal{G}$ are diffeomorphic to $P$; thus, $\mathcal{G}$_satisfies_the assumptions of_Lemma \[vanishing\_proposition\] (1). Therefore $$H^{2}(A_S;\mathcal{S}^k(\nu_S^*))\cong H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*)).$$ Since $\mathcal{G}$ is_transitive, by Lemma \[vanishing\_proposition\] (3), we_have that $$H^{2}_{\mathrm{diff}}(\mathcal{G};\mathcal{S}^k(\nu_S^*))\cong_H^{2}_{\mathrm{diff}}(G;\mathcal{S}^k(\nu_{S,x}^*)).$$_Since_$\nu_{S,x}^*\cong \mathfrak{g}$ as $G$_representations, Theorem \[Theorem1\] implies the conclusion.
By_Lemma \[vanishing\_proposition\] (2), the differentiable cohomology_of compact groups vanishes; thus, Corollary \[Theorem2\]_implies the result.
Let $x\in S$ and_denote by $\mathfrak{g}_x:=\nu_{S,x}^*$ the isotropy_Lie algebra_of the transitive algebroid $A_S$._By hypothesis, $\mathfrak{g}_x$_is reductive,_i.e. it splits as_a direct product of a semisimple_Lie algebra and_its center $\mathfrak{g}_x=\mathfrak{s}_x\oplus \mathfrak{z}_x$, where $\mathfrak{s}_x=[\mathfrak{g}_x,\mathfrak{g}_x]$_and_$\mathfrak{z}_x=Z(\mathfrak{g}_x)$ is the_center_of_$\mathfrak{g}_x$. Since_$\mathfrak{g}:=\nu |
finite words are closed under sum. The cost is $O(n_1\cdot n_2)$ for ${{\sf Last}}$- and ${\mathsf{Sum}}$-automata, and $O(n_1\cdot m_1 \cdot n_2 \cdot m_2)$ for ${\mathsf{Sup}}$-automata.
It is easy to see that the synchronized product of two ${{\sf Last}}$-automata (resp. ${\mathsf{Sum}}$-automata) defines the sum of their languages if the weight of a joint transition is defined as the sum of the weights of the corresponding transitions in the two ${{\sf Last}}$-automata (resp. ${\mathsf{Sum}}$-automata).
The construction for ${\mathsf{Sup}}$-automata is the same as for ${\mathsf{Sup}}$-automata over infinite words given in the proof of \[theo:max-closed-under-sum\].
[|l|\*[4]{}[c|]{}]{} & max. & min. & comp. & sum\
${\mathsf{Sup}}$ & & & [$\times$]{}&\
${{\sf Last}}$ & & & &\
Det. ${\mathsf{Sum}}$ & [$\times$]{}& & &\
Nondet. ${\mathsf{Sum}}$ & & [$\times$]{}& [$\times$]{}&\
[|l|\*[4]{}[c|]{}]{} & max. & min. & comp. & sum\
[[ Sup ]{}]{}& & & [$\times$]{}&\
[[ Linf ]{}]{}& & & [$\times$]{}&\
[[DLsup]{}]{}& & & [$\times$]{}&\
[[NLsup]{}]{}& & & &\
[[DLavg]{}]{}& [$\times$]{}& [$\times$]{}& [$\times$]{}& [$\times$]{}\
[[NLavg]{}]{}& & [$\times$]{}& [$\times$]{}& [$\times$]{}\
[[DDisc]{}]{}& [$\times$]{}& [$\times$]{}& &\
[[NDisc]{}]{}& & [$\times$]{}& [$\times$]{}&\
Closure under $\max$ for infinite words
---------------------------------------
The maximum of two quantitative languages defined by nondeterministic automata can be obtained by an initial nondeterministic choice between the two automata. This observation was also made in [@DrosteR07] for discounted-sum automata. For deterministic automata, a synchronized product can be used for ${\mathsf{Sup}}$ and ${\mathsf{LimSup}}$, while for ${\mathsf{LimInf | finite words are closed under sum. The cost is $ O(n_1\cdot n_2)$ for $ { { \sf Last}}$- and $ { \mathsf{Sum}}$-automata, and $ O(n_1\cdot m_1 \cdot n_2 \cdot m_2)$ for $ { \mathsf{Sup}}$-automata.
It is comfortable to examine that the synchronized product of two $ { { \sf Last}}$-automata (resp. $ { \mathsf{Sum}}$-automata) define the union of their languages if the weight of a joint conversion is define as the sum of the weight of the corresponding transitions in the two $ { { \sf Last}}$-automata (resp. $ { \mathsf{Sum}}$-automata).
The structure for $ { \mathsf{Sup}}$-automata is the same as for $ { \mathsf{Sup}}$-automata over infinite words given in the validation of \[theo: max - closed - under - sum\ ].
[ |l|\*[4]{}[c| ] { } ] { } & max. & min. & comp. & sum\
$ { \mathsf{Sup}}$ & & & [ $ \times$]{}&\
$ { { \sf Last}}$ & & & & \
Det. $ { \mathsf{Sum}}$ & [ $ \times$ ] { } & & & \
Nondet. $ { \mathsf{Sum}}$ & & [ $ \times$ ] { } & [ $ \times$]{}&\
[ |l|\*[4]{}[c| ] { } ] { } & max. & min. & comp. & sum\
[ [ Sup ] { } ] { } & & & [ $ \times$]{}&\
[ [ Linf ] { } ] { } & & & [ $ \times$]{}&\
[ [ DLsup ] { } ] { } & & & [ $ \times$]{}&\
[ [ NLsup ] { } ] { } & & & & \
[ [ DLavg ] { } ] { } & [ $ \times$ ] { } & [ $ \times$ ] { } & [ $ \times$ ] { } & [ $ \times$]{}\
[ [ NLavg ] { } ] { } & & [ $ \times$ ] { } & [ $ \times$ ] { } & [ $ \times$]{}\
[ [ DDisc ] { } ] { } & [ $ \times$ ] { } & [ $ \times$ ] { } & & \
[ [ NDisc ] { } ] { } & & [ $ \times$ ] { } & [ $ \times$]{}&\
Closure under $ \max$ for countless words
---------------------------------------
The utmost of two quantitative linguistic process defined by nondeterministic automata can be obtained by an initial nondeterministic option between the two automata. This observation was also made in [ @DrosteR07 ] for discounted - kernel automata. For deterministic automata, a synchronized product can be used for $ { \mathsf{Sup}}$ and $ { \mathsf{LimSup}}$, while for $ { \mathsf{LimInf | fijite words are closed unaer sum. The cosj us $O(n_1\cvot n_2)$ fkr ${{\sf Lart}}$- and ${\mathsf{Sum}}$-automata, and $O(b_1\cdot m_1 \cdot n_2 \cdot m_2)$ for ${\mxthsf{Sup}}$-altomata.
It is tasy to see that vge syncmxonizsf prmvuct of two ${{\sf Kast}}$-automada (resp. ${\mathsf{Vuo}}$-abtomata) defines the sum of their landuages of the weight of a jpynt fgaksition is defined as the sum of the wepghts of the corrrsponding transitions in tje tao ${{\sf Last}}$-automata (resp. ${\mathsd{Sum}}$-wytomata).
The cunstruction for ${\mathsf{Aup}}$-automata is the same as for ${\oathsy{Sup}}$-automatq iveg infinite wirds diven in the proof ox \[theo:msx-closed-under-smm\].
[|l|\*[4]{}[c|]{}]{} & maz. & min. & comp. & sum\
${\mathvf{Sup}}$ & & & [$\times$]{}&\
${{\sf Lwst}}$ & & & &\
Ded. ${\jathsf{Sum}}$ & [$\times$]{}& & &\
Nindet. ${\matvsf{Sjn}}$ & & [$\tijex$]{}& [$\fimes$]{}&\
[|l|\*[4]{}[f|]{}]{} & jax. & min. & domp. & sum\
[[ Syp ]{}]{}& & & [$\times$]{}&\
[[ Linf ]{}]{}& & & [$\tynes$]{}&\
[[DLsup]{}]{}& & & [$\tijes$]{}&\
[[NLstp]{}]{}& & & &\
[[DLavg]{}]{}& [$\times$]{}& [$\times$]{}& [$\times$]{}& [$\times$]{}\
[[NLavg]{}]{}& & [$\uimes$]{}& [$\times$]{}& [$\times$]{}\
[[DDisc]{}]{}& [$\times$]{}& [$\times$]{}& &\
[[NDisc]{}]{}& & [$\times$]{}& [$\tlmes$]{}&\
Closuwe under $\max$ for infinite words
---------------------------------------
The maximum of two quanvigatnyc lavtuwges defined by nondeterministic automata can ge ontained by an inltial nondeterminixtlc shoice betweev the cso automata. This obsfrvatiog was also madt in [@DtosteR07] for discounted-sum auromata. For dvterninistic automata, c synchronizzd procuct van be used for ${\mathsf{Sbp}}$ and ${\mathsf{LimSkp}}$, while rur ${\mathsf{LimInf | finite words are closed under sum. The $O(n_1\cdot for ${{\sf and ${\mathsf{Sum}}$-automata, and m_2)$ ${\mathsf{Sup}}$-automata. It is to see that synchronized product of two ${{\sf Last}}$-automata ${\mathsf{Sum}}$-automata) defines the sum of their languages if the weight of a joint is defined as the sum of the weights of the corresponding transitions in two Last}}$-automata ${\mathsf{Sum}}$-automata). construction for ${\mathsf{Sup}}$-automata is the same as for ${\mathsf{Sup}}$-automata over infinite words given in the proof \[theo:max-closed-under-sum\]. [|l|\*[4]{}[c|]{}]{} & max. & min. & comp. sum\ ${\mathsf{Sup}}$ & & [$\times$]{}&\ ${{\sf Last}}$ & & &\ ${\mathsf{Sum}}$ & & Nondet. & & [$\times$]{}& [|l|\*[4]{}[c|]{}]{} & max. & min. & comp. & sum\ [[ Sup ]{}]{}& & & [$\times$]{}&\ [[ Linf & & & & [[NLsup]{}]{}& & [[DLavg]{}]{}& [$\times$]{}& [$\times$]{}& [[NLavg]{}]{}& & [$\times$]{}& [$\times$]{}& [$\times$]{}\ [[DDisc]{}]{}& [[NDisc]{}]{}& & [$\times$]{}& [$\times$]{}&\ Closure under $\max$ for words --------------------------------------- maximum of two quantitative languages defined nondeterministic automata can be obtained by an initial choice between the two automata. This observation was also made in [@DrosteR07] for discounted-sum automata. automata, a synchronized product be used for and while ${\mathsf{LimInf | finite words are closed under Sum. The cost Is $O(n_1\cDot N_2)$ foR ${{\sF LasT}}$- and ${\Mathsf{Sum}}$-automATa, anD $O(n_1\cdot m_1 \cdot n_2 \cdot m_2)$ for ${\mAthsf{suP}}$-AutoMAtA.
It is Easy to sEE tHAT thE sYnChrOnIZeD prodUct Of two ${{\sf last}}$-automaTa (rEsP. ${\mathsf{Sum}}$-auTOmAta) defines The Sum of their laNguAges if ThE weIGht of A joInt trAnsitiON is defIned as the SuM Of the wEIghts of THE cOrreSponding transitioNS iN The two ${{\sf Last}}$-auTomata (ReSP. ${\mATHsf{sum}}$-Automata).
ThE cOnstrUCtion foR ${\MaTHSF{SuP}}$-Automata is the Same as for ${\maTHsf{sup}}$-autOmAta OVer infInite WoRDs gIven in the prOof oF \[theo:max-cLosed-uNDer-sum\].
[|l|\*[4]{}[C|]{}]{} & Max. & min. & cOmp. & sum\
${\MatHsf{sup}}$ & & & [$\tIMeS$]{}&\
${{\sF LaSt}}$ & & & &\
dEt. ${\mAThSf{SUM}}$ & [$\tiMes$]{}& & &\
NondeT. ${\mAtHsf{SuM}}$ & & [$\timES$]{}& [$\TIMes$]{}&\
[|l|\*[4]{}[C|]{}]{} & maX. & min. & Comp. & sUm\
[[ Sup ]{}]{}& & & [$\times$]{}&\
[[ LinF ]{}]{}& & & [$\tiMes$]{}&\
[[DlSup]{}]{}& & & [$\Times$]{}&\
[[nLsup]{}]{}& & & &\
[[dLavG]{}]{}& [$\tImes$]{}& [$\tImes$]{}& [$\tiMes$]{}& [$\tiMeS$]{}\
[[NLavg]{}]{}& & [$\times$]{}& [$\timeS$]{}& [$\timEs$]{}\
[[DDisc]{}]{}& [$\tiMes$]{}& [$\TiMes$]{}& &\
[[nDIsc]{}]{}& & [$\tiMEs$]{}& [$\timeS$]{}&\
ClOsuRe under $\Max$ for iNFinItE WORdS
---------------------------------------
The maximum of two quAnTITaTive langUages dEFiNeD By nondetErMinIstiC AUtomaTa caN Be Obtained By an inITiAl NondeteRmInistiC cHoiCe bEtweeN The tWo autoMata. This ObserVAtion was also maDE in [@DrosteR07] for DIsCOUnTEd-suM auTomata. For deTermINistIc auTOmAta, A SynchRonizEd PRoDUct can be used for ${\mathSf{sup}}$ and ${\MathsF{LimSup}}$, while fOr ${\mathsf{LiMiNF | finite words are closed u nder sum.The c ost is $ O(n_ 1\cd ot n_2)$ for $ { {\sf Last}}$- and ${\maths f{Sum }} $ -aut o ma ta, a nd $O(n _ 1\ c d otm_ 1\cd ot n_ 2 \cd otm_2)$ f or ${\math sf{ Su p}}$-automat a .
It is eas y t o see that t hesynchr on ize d prod uct of t wo ${{ \ sf Las t}}$-auto ma t a (res p . ${\ma t h sf {Sum }}$-automata) def i ne s the sum of th eir la ng u ag e s if th e weight o fa joi n t trans i ti o n isd efined as the sum of the wei ghts o fthe corres pondi ng tra nsitions in the two ${{\ sf Las t }}$-aut o mata (r esp. $ {\m ath sf{S u m} }$ -au to m ata ) .
Th e co nstructi on f or ${ \mat h s f { Sup} }$- auto matais the same a s f or $ { \ma thsf{ Sup}} $-au to mataover i nfini te words given in the proof of \ [t heo :m ax-cl o sed-un der -su m\].
[ |l|\*[4 ] {}[ c| ] { } ]{ } & max. & min. &co m p .& sum\
$ {\math s f{ Su p }}$ & &&[$\ time s $ ]{}&\
${{ \ sf Last}}$ & & & &\
D et. ${\ ma thsf{S um }}$ &[$\ti m es$] {}& &&\
Nonde t. ${ \ mathsf{Sum}}$& & [$\times$] { }& [ $\ t imes $]{ }&\
[|l|\* [4]{ } [c|] {}]{ } & ma x . & m in. & c o mp . & sum\
[[ Sup ]{}] {} & & &[$\ti mes$]{}&\
[[Linf ]{}]{ } & & & [$\t imes $ ]{ } &\
[[DLsup]{}] {}& & & [$\time s $]{}&\
[ [NLsu p]{}]{}& & & &\
[ [ D Lavg]{}] {}& [$ \ti mes $ ] {} & [$\times$]{ } & [$\ ti mes$]{} & [ $\times $]{ }\[[N Lav g] {}]{}& &[$\times $] {} &[$ \ti mes$] { }& [$\ti me s$] {} \
[ [DDis c ]{}]{} & [$\ time s$ ]{ } & [ $\times $ ]{ } & &\[[ ND isc] {}] {} & & [ $\ti m es$ ]{}& [$ \times$]{ }&\
Clo su re under$\max$ for in fi nite words
- --- ------ - - -------- -------------------
Th e maximu m o f two qua ntitative la nguage s d e finedby non deter mi nis t i c aut o m at a c an be obtain e d by an i ni tial nondet erministic choiceb etw een the two a uto mata . Th iso bs e rva ti o n w a s also made in [ @DrosteR07 ]f or discounte d -su mautomat a. Fordeter m inistic automata , a synch ro nize d pro duct can b e used f or ${\mat h sf{Su p }} $ and ${ \maths f{ Lim Sup}} $, whi l e f or ${ \maths f{ LimInf | finite_words are_closed under sum. The_cost is_$O(n_1\cdot_n_2)$ for_${{\sf_Last}}$- and ${\mathsf{Sum}}$-automata,_and $O(n_1\cdot m_1_\cdot n_2 \cdot m_2)$_for ${\mathsf{Sup}}$-automata.
It is_easy_to see that the synchronized product of two ${{\sf Last}}$-automata (resp. ${\mathsf{Sum}}$-automata) defines the_sum_of their_languages_if_the weight of a joint_transition is defined as the_sum of_the weights of the corresponding transitions in the_two_${{\sf Last}}$-automata (resp._${\mathsf{Sum}}$-automata).
The construction for ${\mathsf{Sup}}$-automata is the same as for_${\mathsf{Sup}}$-automata over infinite words given in_the proof of_ \[theo:max-closed-under-sum\].
[|l|\*[4]{}[c|]{}]{}_&_max. & min. &_comp. & sum\
${\mathsf{Sup}}$ & & &_[$\times$]{}&\
${{\sf Last}}$ & & & &\
Det._${\mathsf{Sum}}$ & [$\times$]{}& & &\
Nondet. ${\mathsf{Sum}}$ &_& [$\times$]{}& [$\times$]{}&\
[|l|\*[4]{}[c|]{}]{} & max. &_min. & comp. & sum\
[[_Sup ]{}]{}&_& & [$\times$]{}&\
[[ Linf ]{}]{}&_& & [$\times$]{}&\
[[DLsup]{}]{}&_& &_[$\times$]{}&\
[[NLsup]{}]{}& & &_&\
[[DLavg]{}]{}& [$\times$]{}& [$\times$]{}& [$\times$]{}& [$\times$]{}\
[[NLavg]{}]{}& &_[$\times$]{}& [$\times$]{}& [$\times$]{}\
[[DDisc]{}]{}&_[$\times$]{}& [$\times$]{}& &\
[[NDisc]{}]{}& & [$\times$]{}& [$\times$]{}&\
Closure_under_$\max$ for infinite_words
---------------------------------------
The_maximum_of two_quantitative languages defined_by_nondeterministic automata_can_be obtained by an initial nondeterministic_choice_between the two automata. This observation was_also made in [@DrosteR07] for_discounted-sum_automata. For deterministic automata,_a synchronized product can be_used for ${\mathsf{Sup}}$ and ${\mathsf{LimSup}}$, while_for ${\mathsf{LimInf |
uler}$$ where $y'(t)\equiv \frac{d}{dt}y(t)$. The function (\[defi0\]) describes a real growth process.
As far as there is no universal pricing model, the individual evaluations by the market participants differ and the price deviates from the fundamental average. These different evaluations which are changing in time lead to some kind of spontaneous oscillations. As far as each market has another scale it is useful to normalise the market index (price) to make different markets better comparable: $$I(t) \rightarrow i(t)=\frac{I(t)}{I_0(t)}
\label{itrans}$$ The function (\[itrans\]) performs a normalisation which will project all indices of real markets to a unitarian index $i$ with a constant basic trend: $i_0(t) \equiv 1$ and $\lambda=0$. This way the development of markets can be compared in a single scheme.
For further discussions it is necessary to define the market structure. A market is the totality of all market members participating in the trading process [@caldarelli]. The total amount of market members on normalised markets (\[itrans\]) is constant. The normalised DAX can be found in (fig. 2.).
As already mentioned above, the subjective evaluations of the market status differ from each other [@davidson]. The market participants can be separated into three groups: optimists, pessimists and neutral market participants. Each group has a certain concentration which evolves in time $c_k(t)$. Based on the normalisation there is: $$c_o(t)+c_p(t)+c_n(t)=1,
\label{marktnorm}$$ with $c_o(t)$, $c_p(t)$ and $c_n(t)$ as the corresponding concentrations [^2]
The dynamics of the market is a result of the development of the $c_k(t)$ and the index $i(t)$. Each market group has certain features and react on market changes in a different way:
- [**Optimists**]{} consider the market to be priced low. They want to buy.
- [**Pessimists**]{} consider the market to be priced high. They want to sell.
- [**Neutral market members**]{} consider the market to be priced fair. They are passive.
The groups have different sizes | uler}$$ where $ y'(t)\equiv \frac{d}{dt}y(t)$. The function (\[defi0\ ]) describes a real growth procedure.
equally far as there is no universal pricing exemplar, the individual evaluation by the market participants disagree and the price deviates from the fundamental average. These unlike evaluation which are changing in prison term lead to some kind of spontaneous oscillation. equally far as each market has another plate it is useful to normalize the grocery store index (price) to make unlike markets better comparable: $ $ I(t) \rightarrow i(t)=\frac{I(t)}{I_0(t) }
\label{itrans}$$ The affair (\[itrans\ ]) performs a standardization which will project all index of real markets to a unitarian index $ i$ with a changeless basic trend: $ i_0(t) \equiv 1 $ and $ \lambda=0$. This way the development of markets can be compared in a single scheme.
For further discussions it is necessary to define the market structure. A market is the totality of all market members participating in the trading process [ @caldarelli ]. The entire amount of market extremity on normalize markets (\[itrans\ ]) is constant. The normalize DAX can be found in (fig. 2 .).
As already mentioned above, the subjective evaluation of the market status differ from each other [ @davidson ]. The market participants can be distinguish into three groups: optimists, pessimists and neutral market participants. Each group induce a certain concentration which evolves in time $ c_k(t)$. Based on the standardization there is: $ $ c_o(t)+c_p(t)+c_n(t)=1,
\label{marktnorm}$$ with $ c_o(t)$, $ c_p(t)$ and $ c_n(t)$ as the corresponding concentrations [ ^2 ]
The dynamics of the market is a result of the development of the $ c_k(t)$ and the index $ i(t)$. Each market group have sealed features and react on market changes in a different way:
- [ * * Optimists * * ] { } see the market to be price low. They desire to buy.
- [ * * Pessimists * * ] { } consider the marketplace to be priced gamey. They want to sell.
- [ * * Neutral market members * * ] { } consider the market to be priced fair. They are passive.
The group have different size | uleg}$$ where $y'(t)\equiv \frac{d}{dt}y(u)$. The function (\[deyu0\]) desccibes a real gruwth process.
As far as there ms ni unicersal pricing model, tfe indivifual evaouatmons by the markxf partignpanta difyec and the price deviates xrom the fundakevtcl average. These different evaluatiogs whicn wre changing ig tikq lezd to some kind of spontaneous oscjllatiois. As far as eavh market has another scalf it is useful to normwlise the mqrkee index (price) to make dpyferent mariets better comparable: $$I(t) \rightxrrow i(t)=\frac{I(t)}{I_0(j)}
\lzbfn{itrans}$$ The funcnion (\[itrans\]) pcgforms d normakisation which winl project all indices oh real markets to a onitarian hnbex $i$ with a constant bqsic jrend: $i_0(t) \dwuix 1$ znv $\lzmbda=0$. Hhia way the sevelopment of markets can be voiisred in a sihgle sshqme.
For further discussions it is necessdry to define the market srructure. A market is jhe totaliey of all market members participating in the traging 'rucews [@zqlfarelli]. The total amount of market members on hotmslised markets (\[itrans\]) is vojsywnt. The normauised BZX can be found in (flg. 2.).
As ajreadt mentiontd abpve, the subjective evaluatiins of the mcrkwt status differ fxom each othzr [@davodson]. The market participantr cah be separahed into ffree groups: optioisns, passimists and neutral markqt particmpantx. Each erouk has a certain clncenbsation which evolvfs in thme $c_k(t)$. Baded on the normalisation there mx: $$c_o(t)+c_p(t)+c_n(t)=1,
\ldben{marktnoxm}$$ witm $c_o(t)$, $c_p(t)$ and $c_g(t)$ as the corrgsponding concdntrations [^2]
The dyiamics of thq market is a tesult of the developient of rhe $c_k(t)$ xnd the index $o(t)$. Each mcxket group has certain featmres xhd react on marjet changes in a doffdregt wey:
- [**J[timists**]{} convidef tfr maryet to be pvicdd lpw. They want to buy.
- [**Peasimists**]{} consider yhc market jo be prised high. They want to sell.
- [**Nektral macket mrmbgrs**]{} consider the market to be pdiced faig. Tmey are passide.
Thc grjups have bifferent sizes | uler}$$ where $y'(t)\equiv \frac{d}{dt}y(t)$. The function (\[defi0\]) real process. As as there is individual by the market differ and the deviates from the fundamental average. These evaluations which are changing in time lead to some kind of spontaneous oscillations. far as each market has another scale it is useful to normalise the index to different better comparable: $$I(t) \rightarrow i(t)=\frac{I(t)}{I_0(t)} \label{itrans}$$ The function (\[itrans\]) performs a normalisation which will project all of real markets to a unitarian index $i$ a constant basic trend: \equiv 1$ and $\lambda=0$. This the of markets be in single scheme. For discussions it is necessary to define the market structure. A market is the totality of all market participating in process [@caldarelli]. total of members on normalised is constant. The normalised DAX can (fig. 2.). As already mentioned above, the subjective of the status differ from each other [@davidson]. market participants can be separated into three groups: pessimists and neutral market participants. Each group has a certain concentration which evolves in time on the normalisation there $$c_o(t)+c_p(t)+c_n(t)=1, \label{marktnorm}$$ with $c_p(t)$ $c_n(t)$ the concentrations [^2] dynamics of the market is a result of the development of $c_k(t)$ and the index $i(t)$. Each market group has certain react market changes in different way: - [**Optimists**]{} the to be priced low. to - market be high. They want to - [**Neutral market members**]{} consider market to be priced groups have different sizes | uler}$$ where $y'(t)\equiv \frac{d}{dt}y(t)$. the functioN (\[defi0\]) DesCriBeS a reAl grOwth process.
As fAR as tHere is no universal priciNg modEl, THe inDIvIdual EvaluatIOnS BY thE mArKet PaRTiCipanTs dIffer anD the price dEviAtEs from the funDAmEntal averaGe. THese differenT evAluatiOnS whICh are ChaNging In time LEad to sOme kind of SpONtaneoUS oscillATIoNs. As Far as each market haS AnOTher scale it is uSeful tO nORmALIse The Market indeX (pRice) tO Make difFErENT MarKEts better compArable: $$I(t) \rigHTarRow i(t)=\fRaC{I(t)}{i_0(T)}
\label{ItranS}$$ THE fuNction (\[itranS\]) perForms a norMalisaTIon whicH Will proJect alL inDicEs of REaL mArkEtS To a UNiTarIAn iNdex $i$ witH a CoNstanT basIC TREnd: $i_0(T) \eqUiv 1$ aNd $\lamBda=0$. This way the DevElopMEnt Of marKets cAn be CoMpareD in a siNgle sChEme.
For further diScusSions it is NecEsSarY tO defiNE the maRkeT stRucture. a market IS thE tOTALiTy of all market membeRs PARtIcipatinG in the TRaDiNG process [@CaLdaRellI]. tHe totAl amOUnT of markeT membeRS oN nOrmalisEd MarketS (\[iTraNs\]) iS consTAnt. THe normAlised DAx can bE Found in (fig. 2.).
As alREady mentioned ABoVE, ThE SubjEctIve evaluatiOns oF The mArkeT StAtuS DiffeR from EaCH oTHer [@davidson]. The markeT pArticiPants Can be separateD into three GROUps: optimIsts, PEsSImists and neutrAl marKet particiPAnts. Each Group Has a certAin concenTRAtion whiCh eVolVes In tIME $c_K(t)$. Based on the nORMaliSaTion theRe iS: $$c_o(t)+c_p(t)+C_n(t)=1,
\LabEl{mArkTnOrm}$$ with $c_o(T)$, $c_p(t)$ and $c_N(t)$ As ThE cOrrEsponDIng conceNtRatIoNs [^2]
THe dynAMics of The maRket Is A rESulT of the dEVeLOPmenT oF tHe $c_k(T)$ anD tHe indEx $i(t)$. eAch Market gRoup has ceRtaIN feaTuReS and reaCt on market chaNgEs in a diffeReNt wAy:
- [**OptiMISts**]{} consiDer the market to be priced lOW. They waNt tO buy.
- [**PEssiMists**]{} consIdeR the maRkeT To be prIced hiGh. TheY wAnt TO Sell.
- [**NEUTrAl mArKet members**]{} CONsiDer thE mArkeT to be prIced fair. They are pasSIve.
the groups have DifFereNT SiZes | uler}$$ where $y'(t)\equiv \frac{d}{ dt}y( t)$ . T he fun ctio n (\[defi0\])d escr ibes a real growth pro cess.
A s fa r a s the re is n o u n i ver sa lpri ci n gmodel , t he indi vidual eva lua ti ons by the m a rk et partici pan ts differ an d t he pri ce de v iates fr om th e fund a mental average. T h ese di f ferente v al uati ons which are cha n gi n g in time lead to so me ki n d of sp ontaneousos cilla t ions. A s f a r ase ach market ha s another s c ale it is u sef u l to n ormal is e th e market in dex(price) t o make differe n t marke ts bet ter co mpar a bl e: $$ I( t ) \ r ig hta r row i(t)=\f ra c{ I(t)} {I_0 ( t ) }
\la bel {itr ans}$ $ The functio n ( \[it r ans \]) p erfor ms a n ormal isatio n whi ch will project a ll i ndices of re al ma rk ets t o a uni tar ian index$i$ wit h aco n s t an t basic trend: $i_ 0( t ) \ equiv 1$ and $ \ la mb d a=0$. Th is wa y th e devel opme n tof marke ts can be c ompared i n a si ng lesch eme.Forfurthe r discus sions it is necessar y to define th e m a r ke t str uct ure. A mark et i s the tot a li tyo f all mark et me m bers participatingin the t radin g process [@c aldarelli] . T he total amo u nt of market memb ers o n normalis e d market s (\[ itrans\] ) is cons t a nt. Thenor mal ise d D A X c an be found i n (fig .2.).
A s a lreadymen tio ned ab ov e, the su bjective e va lu at ion s oft he marke tsta tu s d iffer from e ach o ther [ @d a vid son]. T h em a rket p ar tici pan ts canbe s e par ated in to threegro u ps:op ti mists,pessimists an dneutral ma rk etpartic i p ants. Ea ch group has a certainc oncentr ati on wh ichevolves i n t ime $c _k( t )$. Ba sed on theno rma l i satio n th ere i s: $$c_o(t ) + c_p (t)+c _n (t)= 1,
\l abel{marktnorm}$$w ith $c_o(t)$, $c _p( t)$a n d$c_ n (t ) $ a st hec o rresponding con centration s[ ^2 ]
The dyn a mic sof themarketis ar esult o f the dev elopmentof the $ c_k (t)$ and t he index $i(t)$.E ach m a rk et gr oup has c er tai n fea turesa ndreact on ma rk et cha ngesin a diffe rent way:
- [**Optim ists** ]{} c ons ider themar k etto be pri cedlow. Theywan t t o buy .
- [* *Pes s im ist s **]{} con s ider them ar ket t obe priced h i g h . T hey w ant to sel l.
- [**Neutral ma r ket members**] {} c o n sid ert he m ar ket to be pric edfa i r . They a re passive.
The grou ps havediffer ent si zes | uler}$$ where_$y'(t)\equiv \frac{d}{dt}y(t)$._The function (\[defi0\]) describes_a real_growth_process.
As far_as_there is no_universal pricing model,_the individual evaluations by_the market participants_differ_and the price deviates from the fundamental average. These different evaluations which are changing_in_time lead_to_some_kind of spontaneous oscillations. As_far as each market has_another scale_it is useful to normalise the market index_(price)_to make different_markets better comparable: $$I(t) \rightarrow i(t)=\frac{I(t)}{I_0(t)}
\label{itrans}$$ The function (\[itrans\])_performs a normalisation which will project_all indices of_real_markets_to a unitarian index_$i$ with a constant basic trend:_$i_0(t) \equiv 1$ and $\lambda=0$. This_way the development of markets can be_compared in a single scheme.
For further_discussions it is necessary to_define the_market structure. A market is_the totality of_all market_members participating in_the trading process [@caldarelli]. The total_amount of market_members on normalised markets (\[itrans\]) is_constant._The normalised DAX_can_be_found in_(fig. 2.).
As already_mentioned_above, the_subjective_evaluations of the market status differ_from_each other [@davidson]. The market participants can_be separated into three_groups:_optimists, pessimists and neutral_market participants. Each group has_a certain concentration which evolves in_time $c_k(t)$._Based on_the normalisation there is: $$c_o(t)+c_p(t)+c_n(t)=1,
\label{marktnorm}$$ with $c_o(t)$, $c_p(t)$ and_$c_n(t)$ as the corresponding concentrations [^2]
The_dynamics of the market_is a_result_of the development_of_the $c_k(t)$_and the index $i(t)$. Each market group_has certain_features and react on market changes_in a different way:
-__ [**Optimists**]{} consider the market to_be priced low. They want to_buy.
- [**Pessimists**]{} consider_the_market_to be priced high. They_want to sell.
- [**Neutral_market members**]{} consider_the market to be priced fair. They_are_passive.
The groups have different sizes |
, i.e. subsets that are the union of some minimizing orbits. More precisely, we would like to know if we can say something about the regularity of such subsets (we will be more precise very soon. It’s a kind of differentiability) and particularly if there is a link between the dynamic of the flow restricted to such a set and the regularity of the set.
The oldest result in this direction concerns the time-dependent case : considering a symplectic twist map of the annulus $T^*{\mathbb {S}}$, G. Birkhoff proved in the 1920’s that any essential invariant curve is the graph of a Lipschitz map (see [@Bir1] or [@He1]). It is easy to prove that such a curve is action minimizing. In the case of higher dimensions, M. Herman proved in [@He2] that any $C^0$-Lagrangian graph of $T^*{\mathbb {T}}^n$ that is invariant by a symplectic twist map is, in fact, the graph of a Lipschitz map. A related result in the autonomous case is that any $C^1$-Hamilton-Jacobi solution of a Tonelli Hamiltonian is, in fact, $C^{1,1}$ (see [@Fa2]). As Rademacher’s theorem says to us that any Lipschitz function is differentiable Lebesgue almost everywhere, these results are a kind of regularity result.
In [@Arna2], we did, in fact, improve these results of regularity in the autonomous case, proving that if a $C^0$-Lagrangian graph is invariant by a Tonelli flow, and if one of the two following hypotheses is satisfied:
1. $\dim M=2$ and all the singularities of $H$ are non degenerate;
2. the dynamic of the restriction of the flow to the invariant graph is Lipschitz conjugate to a translations’ flow;
then the invariant graph is, in fact, $C^1$ almost everywhere (this is stronger than just differentiable). Let us point out that any of the two previous hypotheses implies that the dynamic of the restricted flow to the graph is soft on a certain sense (our arguments are not very precise, but we only want to give a certain intuition of the forthcoming result); indeed, when $\dim M=2$, if we reduce the dynamic modulo the vector field, we obtain a 1-dimension dynamic, and it is known at least in the differentiable case that the Lyap | , i.e. subsets that are the union of some minimizing orbits. More precisely, we would wish to acknowledge if we can say something about the regularity of such subsets (we will be more accurate very soon. It ’s a kind of differentiability) and particularly if there be a link between the dynamic of the stream restricted to such a set and the regularity of the stage set.
The oldest result in this commission concerns the time - subject shell : considering a symplectic twist map of the annulus $ T^*{\mathbb { S}}$, G. Birkhoff proved in the 1920 ’s that any essential changeless curve is the graph of a Lipschitz map (see [ @Bir1 ] or [ @He1 ]). It is comfortable to prove that such a curve is action minimizing. In the case of higher dimension, M. Herman proved in [ @He2 ] that any $ C^0$-Lagrangian graph of $ T^*{\mathbb { T}}^n$ that is invariant by a symplectic twist map is, in fact, the graph of a Lipschitz map. A related resultant role in the autonomous case is that any $ C^1$-Hamilton - Jacobi solution of a Tonelli Hamiltonian is, in fact, $ C^{1,1}$ (see [ @Fa2 ]). As Rademacher ’s theorem says to us that any Lipschitz function is differentiable Lebesgue almost everywhere, these results are a kind of regularity result.
In [ @Arna2 ], we did, in fact, improve these results of regularity in the autonomous case, proving that if a $ C^0$-Lagrangian graph is invariant by a Tonelli flow, and if one of the two following hypotheses is satisfied:
1. $ \dim M=2 $ and all the singularities of $ H$ are non debauched;
2. the moral force of the restriction of the flow to the invariant graph is Lipschitz conjugate solution to a transformation ’ flow;
then the invariant graph is, in fact, $ C^1 $ almost everywhere (this is stronger than just differentiable). Let us point out that any of the two former hypotheses implies that the dynamic of the restricted flow to the graph is soft on a certain sense (our arguments are not very precise, but we only want to feed a sealed intuition of the extroverted result); indeed, when $ \dim M=2 $, if we reduce the dynamic modulo the vector field, we receive a 1 - dimension dynamic, and it is known at least in the differentiable subject that the Lyap | , i.e. subsets that are the unlon of some minimizing mrbits. More prdcisely, we would like to knox if we cqn say something about the reguparity od surh subsets (we will be movz predlse vzrb soon. It’s a kikd of diffesentiability) atd pcrticularly if there is a link betweqn the cyjamic of the fjow geftridneb to such a set and the regularjty of uhe set.
The oldest tesult in this direction clncegns the time-dependfnt case : cobsidqeing a sympldctic twisu kap of the annulus $T^*{\mathbb {S}}$, G. Birkhoff pruved nn the 1920’s thqt anj essential mnvariwnt curve is the gra[h of a Lipschitz map (sex [@Bie1] or [@He1]). It is easy to prove that such a curve is aetion minimizing. In tye case of vighde dkmehsmona, M. Herlan proved in [@He2] that ant $C^0$-Lagrangian graph os $T^*{\mathbb {T}}^n$ tgat is igvariant by a symplectic twist map is, it fzct, the graph of a Lipsxhitz map. A related rgsult in tre autonomous case is that any $C^1$-Hamilton-Jacobi sonutioi uf c Toneuoi Hamiltonian is, in fact, $C^{1,1}$ (see [@Fa2]). As Rademachew’a uheprem says to uf that any Kiosvritz function is diyredentiable Lebesgue almost evertwhere, thtse rrsults are a kind of regulaeity result.
Iu [@Aena2], we did, in fact, improve thzse rexults of regularity in the abtonomkus case, prlving thaf if a $C^0$-Lagrangiav ggaph is invariant by a Tonelli flow, and if pne of ghe jwo foljowing hyplthescv is satisfied:
1. $\dil M=2$ aud aln the singklarities of $H$ are non degeneravx;
2. the dynamiv mf nhe restrnction of the flow eo the invariaut graph is Lkpschitz cknjugatx to a transjations’ flow;
tvgn the invarient graph is, un fqct, $C^1$ auoost everywherr (this is stronger rhan just differenbiablg). Met us point ouc uhqt any of the ywo prqvpoux rfpotheses im[lier tfst thd dynamic on tfe rrstricted flow to tha grzph is soft on a crrbain sensg (our argtments are noy very precise, but we oily wait to bivg a certain intuition of the fodthcoming refult); indeed, rhen $\dim M=2$, if we reduce the dynamic modulo the vector fmeld, we obtain a 1-dimensuon dynamic, and it ns known at lxast ig the difxerentiable case thar the Lyap | , i.e. subsets that are the union minimizing More precisely, would like to something the regularity of subsets (we will more precise very soon. It’s a of differentiability) and particularly if there is a link between the dynamic of flow restricted to such a set and the regularity of the set. The result this concerns time-dependent case : considering a symplectic twist map of the annulus $T^*{\mathbb {S}}$, G. Birkhoff proved the 1920’s that any essential invariant curve is graph of a Lipschitz (see [@Bir1] or [@He1]). It easy prove that a is minimizing. In the of higher dimensions, M. Herman proved in [@He2] that any $C^0$-Lagrangian graph of $T^*{\mathbb {T}}^n$ that is by a map is, fact, graph a Lipschitz map. result in the autonomous case is solution of a Tonelli Hamiltonian is, in fact, (see [@Fa2]). Rademacher’s theorem says to us that Lipschitz function is differentiable Lebesgue almost everywhere, these are a kind of regularity result. In [@Arna2], we did, in fact, improve these results in the autonomous case, that if a graph invariant a flow, and one of the two following hypotheses is satisfied: 1. $\dim M=2$ all the singularities of $H$ are non degenerate; 2. the the of the flow the invariant graph is conjugate a translations’ flow; then graph in everywhere is than just differentiable). Let point out that any of two previous hypotheses implies restricted flow to the graph is soft on certain sense (our arguments are not very but we only want to give a certain intuition of the forthcoming indeed, when if we reduce the dynamic modulo the vector we obtain a 1-dimension and it is known at least in the differentiable that Lyap | , i.e. subsets that are the union oF some minimIzing OrbIts. moRe prEcisEly, we would like TO knoW if we can say something abOut thE rEGulaRItY of suCh subseTS (wE WIll Be MoRe pReCIsE very SooN. It’s a kiNd of differEntIaBility) and parTIcUlarly if thEre Is a link betweEn tHe dynaMiC of THe floW reStricTed to sUCh a set And the regUlARity of THe set.
ThE OLdEst rEsult in this directIOn COncerns the time-DependEnT CaSE : ConSidEring a sympLeCtic tWIst map oF ThE ANNulUS $T^*{\mathbb {S}}$, G. BirKhoff proved IN thE 1920’s that AnY esSEntial InvarIaNT cuRve is the graPh of A LipschitZ map (seE [@bir1] or [@He1]). iT is easy To provE thAt sUch a CUrVe Is aCtIOn mINiMizINg. IN the case Of HiGher dImenSIONS, M. HeRmaN proVed in [@he2] that any $C^0$-LagRanGian GRapH of $T^*{\mAthbb {t}}^n$ thAt Is invAriant By a syMpLectic twist map iS, in fAct, the graPh oF a lipScHitz mAP. A relaTed ResUlt in thE autonoMOus CaSE IS tHat any $C^1$-Hamilton-JacObI SOlUtion of a tonellI haMiLTonian is, In FacT, $C^{1,1}$ (seE [@fA2]). As RaDemaCHeR’s theoreM says tO Us ThAt any LiPsChitz fUnCtiOn iS diffERentIable LEbesgue aLmost EVerywhere, these REsults are a kinD Of REGuLAritY reSult.
In [@Arna2], wE did, IN facT, impROvE thESe resUlts oF rEGuLArity in the autonomouS cAse, proVing tHat if a $C^0$-LagranGian graph iS INVariant bY a ToNElLI flow, and if one oF the tWo followinG HypothesEs is sAtisfied:
1. $\Dim M=2$ and alL THe singulAriTieS of $h$ arE NOn Degenerate;
2. the DYNamiC oF the resTriCtion of The FloW to The InVariant grAph is LipScHiTz CoNjuGate tO A translaTiOns’ FlOw;
tHen thE InvariAnt grAph iS, iN fACt, $C^1$ Almost eVErYWHere (ThIs Is stRonGeR than Just DIffErentiaBle). Let us pOinT Out tHaT aNy of the Two previous hyPoTheses implIeS thAt the dYNAmic of thE restricted flow to the graPH is soft On a CertaIn seNse (our argUmeNts are Not VEry preCise, buT we onLy WanT TO give A CErTaiN iNtuition of THE foRthcoMiNg reSult); indEed, when $\dim M=2$, if we redUCe tHe dynamic moduLo tHe veCTOr FieLD, wE ObtAiN A 1-diMENsion dynamic, and It is known aT lEAsT in the diffERenTiAble casE that thE Lyap | , i.e. subsets that are th e union of some mi nim iz ingorbi ts. More preci s ely, we would like to know if w ec an s a ysomet hing ab o ut t here gu lar it y o f suc h s ubsets(we will b e m or e precise ve r ysoon. It’s akind of diff ere ntiabi li ty) and p art icula rly if thereis a link b e tweent he dyna m i cof t he flow restricte d t o such a set an d there g ul a r ity of the set.
T he ol d est res u lt i n th i s direction c oncerns the tim e-depe nd ent case : cons id e rin g a symplec tictwist map of th e annulu s $T^*{\ mathbb {S }}$ , G. Bi rk hof fp rov e dint he1920’s t ha tany e ssen t i a l inv ari antcurve is the graph of a L i psc hitzmap ( see[@ Bir1] or [@ He1]) .It is easy to p rove that suc h a c urv eis ac t ion mi nim izi ng. Inthe cas e of h i g h er dimensions, M. He rm a n p roved in [@He2 ] t ha t any $C^ 0$ -La gran g i an gr apho f$T^*{\ma thbb { T }} ^n $ thatis invar ia ntbya sym p lect ic twi st map i s, in fact, the grap h of a Lipschi t zm a p. A re lat ed result i n th e aut onom o us ca s e isthatan y $ C ^1$-Hamilton-Jacobi s olutio n ofa Tonelli Ham iltonian i s , in fact, $C^ { 1, 1 }$ (see [@Fa2] ). As Rademache r ’s theor em sa ys to us that any L ipschitz fu nct ion is d if ferentiable L e b esgu ealmosteve rywhere , t hes e r esu lt s are a k ind of r eg ul ar it y r esult .
In [@A rn a2] ,wedid,i n fact , imp rove t he s e r esultso fr e gula ri ty inthe a utono mous cas e, prov ing thatifa $C^ 0$ -L agrangi an graph is i nv ariant byaTon elli f l o w, and i f one of the two follow i ng hypo the ses i s sa tisfied:
1. $\di m M = 2$ and all t he si ng ula r i tieso f $ H$ar e non dege n e rat e;
2 . the dynami c of the restricti o n o f the flow to th e in v a ri ant gr a phis Lip s c hitz conjugateto a trans la t io ns’ flow;the nthe inv ariantgraph is, infact, $C^ 1$ almost e very w h ere (this isstronger than jus t diff e re ntiab le) . Letus po int o ut tha t an y ofthe tw oprevio us hy po theses i mplies that the dynamic of th e res tri cted flow to the graph is sof t on a cer tai n s ense(ou r argu ment s a ren ot ve ry p r ecise, bu t w e o n l ywant to giv e a ce rtain in t uition ofthe forthcoming r e sult); indeed, whe n $\d imM =2$, i f we reduce th e d yn a m ic modul othe vectorfield, w eo btain a 1-d imensi on dyna m i c, and it iskno wn at lea stin the dif fe re n tiable cas ethat t he Lya p | , i.e._subsets that_are the union of_some minimizing_orbits._More precisely,_we_would like to_know if we_can say something about_the regularity of_such_subsets (we will be more precise very soon. It’s a kind of differentiability) and_particularly_if there_is_a_link between the dynamic of_the flow restricted to such_a set_and the regularity of the set.
The oldest result_in_this direction concerns_the time-dependent case : considering a symplectic twist map of_the annulus $T^*{\mathbb {S}}$, G. Birkhoff proved_in the 1920’s_that_any_essential invariant curve is_the graph of a Lipschitz map_(see [@Bir1] or [@He1]). It is_easy to prove that such a curve_is action minimizing. In the case_of higher dimensions, M. Herman proved_in [@He2]_that any $C^0$-Lagrangian graph of_$T^*{\mathbb {T}}^n$ that_is invariant_by a symplectic_twist map is, in fact, the_graph of a_Lipschitz map. A related result in_the_autonomous case is_that_any_$C^1$-Hamilton-Jacobi solution_of a Tonelli_Hamiltonian_is, in_fact,_$C^{1,1}$ (see [@Fa2]). As Rademacher’s theorem_says_to us that any Lipschitz function is_differentiable Lebesgue almost everywhere,_these_results are a kind_of regularity result.
In [@Arna2], we_did, in fact, improve these results_of regularity_in the_autonomous case, proving that if a $C^0$-Lagrangian graph is invariant by_a Tonelli flow, and if one_of the two following_hypotheses is_satisfied:
1._ $\dim M=2$_and_all the_singularities of $H$ are non degenerate;
2. _the dynamic_of the restriction of the flow_to the invariant graph_is_Lipschitz conjugate to a translations’ flow;
then_the invariant graph is, in fact,_$C^1$ almost everywhere (this is_stronger_than_just differentiable). Let us point_out that any of the two_previous hypotheses implies_that the dynamic of the restricted flow_to_the graph is soft on a_certain_sense (our arguments are not very_precise,_but_we only want to give_a certain intuition of the forthcoming_result); indeed, when $\dim M=2$, if we reduce the_dynamic modulo the_vector field, we obtain a_1-dimension_dynamic,_and it is known at least in the differentiable case_that the_Lyap |
each $h \in \mathcal{C}(\mathopen{[}0,R\mathclose{]})$, $Q_\varphi(h) = Q_\varphi(h(\odot)) \in \mathbb{R}$ is defined as the unique solution of the equation $$\int_0^R \varphi^{-1}\left( -h(r) + Q_\varphi(h)\right)\,\mathrm{d}r = 0.$$ It is worth noticing that, compared with the Neumann case, this formulation is slightly less transparent: indeed, due to the non-locality of the periodic boundary conditions, the additional term $Q_\varphi$ appears. An alternative fixed point formulation for, relying on a direct use of coincidence degree theory for the equivalent planar system $u' = \varphi^{-1}(v)$, $v' = - f(r,u)$, has been recently proposed in [@FeZa-17tmna]. $\hfill\lhd$
Two degree lemmas {#section-2.2}
-----------------
Taking advantage of the abstract setting just presented, we now prove two lemmas for the computation of the degree on open balls $B(0,d)$ (with center $0$ and radius $d>0$) of the Banach space $\mathcal{C}(\mathopen{[}0,R\mathclose{]})$ in the framework of the Neumann problem $$\label{eq-phi-ag}
\begin{cases}
\, \bigl{(} r^{N-1}\varphi(u')\bigr{)}' + \lambda r^{N-1} a(r)g(u) = 0, \\
\, u'(0) = u'(R) = 0,
\end{cases}$$ where $a \colon \mathopen{[}0,R\mathclose{]} \to \mathbb{R}$ is an $L^1$-function, $g \colon \mathopen{[}0,+\infty\mathclose{[} \to \mathopen{[}0,+\infty\mathclose{[}$ is a continuous function satisfying $g(0) = 0$, and $\lambda>0$. Notice that, to enter the setting of the previous section, we need for the nonlinearity appearing in the equation to be defined for every $u \in \mathbb{R}$; accordingly, we set $$f(r,u) = \begin{cases}
\, \lambda a(r) g | each $ h \in \mathcal{C}(\mathopen{[}0,R\mathclose{]})$, $ Q_\varphi(h) = Q_\varphi(h(\odot) ) \in \mathbb{R}$ is defined as the unique solution of the equation $ $ \int_0^R \varphi^{-1}\left (-h(r) + Q_\varphi(h)\right)\,\mathrm{d}r = 0.$$ It is deserving comment that, compared with the Neumann case, this conceptualization is slightly less diaphanous: indeed, due to the non - locality of the periodic boundary circumstance, the additional term $ Q_\varphi$ look. An alternative fixed compass point formulation for, relying on a lineal use of coincidence degree theory for the equivalent planar system $ u' = \varphi^{-1}(v)$, $ v' = - f(r, u)$, has been recently propose in [ @FeZa-17tmna ]. $ \hfill\lhd$
Two degree lemmas { # section-2.2 }
-----------------
Taking advantage of the abstract place just presented, we now prove two lemma for the computation of the degree on open balls $ B(0,d)$ (with center $ 0 $ and radius $ d>0 $) of the Banach quad $ \mathcal{C}(\mathopen{[}0,R\mathclose{]})$ in the framework of the Neumann problem $ $ \label{eq - phi - ag }
\begin{cases }
\, \bigl { (} r^{N-1}\varphi(u')\bigr {) }' + \lambda r^{N-1 } a(r)g(u) = 0, \\
\, u'(0) = u'(R) = 0,
\end{cases}$$ where $ a \colon \mathopen{[}0,R\mathclose { ] } \to \mathbb{R}$ is an $ L^1$-function, $ g \colon \mathopen{[}0,+\infty\mathclose { [ } \to \mathopen{[}0,+\infty\mathclose{[}$ is a continuous function satisfying $ g(0) = 0 $, and $ \lambda>0$. Notice that, to enter the setting of the previous section, we need for the nonlinearity appearing in the equation to be defined for every $ u \in \mathbb{R}$; accordingly, we set $ $ f(r, u) = \begin{cases }
\, \lambda a(r) g | eafh $h \in \mathcal{C}(\mathopen{[}0,V\mathclose{]})$, $Q_\varpku(h) = Q_\verphi(h(\osot)) \in \mxthbb{R}$ is defined as the unieuw solytion of the equation $$\knt_0^R \varpji^{-1}\left( -h(e) + Q_\tarphi(h)\right)\,\mathcj{d}r = 0.$$ Ib is slrth ioticing that, cpmpared widh the Neumann cxsz, this formulation is slightly less eranspatejt: indeed, due jo tht njn-lodality of the periodic boundary cohditionv, the additiomal term $Q_\varphi$ appears. Aj alhernative fixed polnt formulajjon dor, relying un a direcu bse of coindidence degree theory for the eduivakent planat aydjem $u' = \varphm^{-1}(v)$, $v' = - f(r,u)$, has beek recendly prolosed in [@FeZa-17tkna]. $\hfull\lhd$
Two degree lemmes {#section-2.2}
-----------------
Taking advwntage of tke abstract setting jysr preventad, wd nod pdote fwo lelmaa for the domputation of the degree on oken valls $B(0,d)$ (with centew $0$ and radius $d>0$) of the Banach space $\mathcdl{C}(\jathopen{[}0,R\mathclose{]})$ in tye framework of the Ngumann proflem $$\label{eq-phi-ag}
\begin{cases}
\, \bigl{(} r^{N-1}\varphi(u')\bigr{)}' + \ldmbda f^{N-1} c(v)n(u) = 0, \\
\, u'(0) = u'(R) = 0,
\end{cases}$$ where $a \colon \mathopen{[}0,R\mathclofs{]} \uo \kathbb{R}$ is an $J^1$-function, $g \cllpg \mathopen{[}0,+\infjy\mathcloss{[} \to \mathopen{[}0,+\infty\lathclofe{[}$ is a contintous function satisfying $g(0) = 0$, ane $\lambda>0$. Notpce rhat, to enter the detting of che prgvious section, we need for thz nonljnearity apoearing ih the equation to be dafined for every $u \in \mathfb{R}$; accorvinglv, we set $$f(r,u) = \begig{cases}
\, \lamhda a(v) g | each $h \in \mathcal{C}(\mathopen{[}0,R\mathclose{]})$, $Q_\varphi(h) = Q_\varphi(h(\odot)) is as the solution of the Q_\varphi(h)\right)\,\mathrm{d}r 0.$$ It is noticing that, compared the Neumann case, this formulation is less transparent: indeed, due to the non-locality of the periodic boundary conditions, the term $Q_\varphi$ appears. An alternative fixed point formulation for, relying on a direct of degree for equivalent planar system $u' = \varphi^{-1}(v)$, $v' = - f(r,u)$, has been recently proposed in [@FeZa-17tmna]. Two degree lemmas {#section-2.2} ----------------- Taking advantage of abstract setting just presented, now prove two lemmas for computation the degree open $B(0,d)$ center $0$ and $d>0$) of the Banach space $\mathcal{C}(\mathopen{[}0,R\mathclose{]})$ in the framework of the Neumann problem $$\label{eq-phi-ag} \begin{cases} \, \bigl{(} + \lambda = 0, \, = = 0, \end{cases}$$ \colon \mathopen{[}0,R\mathclose{]} \to \mathbb{R}$ is an \mathopen{[}0,+\infty\mathclose{[} \to \mathopen{[}0,+\infty\mathclose{[}$ is a continuous function satisfying = 0$, $\lambda>0$. Notice that, to enter the of the previous section, we need for the appearing in the equation to be defined for every $u \in \mathbb{R}$; accordingly, we set \begin{cases} \, \lambda a(r) | each $h \in \mathcal{C}(\mathopen{[}0,R\mAthclose{]})$, $Q_\vArphi(H) = Q_\vArpHi(H(\odoT)) \in \mAthbb{R}$ is defineD As thE unique solution of the eqUatioN $$\iNT_0^R \vaRPhI^{-1}\left( -H(r) + Q_\varpHI(h)\RIGht)\,\MaThRm{d}R = 0.$$ IT Is Worth NotIcing thAt, compared WitH tHe Neumann casE, ThIs formulatIon Is slightly leSs tRanspaReNt: iNDeed, dUe tO the nOn-locaLIty of tHe periodiC bOUndary COnditioNS, ThE addItional term $Q_\varphI$ ApPEars. An alternatIve fixEd POiNT ForMulAtion for, reLyIng on A Direct uSE oF COIncIDence degree thEory for the eQUivAlent pLaNar SYstem $u' = \VarphI^{-1}(v)$, $V' = - F(r,u)$, Has been receNtly Proposed iN [@FeZa-17tMNa]. $\hfill\LHd$
Two deGree leMmaS {#seCtioN-2.2}
-----------------
taKiNg aDvANtaGE oF thE AbsTract setTiNg Just pReseNTED, We noW prOve tWo lemMas for the compUtaTion OF thE degrEe on oPen bAlLs $B(0,d)$ (wIth cenTer $0$ anD rAdius $d>0$) of the BanaCh spAce $\mathcaL{C}(\mAtHopEn{[}0,r\mathCLose{]})$ in The FraMework oF the NeuMAnn PrOBLEm $$\Label{eq-phi-ag}
\begin{cAsES}
\, \BiGl{(} r^{N-1}\varpHi(u')\bigR{)}' + \LaMbDA r^{N-1} a(r)g(u) = 0, \\
\, u'(0) = U'(R) = 0,
\End{CaseS}$$ WHere $a \ColoN \MaThopen{[}0,R\mAthcloSE{]} \tO \mAthbb{R}$ iS aN $L^1$-funcTiOn, $g \ColOn \matHOpen{[}0,+\Infty\mAthclose{[} \To \matHOpen{[}0,+\infty\mathcLOse{[}$ is a continuOUs FUNcTIon sAtiSfying $g(0) = 0$, and $\lAmbdA>0$. notiCe thAT, tO enTEr the SettiNg OF tHE previous section, we nEeD for thE nonlInearity appeaRing in the eQUATion to be DefiNEd FOr every $u \in \mathBb{R}$; acCordingly, wE Set $$f(r,u) = \beGin{caSes}
\, \lambdA a(r) g | each $h \in \mathcal{C}(\ mathopen{[ }0,R\ mat hcl os e{]} )$,$Q_\varphi(h)= Q_\ varphi(h(\odot)) \in \ mathb b{ R }$ i s d efine d as th e u n i que s ol uti on of theequ ation $ $\int_0^R\va rp hi^{-1}\left ( - h(r) + Q_\ var phi(h)\right )\, \mathr m{ d}r = 0.$ $ I t isworthn oticin g that, c om p ared w i th theN e um anncase, this formul a ti o n is slightlyless t ra n sp a r ent : i ndeed, due t o the non-loc a li t y oft he periodic b oundary con d iti ons, t he ad d itiona l ter m$ Q_\ varphi$ app ears . An alte rnativ e fixedp oint fo rmulat ion fo r, r e ly in g o na di r ec t u s e o f coinci de nc e deg reet h e o ry f ortheequiv alent planarsys tem$ u'= \va rphi^ {-1} (v )$, $ v' = - f(r, u) $, has been rec entl y propose d i n[@F eZ a-17t m na]. $ \hf ill \lhd$
Two deg r eele m m a s{#section-2.2}
--- -- - - -- --------
Taki n gad v antage o fthe abs t r act s etti n gjust pre sented , w enow pro ve two l em mas fo r the comp utatio n of the degr e e on open ball s $B(0,d)$ (wi t hc e nt e r $0 $ a nd radius $ d>0$ ) oftheB an ach space $\ma th c al { C}(\mathopen{[}0,R\ ma thclos e{]}) $ in the fram ework of t h e Neumannprob l em $$\label{eq-ph i-ag}
\begin{ca s es}
\, \ bigl{ (} r^{N- 1}\varphi ( u ')\bigr{ )}' +\la mbd a r^ {N-1} a(r)g(u ) = 0, \ \
\, u' (0) = u'(R ) = 0,
\e nd{ ca ses}$$ wh ere $a \ co lo n\m ath open{ [ }0,R\mat hc los e{ ]}\to \ m athbb{ R}$ i s an $ L^ 1 $-f unction , $ g \col on \ math ope n{ [}0,+ \inf t y\m athclos e{[} \to\ma t hope n{ [} 0,+\inf ty\mathclose{ [} $ is a con ti nuo us fun c t ion sati sfying $g(0) = 0$, and$ \lambda >0$ . Not icethat, toent er the se t ting o f theprevi ou s s e c tion, w enee dfor the no n l ine arity a ppea ring in the equation to b e de fined for eve ry$u \ i n \ mat h bb { R}$ ;a cco r d ingly, we set $ $f(r,u) =\b e gi n{cases}
\ , \l am bda a(r ) g | each_$h \in_\mathcal{C}(\mathopen{[}0,R\mathclose{]})$, $Q_\varphi(h) = Q_\varphi(h(\odot))_\in \mathbb{R}$_is_defined as_the_unique solution of_the equation $$\int_0^R_\varphi^{-1}\left( -h(r) + Q_\varphi(h)\right)\,\mathrm{d}r_= 0.$$ It_is_worth noticing that, compared with the Neumann case, this formulation is slightly less transparent:_indeed,_due to_the_non-locality_of the periodic boundary conditions,_the additional term $Q_\varphi$ appears._An alternative_fixed point formulation for, relying on a direct_use_of coincidence degree_theory for the equivalent planar system $u' = \varphi^{-1}(v)$,_$v' = - f(r,u)$, has been_recently proposed in_[@FeZa-17tmna]._$\hfill\lhd$
Two_degree lemmas {#section-2.2}
-----------------
Taking advantage_of the abstract setting just presented,_we now prove two lemmas for_the computation of the degree on open_balls $B(0,d)$ (with center $0$ and_radius $d>0$) of the Banach_space $\mathcal{C}(\mathopen{[}0,R\mathclose{]})$_in the framework of the_Neumann problem $$\label{eq-phi-ag}
\begin{cases}
\,_\bigl{(} r^{N-1}\varphi(u')\bigr{)}'_+ \lambda r^{N-1}_a(r)g(u) = 0, \\
\, u'(0) =_u'(R) = 0,
\end{cases}$$_where $a \colon \mathopen{[}0,R\mathclose{]} \to \mathbb{R}$_is_an $L^1$-function, $g_\colon_\mathopen{[}0,+\infty\mathclose{[}_\to \mathopen{[}0,+\infty\mathclose{[}$_is a continuous_function_satisfying $g(0)_=_0$, and $\lambda>0$. Notice that, to_enter_the setting of the previous section, we_need for the nonlinearity_appearing_in the equation to_be defined for every $u_\in \mathbb{R}$; accordingly, we set $$f(r,u)_= \begin{cases}
\,_\lambda a(r)_g |
$T\neq0$ {#couplatt}
--------------------------------
We extend here our studies of the coupling at zero temperature to finite temperatures below and above deconfinement following the conceptual approach given in [@Kaczmarek:2004gv]. In this case the appropriate observable is the color singlet quark anti-quark free energy and its derivative. We use the perturbative short and large distance relation from one gluon exchange [@Nadkarni:1986as; @Nadkarni:1986cz; @McLerran:1981pb], [*i.e.*]{} in the limit $r\Lambda_{\text{QCD}}\ll1$ zero temperature perturbation theory suggests $$\begin{aligned}
F_1(r,T)\;\equiv\;V(r)&\simeq&-\frac{4}{3}\frac{\alpha(r)}{r}\;,\label{alp_rT1}\end{aligned}$$ while high temperature perturbation theory, [*i.e.*]{} $rT\gg1$ and $T$ well above $T_c$, yields $$\begin{aligned}
F_1(r,T)&\simeq&-\frac{4}{3}\frac{\alpha(T)}{r}e^{-m_D(T)r}\;.\label{alp_rT2}\end{aligned}$$ In both relations we have neglected any constant contributions to the free energies which, in particular, at high temperatures will dominate the large distance behavior of the free energies. Moreover, we already anticipated here the running of the couplings with the expected dominant scales $r$ and $T$ in both limits. At finite temperature we define the running coupling in analogy to $T=0$ as (see [@Kaczmarek:2002mc; @Kaczmarek:2004gv]),$$\begin{aligned}
\alpha_{qq}(r,T)&\equiv&\frac{3}{4}r^2 \frac{dF_1(r,T)}{dr}\;.\label{alp_rT}\end{aligned}$$ With this definition any undetermined constant contributions to the free energies are eliminated and the coupling defined here at finite temperature will recover the coupling at zero temperature defined in (\[alp\_qq\]) in the limit of small distances. Therefore $\alpha_{qq}(r,T)$ will show the (zero temperature) weakening in the short distance perturbative regime. In the large distance limit, however, the coupling | $ T\neq0 $ { # couplatt }
--------------------------------
We extend here our studies of the coupling at zero temperature to finite temperatures downstairs and above deconfinement take after the conceptual approach given in [ @Kaczmarek:2004gv ]. In this case the appropriate discernible is the color singlet quark anti - quark free department of energy and its derivative instrument. We use the perturbative inadequate and large distance relation from one gluon substitution [ @Nadkarni:1986as; @Nadkarni:1986cz; @McLerran:1981pb ], [ * i.e. * ] { } in the limit $ r\Lambda_{\text{QCD}}\ll1 $ zero temperature perturbation hypothesis suggests $ $ \begin{aligned }
F_1(r, T)\;\equiv\;V(r)&\simeq&-\frac{4}{3}\frac{\alpha(r)}{r}\;,\label{alp_rT1}\end{aligned}$$ while high temperature disturbance theory, [ * i.e. * ] { } $ rT\gg1 $ and $ T$ well above $ T_c$, yields $ $ \begin{aligned }
F_1(r, T)&\simeq&-\frac{4}{3}\frac{\alpha(T)}{r}e^{-m_D(T)r}\;.\label{alp_rT2}\end{aligned}$$ In both relation we have neglected any constant contributions to the barren energies which, in particular, at high temperatures will predominate the large distance behavior of the free energies. Moreover, we already anticipate here the running of the couplings with the expected dominant scales $ r$ and $ T$ in both limit. At finite temperature we specify the running coupling in analogy to $ T=0 $ as (see [ @Kaczmarek:2002mc; @Kaczmarek:2004gv]),$$\begin{aligned }
\alpha_{qq}(r, T)&\equiv&\frac{3}{4}r^2 \frac{dF_1(r, T)}{dr}\;.\label{alp_rT}\end{aligned}$$ With this definition any undetermined constant contributions to the free energy are eliminated and the coupling define here at finite temperature will reclaim the coupling at zero temperature defined in (\[alp\_qq\ ]) in the terminus ad quem of small distance. Therefore $ \alpha_{qq}(r, T)$ will show the (zero temperature) dampen in the short distance perturbative regime. In the large distance limit, however, the yoke | $T\nfq0$ {#couplatt}
--------------------------------
We extend hert our studies of jhw coupning af zero tdmperature to finite temperavurew beliw and above deconfineoent folllwing thw coiceptual approaci given lu [@Kacalaren:2004jv]. In this case the appro[riate observatld ns the color singlet quark anti-quark free emeggy and its detivatpvq. We lst the perturbative short and largs distaice relation frpm one gluon exchange [@Nadkwrni:1986ws; @Nadkarni:1986cz; @McLegran:1981pb], [*i.e.*]{} ib thq limit $r\Lambaa_{\text{QCD}}\ll1$ zero tempetature perturbation theory suggerts $$\bzgin{aligned}
D_1(r,R)\;\eqkhv\;V(r)&\simeq&-\frec{4}{3}\frac{\wlpha(r)}{r}\;,\label{alp_rT1}\end{dligned}$$ while high tekpecatuee perturbation theorb, [*i.e.*]{} $rT\gg1$ and $T$ well above $T_c$, ynelds $$\begin{aligned}
F_1(r,T)&\wineq&-\frdc{4}{3}\frdc{\aloya(T)}{f}e^{-m_S(T)c}\;.\lagel{alp_gT2}\eid{aligned}$$ Ih both relarions we have neglevtqe any constanf contwifutions to the free energies which, in pdrtjcular, at high temperatyres will dominate thg large diftance behavior of the free energies. Moreover, we dlreavy anuiglpatde jere the running of the couplings with the ex[scued dominant scalcs $r$ and $T$ in both llmojs. At finite tgmperatbde we define the runjing coopling in analody tp $T=0$ as (see [@Kaczmarek:2002mc; @Kacznarek:2004gv]),$$\begin{clitned}
\alpha_{qq}(r,T)&\equnv&\frac{3}{4}r^2 \frac{bF_1(r,T)}{dr}\;.\kabel{slp_rT}\end{aligned}$$ With thns defjnition any undetermjved constant congrinuthons to ufe free energies wre elimiiated and thd coopling qefined hege at finite temperature will racover the coupling at zero temperature dxhined in (\[alp\_qa\]) hn nhe limit of skall distancef. Therefore $\alkha_{qq}(r,T)$ wnll shuw the (zerk tempecature) weakeging in the svlrt distance perturbwtivw retime. In ghe large distsnce limin, kowever, tye coupling | $T\neq0$ {#couplatt} -------------------------------- We extend here our the at zero to finite temperatures the approach given in In this case appropriate observable is the color singlet anti-quark free energy and its derivative. We use the perturbative short and large relation from one gluon exchange [@Nadkarni:1986as; @Nadkarni:1986cz; @McLerran:1981pb], [*i.e.*]{} in the limit $r\Lambda_{\text{QCD}}\ll1$ temperature theory $$\begin{aligned} while high temperature perturbation theory, [*i.e.*]{} $rT\gg1$ and $T$ well above $T_c$, yields $$\begin{aligned} F_1(r,T)&\simeq&-\frac{4}{3}\frac{\alpha(T)}{r}e^{-m_D(T)r}\;.\label{alp_rT2}\end{aligned}$$ In relations we have neglected any constant contributions to free energies which, in at high temperatures will dominate large behavior of free Moreover, already anticipated here running of the couplings with the expected dominant scales $r$ and $T$ in both limits. At finite we define coupling in to as [@Kaczmarek:2002mc; @Kaczmarek:2004gv]),$$\begin{aligned} \alpha_{qq}(r,T)&\equiv&\frac{3}{4}r^2 this definition any undetermined constant contributions energies are eliminated and the coupling defined here finite temperature recover the coupling at zero temperature in (\[alp\_qq\]) in the limit of small distances. $\alpha_{qq}(r,T)$ will show the (zero temperature) weakening in the short distance perturbative regime. In the limit, however, the coupling | $T\neq0$ {#couplatt}
--------------------------------
We extend here oUr studies oF the cOupLinG aT zerO temPerature to finiTE temPeratures below and above DeconFiNEmenT FoLlowiNg the coNCePTUal ApPrOacH gIVeN in [@KaCzmArek:2004gv]. IN this case tHe aPpRopriate obseRVaBle is the coLor Singlet quark AntI-quark FrEe eNErgy aNd iTs derIvativE. we use tHe perturbAtIVe shorT And largE DIsTancE relation from one gLUoN Exchange [@NadkarNi:1986as; @NaDkARnI:1986CZ; @MclerRan:1981pb], [*i.e.*]{} in tHe Limit $R\lambda_{\tEXt{qcd}}\Ll1$ zERo temperature PerturbatioN TheOry sugGeSts $$\BEgin{alIgned}
f_1(r,t)\;\EquIv\;V(r)&\simeq&-\frAc{4}{3}\frAc{\alpha(r)}{r}\;,\Label{aLP_rT1}\end{aLIgned}$$ whIle higH teMpeRatuRE pErTurBaTIon THeOry, [*I.E.*]{} $rT\Gg1$ and $T$ weLl AbOve $T_c$, YielDS $$\BEGin{aLigNed}
F_1(R,T)&\simEq&-\frac{4}{3}\frac{\alpHa(T)}{R}e^{-m_D(t)R}\;.\laBel{alP_rT2}\enD{aliGnEd}$$ In bOth relAtionS wE have neglected aNy coNstant conTriBuTioNs To the FRee eneRgiEs wHich, in pArticulAR, at HiGH TEmPeratures will dominAtE THe Large disTance bEHaViOR of the frEe EneRgieS. mOreovEr, we ALrEady antiCipateD HeRe The runnInG of the CoUplIngS with THe exPected Dominant ScaleS $R$ and $T$ in both limITs. At finite temPErATUrE We deFinE the running CoupLIng iN anaLOgY to $t=0$ As (see [@kaczmArEK:2002mC; @kaczmarek:2004gv]),$$\begin{aliGnEd}
\alphA_{qq}(r,T)&\Equiv&\frac{3}{4}r^2 \fraC{dF_1(r,T)}{dr}\;.\labEL{ALp_rT}\end{aLignED}$$ WITh this definitiOn any UndeterminED constanT contRibutionS to the freE ENergies aRe eLimInaTed AND tHe coupling defINEd heRe At finitE teMperatuRe wIll RecOveR tHe couplinG at zero tEmPeRaTuRe dEfineD In (\[alp\_qq\]) iN tHe lImIt oF smalL DistanCes. ThErefOrE $\aLPha_{Qq}(r,T)$ wilL ShOW The (zErO tEmpeRatUrE) weakEninG In tHe short Distance pErtURbatIvE rEgime. In The large distaNcE limit, howeVeR, thE couplING | $T\neq0$ {#couplatt}
---- ---------- ----- --- --- -- ---- -
W e extend hereo ur s tudies of the coupling at z er o tem p er ature to fin i te t emp er at ure sb el ow an d a bove de confinemen t f ol lowing the c o nc eptual app roa ch given in[@K aczmar ek :20 0 4gv]. In this caset he app ropriateob s ervabl e is the c ol or s inglet quark anti - qu a rk free energy and i ts de r i vat ive . We use t he pert u rbative sh o r t an d large distan ce relation fro m onegl uon exchan ge [@ Na d kar ni:1986as;@Nad karni:198 6cz; @ M cLerran : 1981pb] , [*i. e.* ]{} int he l imi t$ r\L a mb da_ { \te xt{QCD}} \l l1 $ zer o te m p e r atur e p ertu rbati on theory sug ges ts $ $ \be gin{a ligne d}
F _1 (r,T) \;\equ iv\;V (r )&\simeq&-\frac {4}{ 3}\frac{\ alp ha (r) }{ r}\;, \ label{ alp _rT 1}\end{ aligned } $$wh i l e h igh temperature pe rt u r ba tion the ory, [ * i. e. * ]{} $rT\ gg 1$and$ T $ wel l ab o ve $T_c$,yields $$ \b egin{al ig ned}
F _1 (r, T)& \sime q &-\f rac{4} {3}\frac {\alp h a(T)}{r}e^{-m_ D (T)r}\;.\labe l {a l p _r T 2}\e nd{ aligned}$$In b o th r elat i on s w e have negl ec t ed any constant contri bu tionsto th e free energi es which,i n particul ar,a th igh temperatur es wi ll dominat e the lar ge di stance b ehavior o f the free en erg ies . M o r eo ver, we alrea d y ant ic ipatedher e the r unn ing of th ecouplings with th eex pe ct eddomin a nt scale s$r$ a nd$T$ i n bothlimit s. A tfi n ite temper a tu r e wede fi ne t heru nning cou p lin g in an alogy to$T= 0 $ as ( se e [@Kac zmarek:2002mc ;@Kaczmarek :2 004 gv]),$ $ \ begin{al igned}
\alpha_{qq}(r, T )&\equi v&\ frac{ 3}{4 }r^2 \fra c{d F_1(r, T)} { dr}\;. \label {alp_ rT }\e n d {alig n e d} $$Wi th this de f i nit ion a ny und etermin ed constant contri b uti ons to the fr eeener g i es ar e e l imi na t eda n d the couplingdefined he re at finite te m per at ure wil l recov er th e coupli ng at zer o tempera tu re d e f ine d in (\[al p\_qq\]) in the l i mit o f s malldis tances .The refor e $\al p ha_ {qq}( r,T)$wi ll sho w the ( zero tem perature) weakening inthe sh ort d ist ance pert urb a tiv e regime. Inthe largedis tan ce li mit , howe ver, th e c o uplin g | $T\neq0$_{#couplatt}
--------------------------------
We extend_here our studies of_the coupling_at_zero temperature_to_finite temperatures below_and above deconfinement_following the conceptual approach_given in [@Kaczmarek:2004gv]._In_this case the appropriate observable is the color singlet quark anti-quark free energy and_its_derivative. We_use_the_perturbative short and large distance_relation from one gluon exchange_[@Nadkarni:1986as; @Nadkarni:1986cz;_@McLerran:1981pb], [*i.e.*]{} in the limit $r\Lambda_{\text{QCD}}\ll1$ zero temperature_perturbation_theory suggests $$\begin{aligned}
F_1(r,T)\;\equiv\;V(r)&\simeq&-\frac{4}{3}\frac{\alpha(r)}{r}\;,\label{alp_rT1}\end{aligned}$$_while high temperature perturbation theory, [*i.e.*]{} $rT\gg1$ and $T$_well above $T_c$, yields $$\begin{aligned}
F_1(r,T)&\simeq&-\frac{4}{3}\frac{\alpha(T)}{r}e^{-m_D(T)r}\;.\label{alp_rT2}\end{aligned}$$ In_both relations we_have_neglected_any constant contributions to_the free energies which, in particular,_at high temperatures will dominate the_large distance behavior of the free energies._Moreover, we already anticipated here the_running of the couplings with_the expected_dominant scales $r$ and $T$_in both limits._At finite_temperature we define_the running coupling in analogy to_$T=0$ as (see_[@Kaczmarek:2002mc; @Kaczmarek:2004gv]),$$\begin{aligned}
\alpha_{qq}(r,T)&\equiv&\frac{3}{4}r^2 \frac{dF_1(r,T)}{dr}\;.\label{alp_rT}\end{aligned}$$ With_this_definition any undetermined_constant_contributions_to the_free energies are_eliminated_and the_coupling_defined here at finite temperature will_recover_the coupling at zero temperature defined in_(\[alp\_qq\]) in the limit_of_small distances. Therefore $\alpha_{qq}(r,T)$_will show the (zero temperature)_weakening in the short distance perturbative_regime. In_the large_distance limit, however, the coupling |
$u \in \Aut^{2,q}(P)$ is the resulting gauge transformation, depending on $A$ and $A_0$, such that $$d_{A_0}^*(u(A)-A_0) = 0,$$ then $$\|u\|_{W_{A_1}^{2,p}(X)} \leq C.$$
Write $u = u_0+\gamma$ as in the proof of Proposition \[prop:Feehan\_2001\_lemma\_6-6\], with $u_0 \in (\Ker \Delta_{A_0})^\perp$ and $\gamma \in \Ker \Delta_{A_0}$, and observe that $$\begin{aligned}
\|u\|_{W_{A_1}^{2,p}(X)} &\leq C\left(\|\Delta_{A_0}u\|_{L^p(X)} + \|u\|_{L^p(X)}\right)
\quad \text{(by Proposition \ref{prop:W2p_apriori_estimate_Delta_A_Sobolev})}
\\
&=
C\left(\|\Delta_{A_0}u_0\|_{L^p(X)} + \|u\|_{L^p(X)}\right)
\\
&\leq
C\left(\|d_{A_0}^*(A-A_0)\|_{L^p(X)} + \|u\|_{L^p(X)}\right)
\quad \text{(by \eqref{eq:Feehan_2001_6-10})}
\\
&\leq
C\left(1 + \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)\|A-A_0\|_{W^{1,p}_{A_1}(X)} + C\|u\|_{L^p(X)}
\quad \text{(by \eqref{eq:Lp_bound_dA0star_A_minus_A0})}
\\
&\leq
C\left(1 + \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)(\varepsilon + C\Vol_g(X)^{1/p},\end{aligned}$$ where the last inequality follows from with $\eps \in (0,1]$, the Sobolev embedding $W^{1,p}(X) \subset L^s | $ u \in \Aut^{2,q}(P)$ is the resulting gauge transformation, depending on $ A$ and $ A_0 $, such that $ $ d_{A_0}^*(u(A)-A_0) = 0,$$ then $ $ \|u\|_{W_{A_1}^{2,p}(X) } \leq C.$$
Write $ uracil = u_0+\gamma$ as in the validation of Proposition \[prop: Feehan\_2001\_lemma\_6 - 6\ ], with $ u_0 \in (\Ker \Delta_{A_0})^\perp$ and $ \gamma \in \Ker \Delta_{A_0}$, and observe that $ $ \begin{aligned }
\|u\|_{W_{A_1}^{2,p}(X) } & \leq C\left(\|\Delta_{A_0}u\|_{L^p(X) } + \|u\|_{L^p(X)}\right)
\quad \text{(by Proposition \ref{prop: W2p_apriori_estimate_Delta_A_Sobolev }) }
\\
& =
C\left(\|\Delta_{A_0}u_0\|_{L^p(X) } + \|u\|_{L^p(X)}\right)
\\
& \leq
C\left(\|d_{A_0}^*(A - A_0)\|_{L^p(X) } + \|u\|_{L^p(X)}\right)
\quad \text{(by \eqref{eq: Feehan_2001_6 - 10 }) }
\\
& \leq
C\left(1 + \|A_0 - A_1\|_{W^{1,p}_{A_1}(X)}\right)\|A - A_0\|_{W^{1,p}_{A_1}(X) } + C\|u\|_{L^p(X) }
\quad \text{(by \eqref{eq: Lp_bound_dA0star_A_minus_A0 }) }
\\
& \leq
C\left(1 + \|A_0 - A_1\|_{W^{1,p}_{A_1}(X)}\right)(\varepsilon + C\Vol_g(X)^{1 / p},\end{aligned}$$ where the last inequality follow from with $ \eps \in (0,1]$, the Sobolev embed $ W^{1,p}(X) \subset L^s | $u \ln \Aut^{2,q}(P)$ is the resultinn gauge transformation, vependihg on $A$ xnd $A_0$, such that $$d_{A_0}^*(u(A)-A_0) = 0,$$ then $$\|u\|_{Q_{A_1}^{2,p}(X)} \oeq C.$$
Write $u = u_0+\gamma$ ar in the iroof of Propiwition \[pro':Reehan\_2001\_lcima\_6-6\], slth $u_0 \mn (\Ker \Delta_{A_0})^\petp$ and $\gamma \in \Ker \Delta_{A_0}$, avd observe that $$\begin{aligned}
\|u\|_{W_{A_1}^{2,p}(X)} &\leq C\left(\|\Drlha_{A_0}u\|_{L^p(X)} + \|u\|_{L^p(X)}\ryght)
\atad \fvxu{(by Proposition \ref{prop:W2p_apriori_eatimate_Velta_A_Sobolev})}
\\
&=
C\lrft(\|\Delta_{A_0}u_0\|_{L^p(X)} + \|u\|_{L^p(X)}\right)
\\
&\lee
C\levt(\|d_{A_0}^*(A-A_0)\|_{L^p(X)} + \|u\|_{L^p(X)}\rigjt)
\quad \text{(vy \ezeef{eq:Feehan_2001_6-10})}
\\
&\ldq
C\left(1 + \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)\|A-A_0\|_{S^{1,p}_{A_1}(X)} + C\|u\|_{L^p(X)}
\quad \text{(by \eqref{eq:Lo_bounb_dA0star_A_minos_Z0})}
\\
&\lfx
C\left(1 + \|A_0-A_1\|_{W^{1,'}_{A_1}(X)}\rigrt)(\varepsilon + C\Vol_g(X)^{1/[},\end{alibned}$$ where the lavt unequality follows frmm with $\eps \in (0,1]$, thg Sobolev amyedding $W^{1,p}(X) \subset L^s | $u \in \Aut^{2,q}(P)$ is the resulting gauge on and $A_0$, that $$d_{A_0}^*(u(A)-A_0) = Write = u_0+\gamma$ as the proof of \[prop:Feehan\_2001\_lemma\_6-6\], with $u_0 \in (\Ker \Delta_{A_0})^\perp$ $\gamma \in \Ker \Delta_{A_0}$, and observe that $$\begin{aligned} \|u\|_{W_{A_1}^{2,p}(X)} &\leq C\left(\|\Delta_{A_0}u\|_{L^p(X)} + \|u\|_{L^p(X)}\right) \text{(by Proposition \ref{prop:W2p_apriori_estimate_Delta_A_Sobolev})} \\ &= C\left(\|\Delta_{A_0}u_0\|_{L^p(X)} + \|u\|_{L^p(X)}\right) \\ &\leq C\left(\|d_{A_0}^*(A-A_0)\|_{L^p(X)} + \|u\|_{L^p(X)}\right) \text{(by \\ C\left(1 \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)\|A-A_0\|_{W^{1,p}_{A_1}(X)} + C\|u\|_{L^p(X)} \quad \text{(by \eqref{eq:Lp_bound_dA0star_A_minus_A0})} \\ &\leq C\left(1 + \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)(\varepsilon + C\Vol_g(X)^{1/p},\end{aligned}$$ where the last follows from with $\eps \in (0,1]$, the Sobolev $W^{1,p}(X) \subset L^s | $u \in \Aut^{2,q}(P)$ is the resulting gauGe transforMatioN, dePenDiNg on $a$ and $a_0$, such that $$d_{A_0}^*(u(A)-A_0) = 0,$$ THen $$\|u\|_{w_{A_1}^{2,p}(X)} \leq C.$$
Write $u = u_0+\gamma$ as In the PrOOf of pRoPositIon \[prop:fEeHAN\_2001\_leMmA\_6-6\], wIth $U_0 \iN (\keR \DeltA_{A_0})^\pErp$ and $\gAmma \in \Ker \DEltA_{A_0}$, And observe thAT $$\bEgin{aligneD}
\|u\|_{W_{a_1}^{2,p}(X)} &\leq C\left(\|\DEltA_{A_0}u\|_{L^p(X)} + \|U\|_{L^P(X)}\rIGht)
\quAd \tExt{(by proposITion \reF{prop:W2p_apRiORi_estiMAte_DeltA_a_soBoleV})}
\\
&=
C\left(\|\Delta_{A_0}u_0\|_{L^p(X)} + \|u\|_{l^P(X)}\RIght)
\\
&\leq
C\left(\|d_{A_0}^*(a-A_0)\|_{L^p(X)} + \|u\|_{l^p(x)}\RiGHT)
\quAd \tExt{(by \eqref{Eq:feehaN_2001_6-10})}
\\
&\Leq
C\lefT(1 + \|a_0-A_1\|_{w^{1,P}_{a_1}(x)}\riGHt)\|A-A_0\|_{W^{1,p}_{A_1}(X)} + C\|u\|_{L^p(X)}
\Quad \text{(by \eQRef{Eq:Lp_boUnD_dA0STar_A_miNus_A0})}
\\
&\lEq
c\LefT(1 + \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\righT)(\varEpsilon + C\VOl_g(X)^{1/p},\eND{aligneD}$$ Where thE last iNeqUalIty fOLlOwS frOm WIth $\EPs \In (0,1]$, tHE SoBolev embEdDiNg $W^{1,p}(X) \SubsET l^S | $u \in \Aut^{2,q}(P)$ isthe result ing g aug e t ra nsfo rmat ion, depending on $ A$ and $A_0$, such tha t $$d _{ A _0}^ * (u (A)-A _0) = 0 , $$ t hen $ $\ |u\ |_ { W_ {A_1} ^{2 ,p}(X)} \leq C.$$
W ri te $u = u_0+ \ ga mma$ as in th e proof of P rop ositio n\[p r op:Fe eha n\_20 01\_le m ma\_6- 6\], with $ u _0 \in (\Ker \ D e lt a_{A _0})^\perp$ and $ \ ga m ma \in \Ker \D elta_{ A_ 0 }$ , and ob serve that $ $\beg i n{align e d} \ | u\| _ {W_{A_1}^{2,p }(X)} &\leq C\l eft(\| \D elt a _{A_0} u\|_{ L^ p (X) } + \|u\|_{ L^p( X)}\right )
\qua d \text{ ( by Prop ositio n \ ref {pro p :W 2p _ap ri o ri_ e st ima t e_D elta_A_S ob ol ev})}
\\& = C \lef t(\ |\De lta_{ A_0}u_0\|_{L^ p(X )} + \|u \|_{L ^p(X) }\ri gh t)
\\
&\leq
C\le ft (\|d_{A_0}^*(A- A_0) \|_{L^p(X )}+\|u \| _{L^p ( X)}\ri ght )
\ quad \t ext{(by \eq re f { e q: Feehan_2001_6-10}) }\ \
& \leq
C\l eft(1+ \ |A _ 0-A_1\|_ {W ^{1 ,p}_ { A _1}(X )}\r i gh t)\|A-A_ 0\|_{W ^ {1 ,p }_{A_1} (X )} + C \| u\| _{L ^p(X) }
\qu ad \te xt{(by \ eqref { eq:Lp_bound_dA 0 star_A_minus_ A 0} ) }
\ \
&\l eqC\left(1 +\|A_ 0 -A_1 \|_{ W ^{ 1,p } _{A_1 }(X)} \r i gh t )(\varepsilon + C\V ol _g(X)^ {1/p} ,\end{aligned }$$ wheret h e last in equa l it y follows fromwith$\eps \in( 0,1]$, t he So bolev em bedding $ W ^ {1,p}(X) \s ubs etL^s | $u_\in \Aut^{2,q}(P)$_is the resulting gauge_transformation, depending_on_$A$ and_$A_0$,_such that $$d_{A_0}^*(u(A)-A_0)_= 0,$$ then_$$\|u\|_{W_{A_1}^{2,p}(X)} \leq C.$$
Write $u_= u_0+\gamma$ as_in_the proof of Proposition \[prop:Feehan\_2001\_lemma\_6-6\], with $u_0 \in (\Ker \Delta_{A_0})^\perp$ and $\gamma \in \Ker_\Delta_{A_0}$,_and observe_that_$$\begin{aligned}
\|u\|_{W_{A_1}^{2,p}(X)}_&\leq C\left(\|\Delta_{A_0}u\|_{L^p(X)} + \|u\|_{L^p(X)}\right)
\quad \text{(by_Proposition \ref{prop:W2p_apriori_estimate_Delta_A_Sobolev})}
\\
&=
C\left(\|\Delta_{A_0}u_0\|_{L^p(X)} + \|u\|_{L^p(X)}\right)
\\
&\leq
C\left(\|d_{A_0}^*(A-A_0)\|_{L^p(X)} +_\|u\|_{L^p(X)}\right)
\quad \text{(by_\eqref{eq:Feehan_2001_6-10})}
\\
&\leq
C\left(1 + \|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)\|A-A_0\|_{W^{1,p}_{A_1}(X)} + C\|u\|_{L^p(X)}
\quad \text{(by \eqref{eq:Lp_bound_dA0star_A_minus_A0})}
\\
&\leq
C\left(1 +_\|A_0-A_1\|_{W^{1,p}_{A_1}(X)}\right)(\varepsilon_+ C\Vol_g(X)^{1/p},\end{aligned}$$ where_the last inequality follows from with $\eps \in (0,1]$,_the Sobolev embedding $W^{1,p}(X) \subset L^s |
a(t)$ via a designer approach, we argue that no sensible Minkowski limit exists for this theory once this connection has been made. For a brief discussion of this see Appendix \[app:B\]. In the context of the Equation of State approach, in the limit of $H \rightarrow 0$ we see that $\rho$, $P \rightarrow 0$ from and. Therefore, the expressions for $w_{\mathrm{de}}\Pi^S$ and $w_{\mathrm{de}}\Gamma$ cannot be computed since $w_{\mathrm{de}}\Pi^S$ appears as $P\Pi^S$ from the perturbed energy momentum tensor and $w_{\mathrm{de}}\Gamma$ can be written as $w_{\mathrm{de}}\rho\Gamma = \left(\frac{\delta P}{\delta \rho}-\frac{dP}{d\rho} \right) \delta \rho$.
Special cases
-------------
### $w_{\mathrm{de}}=-1$
Consider the case where we have exactly $w_{\mathrm{de}}=-1$, equivalent to $\Lambda\mathrm{CDM}$. From we have an analytical solution given by and in this case the $c_\Pi$ and $c_\Gamma$ coefficients reduce to $$c_{\Pi\Delta} = \frac{c_{13}}{c_{14}}, \quad
c_{\Pi\Theta} = \frac{1}{2}\left(1+\epsilon_H \right) - \frac{c_{13}}{c_{14}},\quad
c_{\Pi X} = 0, \quad
c_{\Pi Y} = -\frac{c_{13}}{3\alpha}\left( 1+\frac{M^2\mathcal{F}_0}{6\Omega_{\mathrm{de},0}H_0^2}\right)\left( \frac{H}{H_0}\right) ,$$ and also $$c_{\Gamma \Delta} = -c_{\Gamma \Theta} = -\frac{dP}{d\rho}=1, \quad
c_{\Gamma W} = c_{\Gamma X} = c_{\Gamma Y} = 0,$$ and hence $\Gamma = \delta$. Here we see that from $c_{\Pi Y}$, as with the background evolution, $M$ and $\mathcal{F}_0$ are degenerate.
This case is indistinguishable from $\Lambda\mathrm{CDM}$ at background order, but at | a(t)$ via a designer approach, we argue that no sensible Minkowski limit exist for this hypothesis once this connection has been made. For a abbreviated discussion of this see Appendix \[app: B\ ]. In the context of the Equation of State approach path, in the limit of $ H \rightarrow 0 $ we watch that $ \rho$, $ P \rightarrow 0 $ from and. Therefore, the expression for $ w_{\mathrm{de}}\Pi^S$ and $ w_{\mathrm{de}}\Gamma$ cannot be computed since $ w_{\mathrm{de}}\Pi^S$ appears as $ P\Pi^S$ from the perturb energy momentum tensor and $ w_{\mathrm{de}}\Gamma$ can be written as $ w_{\mathrm{de}}\rho\Gamma = \left(\frac{\delta P}{\delta \rho}-\frac{dP}{d\rho } \right) \delta \rho$.
limited cases
-------------
# # # $ w_{\mathrm{de}}=-1 $
Consider the case where we have exactly $ w_{\mathrm{de}}=-1 $, equivalent to $ \Lambda\mathrm{CDM}$. From we have an analytic solution given by and in this case the $ c_\Pi$ and $ c_\Gamma$ coefficients abridge to $ $ c_{\Pi\Delta } = \frac{c_{13}}{c_{14 } }, \quad
c_{\Pi\Theta } = \frac{1}{2}\left(1+\epsilon_H \right) - \frac{c_{13}}{c_{14}},\quad
c_{\Pi X } = 0, \quad
c_{\Pi Y } = -\frac{c_{13}}{3\alpha}\left (1+\frac{M^2\mathcal{F}_0}{6\Omega_{\mathrm{de},0}H_0 ^ 2}\right)\left (\frac{H}{H_0}\right) , $ $ and also $ $ c_{\Gamma \Delta } = -c_{\Gamma \Theta } = -\frac{dP}{d\rho}=1, \quad
c_{\Gamma W } = c_{\Gamma X } = c_{\Gamma Y } = 0,$$ and therefore $ \Gamma = \delta$. Here we see that from $ c_{\Pi Y}$, as with the backdrop evolution, $ M$ and $ \mathcal{F}_0 $ are degenerate.
This case is identical from $ \Lambda\mathrm{CDM}$ at background order, but at | a(t)$ gia a designer approach, de argue that no sensiule Miniowski lkmit exists for this theory lnxe thus connection has been made. For a brief disrussion of this see Appekbix \[ali:B\]. In vhe context of jhe Equation of State apprmazh, in the limit of $H \rightarrow 0$ we sqe that $\rjo$, $P \rightarror 0$ fgoi ans. Therefore, the expressions for $w_{\mzthrm{de}}\Ki^S$ and $w_{\mathrm{de}}\Gsmma$ cannot be computed sijce $a_{\mathrm{de}}\Pi^S$ appeags as $P\Pi^S$ drom rhe perturbea energy momentum tenskr and $w_{\mathrm{de}}\Gamma$ can be wrktten as $w_{\mathrn{dw}}\rhl\Camma = \left(\hrac{\dejta P}{\delta \rmp}-\frac{d[}{d\rho} \roght) \delta \rho$.
Xperial cases
-------------
### $w_{\mathrm{de}}=-1$
Consiver the case where wg have exawtmy $w_{\mathrm{de}}=-1$, equicaoent jo $\Lakbda\oqthfm{CSM}$. Fdom we hate an analyfical solutuon given by and in trpx case the $c_\Li$ and $c_\Damma$ coefficients reduce to $$c_{\Pi\Delta} = \xrad{c_{13}}{c_{14}}, \quad
c_{\Pi\Theta} = \frac{1}{2}\lwft(1+\epsilon_H \right) - \frwc{c_{13}}{c_{14}},\quad
c_{\[i X} = 0, \quad
c_{\Pi Y} = -\frac{c_{13}}{3\alpha}\left( 1+\frac{M^2\mathcal{F}_0}{6\Omaga_{\mavhfm{dt},0}H_0^2}\vlght)\uwfh( \frac{H}{H_0}\right) ,$$ and also $$c_{\Gamma \Delta} = -c_{\Gamma \Fhtta} = -\frac{dP}{d\rho}=1, \qmad
c_{\Gamma W} = c_{\Gamms D} = s_{\Gamma Y} = 0,$$ ana hencz $\Gzmma = \delta$. Here wf see trat feom $c_{\Pi Y}$, as eith the background evolutiin, $M$ and $\matkcao{F}_0$ are degenerate.
Tkis case is nndistonguixhable from $\Lambda\mathro{CDM}$ at backgroknd order, cut at | a(t)$ via a designer approach, we argue sensible limit exists this theory once For brief discussion of see Appendix \[app:B\]. the context of the Equation of approach, in the limit of $H \rightarrow 0$ we see that $\rho$, $P 0$ from and. Therefore, the expressions for $w_{\mathrm{de}}\Pi^S$ and $w_{\mathrm{de}}\Gamma$ cannot be computed $w_{\mathrm{de}}\Pi^S$ as from perturbed energy momentum tensor and $w_{\mathrm{de}}\Gamma$ can be written as $w_{\mathrm{de}}\rho\Gamma = \left(\frac{\delta P}{\delta \rho}-\frac{dP}{d\rho} \right) \rho$. Special cases ------------- ### $w_{\mathrm{de}}=-1$ Consider the where we have exactly equivalent to $\Lambda\mathrm{CDM}$. From we an solution given and this the $c_\Pi$ and coefficients reduce to $$c_{\Pi\Delta} = \frac{c_{13}}{c_{14}}, \quad c_{\Pi\Theta} = \frac{1}{2}\left(1+\epsilon_H \right) - \frac{c_{13}}{c_{14}},\quad c_{\Pi X} = 0, c_{\Pi Y} 1+\frac{M^2\mathcal{F}_0}{6\Omega_{\mathrm{de},0}H_0^2}\right)\left( \frac{H}{H_0}\right) and $$c_{\Gamma = -c_{\Gamma \Theta} \quad c_{\Gamma W} = c_{\Gamma X} = 0,$$ and hence $\Gamma = \delta$. Here see that $c_{\Pi Y}$, as with the background $M$ and $\mathcal{F}_0$ are degenerate. This case is from $\Lambda\mathrm{CDM}$ at background order, but at | a(t)$ via a designer approach, we aRgue that no SensiBle minKoWski LimiT exists for this THeorY once this connection has Been mAdE. for a BRiEf disCussion OF tHIS seE APpEndIx \[APp:b\]. In thE coNtext of The EquatioN of stAte approach, iN ThE limit of $H \rIghTarrow 0$ we see tHat $\Rho$, $P \riGhTarROw 0$ froM anD. TherEfore, tHE expreSsions for $W_{\mAThrm{de}}\pI^S$ and $w_{\mATHrM{de}}\GAmma$ cannot be compuTEd SInce $w_{\mathrm{de}}\PI^S$ appeArS As $p\pI^S$ fRom The perturbEd EnergY MomentuM TeNSOR anD $W_{\mathrm{de}}\GammA$ can be writtEN as $W_{\mathrM{dE}}\rhO\gamma = \lEft(\frAc{\DEltA P}{\delta \rho}-\fRac{dp}{d\rho} \righT) \delta \RHo$.
SpeciAL cases
-------------
### $w_{\Mathrm{De}}=-1$
COnsIder THe CaSe wHeRE we HAvE exACtlY $w_{\mathrm{De}}=-1$, EqUivalEnt tO $\lAMBda\mAthRm{CDm}$. From We have an analyTicAl soLUtiOn givEn by aNd in ThIs casE the $c_\PI$ and $c_\gaMma$ coefficients ReduCe to $$c_{\Pi\DeLta} = \FrAc{c_{13}}{C_{14}}, \qUad
c_{\PI\theta} = \fRac{1}{2}\LefT(1+\epsiloN_H \right) - \FRac{C_{13}}{c_{14}},\QUAD
c_{\pi X} = 0, \quad
c_{\Pi Y} = -\frac{c_{13}}{3\alPhA}\LEfT( 1+\frac{M^2\maThcal{F}_0}{6\oMeGa_{\MAthrm{de},0}H_0^2}\RiGht)\Left( \FRAc{H}{H_0}\rIght) ,$$ ANd Also $$c_{\GamMa \DeltA} = -C_{\GAmMa \Theta} = -\FrAc{dP}{d\rHo}=1, \QuaD
c_{\GAmma W} = C_{\gammA X} = c_{\GamMa Y} = 0,$$ and heNce $\GaMMa = \delta$. Here we sEE that from $c_{\Pi Y}$, AS wITH tHE bacKgrOund evolutiOn, $M$ aND $\matHcal{f}_0$ ArE deGEneraTe.
ThiS cASe IS indistinguishable fRoM $\LambdA\mathRm{CDM}$ at backgrOund order, bUT AT | a(t)$ via a designer appro ach, we ar gue t hat no s ensi bleMinkowski limi t exi sts for this theory on ce th is conn e ct ion h as been ma d e . F or a br ie f d iscus sio n of th is see App end ix \[app:B\].I nthe contex t o f the Equati onof Sta te ap p roach , i n the limit of $H\rightarr ow 0$ wes ee that $ \r ho$, $P \rightarrow 0 $ f r om and. Theref ore, t he ex p r ess ion s for $w_{ \m athrm { de}}\Pi ^ S$ a n d $ w _{\mathrm{de} }\Gamma$ ca n not be co mp ute d since $w_{ \m a thr m{de}}\Pi^S $ ap pears as$P\Pi^ S $ fromt he pert urbedene rgy mom e nt um te ns o r a n d$w_ { \ma thrm{de} }\ Ga mma$canb e w ritt enas $ w_{\m athrm{de}}\rh o\G amma = \ left( \frac {\de lt a P}{ \delta \rho }- \frac{dP}{d\rho } \r ight) \de lta \ rho $.
Spe c ial ca ses
-- ------- ----
# # # $ w_ { \ m at hrm{de}}=-1$
Cons id e r t he casewherew eha v e exactl y$w_ {\ma t h rm{de }}=- 1 $, equival ent to $\ La mbda\ma th rm{CDM }$ . F rom we h a ve a n anal ytical s oluti o n given by and in this caset he $ c_ \ Pi$and $c_\Gamma$ coe f fici ents re duc e to $ $c_{\ Pi \ De l ta} = \frac{c_{13}} {c _{14}} , \qu ad
c_{\Pi\The ta} = \fra c { 1 }{2}\lef t(1+ \ ep s ilon_H \right) - \f rac{c_{13} } {c_{14}} ,\qua d
c_{\Pi X} = 0,\ q uad
c_{\ PiY}= -\f r a c{ c_{13}}{3\alp h a }\le ft ( 1+\fr ac{ M^2\mat hca l{F }_0 }{6 \O mega_{\ma thrm{de} ,0 }H _0 ^2 }\r ight) \ left( \f ra c{H }{ H_0 }\rig h t) ,$ $ and als o$$ c _{\ Gamma \ D el t a } =-c _{ \Gam ma\T heta} = - \ fra c{dP}{d \rho}=1,\qu a d
c_ {\ Ga mma W}= c_{\Gamma X }= c_{\Gamm aY}= 0,$$ a nd hence $\Gamma = \delta$. Her e we see th at fr om $ c_{\Pi Y} $,as wit h t h e back ground evol ut ion , $M$ a n d $ \ma th cal{F}_0$a r e d egene ra te.
This c ase is indistingui s hab le from $\Lam bda \mat h r m{ CDM } $a t b ac k gro u n d order, but at | a(t)$ via_a designer_approach, we argue that_no sensible_Minkowski_limit exists_for_this theory once_this connection has_been made. For a_brief discussion of_this_see Appendix \[app:B\]. In the context of the Equation of State approach, in the_limit_of $H_\rightarrow_0$_we see that $\rho$, $P_\rightarrow 0$ from and. Therefore,_the expressions_for $w_{\mathrm{de}}\Pi^S$ and $w_{\mathrm{de}}\Gamma$ cannot be computed since_$w_{\mathrm{de}}\Pi^S$_appears as $P\Pi^S$_from the perturbed energy momentum tensor and $w_{\mathrm{de}}\Gamma$ can_be written as $w_{\mathrm{de}}\rho\Gamma = \left(\frac{\delta_P}{\delta \rho}-\frac{dP}{d\rho} \right)_\delta_\rho$.
Special_cases
-------------
### $w_{\mathrm{de}}=-1$
Consider the case_where we have exactly $w_{\mathrm{de}}=-1$, equivalent_to $\Lambda\mathrm{CDM}$. From we have an_analytical solution given by and in this_case the $c_\Pi$ and $c_\Gamma$ coefficients_reduce to $$c_{\Pi\Delta} = \frac{c_{13}}{c_{14}},_\quad
c_{\Pi\Theta} =_\frac{1}{2}\left(1+\epsilon_H \right) - \frac{c_{13}}{c_{14}},\quad
c_{\Pi X}_= 0, \quad
c_{\Pi_Y} =_ -\frac{c_{13}}{3\alpha}\left( 1+\frac{M^2\mathcal{F}_0}{6\Omega_{\mathrm{de},0}H_0^2}\right)\left(_\frac{H}{H_0}\right) ,$$ and also $$c_{\Gamma_\Delta} = -c_{\Gamma_\Theta} = -\frac{dP}{d\rho}=1, \quad
c_{\Gamma W} =_c_{\Gamma_X} = c_{\Gamma_Y}_=_0,$$ and_hence $\Gamma =_\delta$._Here we_see_that from $c_{\Pi Y}$, as with_the_background evolution, $M$ and $\mathcal{F}_0$ are degenerate.
This_case is indistinguishable from_$\Lambda\mathrm{CDM}$_at background order, but_at |
-step nilpotent metric Lie algebras ${\mathfrak{n}}$ with $\dim({\mathfrak{n}}')\leq 2$, and for each Lie algebra in the remaining case $\dim({\mathfrak{n}}')=3$ we construct purely coclosed G$_2$-structures, as well as metrics which are not compatible with any purely coclosed G$_2$-structure.
Preliminaries on G$_2$-structures {#sectPCCG2}
=================================
Basic definitions {#BasicDef}
-----------------
A G$_2$-structure on a 7-dimensional vector space $V$ is defined by a 3-form ${\varphi}\in\Lambda^3V^*$ satisfying the non-degeneracy condition $$\label{NonDegG2}
v{\lrcorner}{\varphi}{\wedge}v {\lrcorner}{\varphi}{\wedge}{\varphi}\neq 0, \quad \forall~v\in V\smallsetminus\{0\}.$$ Since the stabilizer ${{\mathrm {GL}}}(V)_{\varphi}\subset{{\mathrm {GL}}}(V)$ of any such 3-form is isomorphic to the exceptional Lie group G$_2$, the set $\Lambda^3_{{\scriptscriptstyle}+}V^*$ of all G$_2$-structures on $V$ is isomorphic to ${{\mathrm {GL}}}(7,{\mathbb R})/{{\mathrm G}}_2$ and thus open in $\Lambda^3V^*$.
A G$_2$-structure ${\varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ gives rise to a unique inner product $g_{\varphi}$ and orientation on $V$ with corresponding volume form $\operatorname{vol}_{\varphi}$ satisfying $$\label{metvolG2}
g_{\varphi}(v,w)\operatorname{vol}_{\varphi}= \frac16\, v{\lrcorner}{\varphi}{\wedge}w{\lrcorner}{\varphi}{\wedge}{\varphi}.$$ Moreover, there exists a $g_{\varphi}$-orthonormal basis $\mathcal{B} = \{e_1,\ldots,e_7\}$ of $V$ with dual basis $\mathcal{B}^*=\{e^1,\ldots,e^7\}$ such that $$\label{G2adapted}
\begin{split}
{\varphi}&= e^{127}+e^{347}+e^{567} + e^{135}-e^{146}-e^{236}-e^{245}, \\
*_{\varphi}{\varphi}&= e^{1234 | -step nilpotent metric Lie algebras $ { \mathfrak{n}}$ with $ \dim({\mathfrak{n}}')\leq 2 $, and for each Lie algebra in the remaining case $ \dim({\mathfrak{n}}')=3 $ we manufacture strictly coclosed G$_2$-structures, as well as metrics which are not compatible with any strictly coclosed G$_2$-structure.
Preliminaries on G$_2$-structures { # sectPCCG2 }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Basic definition { # BasicDef }
-----------------
A G$_2$-structure on a 7 - dimensional vector space $ V$ is defined by a 3 - phase $ { \varphi}\in\Lambda^3V^*$ satisfying the non - degeneracy condition $ $ \label{NonDegG2 }
v{\lrcorner}{\varphi}{\wedge}v { \lrcorner}{\varphi}{\wedge}{\varphi}\neq 0, \quad \forall ~ v\in V\smallsetminus\{0\}.$$ Since the stabilizer $ { { \mathrm { GL}}}(V)_{\varphi}\subset{{\mathrm { GL}}}(V)$ of any such 3 - kind is isomorphic to the exceptional Lie group G$_2 $, the hardening $ \Lambda^3_{{\scriptscriptstyle}+}V^*$ of all G$_2$-structures on $ V$ is isomorphic to $ { { \mathrm { GL}}}(7,{\mathbb R})/{{\mathrm G}}_2 $ and thus unfold in $ \Lambda^3V^*$.
A G$_2$-structure $ { \varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ gives rise to a unique inner product $ g_{\varphi}$ and predilection on $ V$ with corresponding volume form $ \operatorname{vol}_{\varphi}$ meet $ $ \label{metvolG2 }
g_{\varphi}(v, w)\operatorname{vol}_{\varphi}= \frac16\, v{\lrcorner}{\varphi}{\wedge}w{\lrcorner}{\varphi}{\wedge}{\varphi}.$$ Moreover, there exists a $ g_{\varphi}$-orthonormal basis $ \mathcal{B } = \{e_1,\ldots, e_7\}$ of $ V$ with double basis $ \mathcal{B}^*=\{e^1,\ldots, e^7\}$ such that $ $ \label{G2adapted }
\begin{split }
{ \varphi}&= e^{127}+e^{347}+e^{567 } + e^{135}-e^{146}-e^{236}-e^{245 }, \\
* _ { \varphi}{\varphi}&= e^{1234 | -steo nilpotent metric Lie augebras ${\mathfrak{n}}$ with $\dim({\mafhfrak{n}}')\ldq 2$, and for each Lie algebra ib the remaining case $\dim({\matffrak{n}}')=3$ we construxt pnrely coclosed G$_2$-structurcf, as aell es metrics whicm are not cmmpatible with avy purely coclosed G$_2$-structure.
Preliminwries om H$_2$-structures {#sestPCBG2}
=================================
Fasid definitions {#BasicDef}
-----------------
A G$_2$-structure on a 7-dpmensional vector space $V$ is defined by a 3-flrm ${\garphi}\in\Lambda^3V^*$ sahisfying thg nog-eegeneracy cundition $$\label{NonDegG2}
v{\mrcorner}{\varphi}{\wedge}v {\lrcorner}{\vafphi}{\wzdge}{\varphi}\ngq 0, \eoad \forall~v\ii V\smajlsetminus\{0\}.$$ Slmce tha stabikizer ${{\mathrm {GK}}}(V)_{\verphu}\subset{{\mathrm {GL}}}(V)$ of eny such 3-form is isoiorphic tm che exceptional Lie geoyp G$_2$, jhe sat $\Lxnbdx^3_{{\scdi'tsdriptshylx}+}V^*$ of all G$_2$-atructures in $V$ is isomorphic uo ${{\iqthrm {GL}}}(7,{\mathbg R})/{{\matrri G}}_2$ and thus open in $\Lambda^3V^*$.
A G$_2$-structurt ${\varlhi}\in\Lambda^3_{{\scriptscriptwtyle}+}V^*$ gives rise to w unique ynner product $g_{\varphi}$ and orientation on $V$ with cmrres'ovdiun xilkme form $\operatorname{vol}_{\varphi}$ satisfying $$\labqm{mttvplG2}
g_{\varphi}(v,w)\opcratorname{vol}_{\varpho}= \vrss16\, v{\lrcorner}{\vatphi}{\wedys}w{\mrcorner}{\varphi}{\wedgf}{\varphi}.$$ Moreiver, thert exixts a $g_{\varphi}$-orthonormal bawis $\mathcal{B} = \{e_1,\ldots,e_7\}$ of $V$ with dual basis $\matncal{B}^*=\{r^1,\ldots,e^7\}$ such that $$\label{Y2adaptsd}
\begin{splih}
{\varphi}&= e^{127}+s^{347}+d^{567} + e^{135}-e^{146}-e^{236}-e^{245}, \\
*_{\varphi}{\vafphp}&= e^{1234 | -step nilpotent metric Lie algebras ${\mathfrak{n}}$ with and each Lie in the remaining coclosed as well as which are not with any purely coclosed G$_2$-structure. Preliminaries G$_2$-structures {#sectPCCG2} ================================= Basic definitions {#BasicDef} ----------------- A G$_2$-structure on a 7-dimensional vector $V$ is defined by a 3-form ${\varphi}\in\Lambda^3V^*$ satisfying the non-degeneracy condition $$\label{NonDegG2} v{\lrcorner}{\varphi}{\wedge}v 0, \forall~v\in Since stabilizer ${{\mathrm {GL}}}(V)_{\varphi}\subset{{\mathrm {GL}}}(V)$ of any such 3-form is isomorphic to the exceptional Lie group G$_2$, set $\Lambda^3_{{\scriptscriptstyle}+}V^*$ of all G$_2$-structures on $V$ is to ${{\mathrm {GL}}}(7,{\mathbb R})/{{\mathrm and thus open in $\Lambda^3V^*$. G$_2$-structure gives rise a inner $g_{\varphi}$ and orientation $V$ with corresponding volume form $\operatorname{vol}_{\varphi}$ satisfying $$\label{metvolG2} g_{\varphi}(v,w)\operatorname{vol}_{\varphi}= \frac16\, v{\lrcorner}{\varphi}{\wedge}w{\lrcorner}{\varphi}{\wedge}{\varphi}.$$ Moreover, there exists a $g_{\varphi}$-orthonormal basis = \{e_1,\ldots,e_7\}$ with dual $\mathcal{B}^*=\{e^1,\ldots,e^7\}$ that \begin{split} {\varphi}&= e^{127}+e^{347}+e^{567} \\ *_{\varphi}{\varphi}&= e^{1234 | -step nilpotent metric Lie algEbras ${\mathfRak{n}}$ wIth $\Dim({\MaThfrAk{n}}')\lEq 2$, and for each LiE AlgeBra in the remaining case $\dIm({\matHfRAk{n}}')=3$ wE CoNstruCt purelY CoCLOseD G$_2$-StRucTuREs, As welL as Metrics Which are noT coMpAtible with anY PuRely coclosEd G$_2$-Structure.
PreLimInarieS oN G$_2$-sTRuctuRes {#SectPcCG2}
=================================
BasIC definItions {#BasIcdEf}
-----------------
A G$_2$-stRUcture oN A 7-DiMensIonal vector space $V$ IS dEFined by a 3-form ${\vaRphi}\in\laMBdA^3v^*$ SatIsfYing the non-DeGenerACy condiTIoN $$\LABel{nOnDegG2}
v{\lrcornEr}{\varphi}{\wedGE}v {\lRcorneR}{\vArpHI}{\wedge}{\VarphI}\nEQ 0, \quAd \forall~v\in v\smaLlsetminuS\{0\}.$$ Since THe stabiLIzer ${{\matHrm {GL}}}(V)_{\VarPhi}\SubsET{{\mAtHrm {gL}}}(v)$ Of aNY sUch 3-FOrm Is isomorPhIc To the ExcePTIONal LIe gRoup g$_2$, the sEt $\Lambda^3_{{\scripTscRiptSTylE}+}V^*$ of aLl G$_2$-stRuctUrEs on $V$ Is isomOrphiC tO ${{\mathrm {GL}}}(7,{\mathbb r})/{{\matHrm G}}_2$ and thUs oPeN in $\laMbda^3V^*$.
a g$_2$-strucTurE ${\vaRphi}\in\LAmbda^3_{{\scRIptScRIPTsTyle}+}V^*$ gives rise to a uNiQUE iNner prodUct $g_{\vaRPhI}$ aND orientaTiOn oN $V$ wiTH CorreSponDInG volume fOrm $\opeRAtOrName{vol}_{\VaRphi}$ saTiSfyIng $$\Label{MEtvoLG2}
g_{\varPhi}(v,w)\opeRatorNAme{vol}_{\varphi}= \frAC16\, v{\lrcorner}{\varPHi}{\WEDgE}W{\lrcOrnEr}{\varphi}{\wedGe}{\vaRPhi}.$$ MOreoVEr, TheRE exisTs a $g_{\vArPHi}$-ORthonormal basis $\mathCaL{B} = \{e_1,\ldoTs,e_7\}$ of $v$ with dual basiS $\mathcal{B}^*=\{e^1,\LDOTs,e^7\}$ such tHat $$\lABeL{g2adapted}
\begin{sPlit}
{\vArphi}&= e^{127}+e^{347}+e^{567} + e^{135}-e^{146}-E^{236}-E^{245}, \\
*_{\varphi}{\vArphi}&= E^{1234 | -step nilpotent metric Lie algebras${\ma thf rak {n }}$with $\dim({\mathf r ak{n }}')\leq 2$, and for e ach L ie alge b ra in t he rema i ni n g ca se $ \di m( { \m athfr ak{ n}}')=3 $ we const ruc tpurely coclo s ed G$_2$-str uct ures, as wel l a s metr ic s w h ich a renot c ompati b le wit h any pur el y coclo s ed G$_2 $ - st ruct ure.
Preliminari e so n G$_2$-struct ures { #s e ct P C CG2 }
= ========== == ===== = ======= = == = = = =
B asic definiti ons {#Basic D ef}
----- -- --- - ------
A G $_ 2 $-s tructure on a 7 -dimensio nal ve c tor spa c e $V$ i s defi ned by a 3 - fo rm ${ \v a rph i }\ in\ L amb da^3V^*$ s at isfyi ng t h e n on-d ege nera cy co ndition $$\la bel {Non D egG 2}
v{ \lrco rner }{ \varp hi}{\w edge} v{\lrcorner}{\va rphi }{\wedge} {\v ar phi }\ neq 0 , \quad \f ora ll~v\in V\smal l set mi n u s \{ 0\}.$$ Since the s ta b i li zer ${{\ mathrm {G L} } }(V)_{\v ar phi }\su b s et{{\ math r m{GL}}}(V )$ ofa ny s uch 3-f or m is i so mor phi c tot he e xcepti onal Lie grou p G$_2$, the se t $\Lambda^3_{ { \s c r ip t scri pts tyle}+}V^*$ ofa ll G $_2$ - st ruc t ureson $V $i si somorphic to ${{\ma th rm {GL }}}(7 ,{\mathbb R}) /{{\mathrm G } }_2$ and thu s o p en in $\Lambda ^3V^* $.
A G$_2 $ -structu re ${ \varphi} \in\Lambd a ^ 3_{{\scr ipt scr ipt sty l e }+ }V^*$ gives r i s e to a unique in ner pro duc t $ g_{ \va rp hi}$ andorientat io non $ V$withc orrespon di ngvo lum e for m $\ope rator name {v ol } _{\ varphi} $ s a t isfy in g$$\l abe l{ metvo lG2} g_{ \varphi }(v,w)\op era t orna me {v ol}_{\v arphi}= \frac 16 \, v{\lrco rn er} {\varp h i }{\wedge }w{\lrcorner}{\varphi}{ \ wedge}{ \va rphi} .$$Moreover, th ere ex ist s a $g_ {\varp hi}$- or tho n o rmalb a si s $ \m athcal{B}= \{e _1,\l do ts,e _7\}$ o f $V$ with dual ba s is$\mathcal{B}^ *=\ {e^1 , \ ld ots , e^ 7 \}$ s u cht h at $$\label{G2a dapted}
\b eg i n{ split}
{\v a rph i} &= e^{1 27}+e^{ 347}+ e ^{567}+ e^{135} -e^{146}- e^ {236 } - e^{ 245}, \\
* _{\varph i}{\varph i }&= e ^ {1 234 | -step nilpotent_metric Lie_algebras ${\mathfrak{n}}$ with $\dim({\mathfrak{n}}')\leq_2$, and_for_each Lie_algebra_in the remaining_case $\dim({\mathfrak{n}}')=3$ we_construct purely coclosed G$_2$-structures,_as well as_metrics_which are not compatible with any purely coclosed G$_2$-structure.
Preliminaries on G$_2$-structures {#sectPCCG2}
=================================
Basic definitions {#BasicDef}
-----------------
A_G$_2$-structure_on a_7-dimensional_vector_space $V$ is defined by_a 3-form ${\varphi}\in\Lambda^3V^*$ satisfying the_non-degeneracy condition_$$\label{NonDegG2}
v{\lrcorner}{\varphi}{\wedge}v {\lrcorner}{\varphi}{\wedge}{\varphi}\neq 0, \quad \forall~v\in V\smallsetminus\{0\}.$$ Since the_stabilizer_${{\mathrm {GL}}}(V)_{\varphi}\subset{{\mathrm {GL}}}(V)$_of any such 3-form is isomorphic to the exceptional_Lie group G$_2$, the set $\Lambda^3_{{\scriptscriptstyle}+}V^*$_of all G$_2$-structures_on_$V$_is isomorphic to ${{\mathrm_{GL}}}(7,{\mathbb R})/{{\mathrm G}}_2$ and thus open_in $\Lambda^3V^*$.
A G$_2$-structure ${\varphi}\in\Lambda^3_{{\scriptscriptstyle}+}V^*$ gives rise_to a unique inner product $g_{\varphi}$ and_orientation on $V$ with corresponding volume_form $\operatorname{vol}_{\varphi}$ satisfying $$\label{metvolG2}
g_{\varphi}(v,w)\operatorname{vol}_{\varphi}= \frac16\,_v{\lrcorner}{\varphi}{\wedge}w{\lrcorner}{\varphi}{\wedge}{\varphi}.$$ Moreover,_there exists a $g_{\varphi}$-orthonormal basis_$\mathcal{B} = \{e_1,\ldots,e_7\}$_of $V$_with dual basis_$\mathcal{B}^*=\{e^1,\ldots,e^7\}$ such that $$\label{G2adapted}
\begin{split}
{\varphi}&= e^{127}+e^{347}+e^{567} +_e^{135}-e^{146}-e^{236}-e^{245}, \\
*_{\varphi}{\varphi}&= e^{1234 |
, Phys. Rev. B 72 (2005) 165204. T. Omiya, F. Matsukura, A. Shen, Y. Ohno, H. Ohno, Physica E 10 (2001) 206. M. Reinwald, U. Wurstbauer, M. D[ö]{}ppe, W. Kipferl, K. Wagenhuber, H.-P. Tranitz, D. Weiss, W. Wegscheider, J. Cryst. Growth 278 (2005) 690. K. Y. Wang, K. W. Edmonds, L. X. Zhao, M. Sawicki, R. P. Campion, B. L. Gallagher, C. T. Foxon, Phys. Rev. B 72 (2005) 115207. J. Daeubler, M. Glunk, W. Schoch, W. Limmer, R. Sauer, Appl. Phys. Lett. 88 (2006) 051904. W. Limmer, K. Bitzer, R. Sauer, Physica E 21 (2004) 573. F. Matsukura, M. Sawicki, T. Dietl, D. Chiba, H. Ohno, Physica E 21 (2004) 1032. X. Liu, W. L. Lim, L. V. Titova, M. Dobrowolska, J. K. Furdyna, M. Kutrowski, T. Wojtowicz, J. Appl. Phys. 98 (2005) 063904. S. T. B. Goennenwein, S. Russo, A. F. Morpurgo, T. M. Klapwijk, W. Van Roy, J. De Boeck, Phys. Rev. B 71 (2005) 193306. W. Limmer, M. Glunk, J. Daeubler, T. Hummel, W. Schoch, C. Bihler, H. Huebl, M. S. Brandt, S. T. B. Goennenwein, R. Sauer, Microelectron. J., accepted for publication.
---
abstract: 'In this paper, we introduce a corresponding between bipartite graphs with a perfect matching and digraphs, which implicates an equivalent relation between the extendibility of bipartite graphs and the strongly connectivity of digraphs | , Phys. Rev. B 72 (2005) 165204. T. Omiya, F. Matsukura, A. Shen, Y. Ohno, H. Ohno, Physica E 10 (2001) 206. M. Reinwald, U. Wurstbauer, M. D[ö]{}ppe, W. Kipferl, K. Wagenhuber, H.-P. Tranitz, D. Weiss, W. Wegscheider, J. Cryst. Growth 278 (2005) 690. K. Y. Wang, K. W. Edmonds, L. X. Zhao, M. Sawicki, R. P. Campion, B. L. Gallagher, C. T. Foxon, Phys. Rev. B 72 (2005) 115207. J. Daeubler, M. Glunk, W. Schoch, W. Limmer, R. Sauer, Appl. Phys. Lett. 88 (2006) 051904. W. Limmer, K. Bitzer, R. Sauer, Physica E 21 (2004) 573. F. Matsukura, M. Sawicki, T. Dietl, D. Chiba, H. Ohno, Physica E 21 (2004) 1032. X. Liu, W. L. Lim, L. V. Titova, M. Dobrowolska, J. K. Furdyna, M. Kutrowski, T. Wojtowicz, J. Appl. Phys. 98 (2005) 063904. S. T. B. Goennenwein, S. Russo, A. F. Morpurgo, T. M. Klapwijk, W. Van Roy, J. De Boeck, Phys. Rev. B 71 (2005) 193306. W. Limmer, M. Glunk, J. Daeubler, T. Hummel, W. Schoch, C. Bihler, H. Huebl, M. S. Brandt, S. T. B. Goennenwein, R. Sauer, Microelectron. J., accepted for publication.
---
abstract:' In this paper, we insert a corresponding between bipartite graph with a perfect matching and digraphs, which implicate an equivalent relative between the extendibility of bipartite graphs and the strongly connectivity of digraph | , Phjs. Rev. B 72 (2005) 165204. T. Omiya, F. Mausukura, A. Shen, Y. Ohno, H. Mhno, Pgysica E 10 (2001) 206. M. Reinwald, U. Wurstbauer, L. E[ö]{}ppe, Q. Kipferl, K. Wagenhuber, H.-P. Traninz, D. Weisw, W. Xegscheider, J. Crbat. Growbk 278 (2005) 690. I. Y. Wcnj, K. W. Edmonds, L. X. Zhao, M. Vawicki, R. P. Cakpkou, B. L. Gallagher, C. T. Foxon, Phys. Rev. B 72 (2005) 115207. J. Dsekbler, M. Glunk, R. Scnjch, S. Limmer, R. Sauer, Appl. Phys. Lett. 88 (2006) 051904. W. Limmtr, K. Bitzer, R. Sauet, Physica E 21 (2004) 573. F. Matsukura, M. Swwicki, T. Dietl, D. Cjiba, H. Ohno, Phyfuca E 21 (2004) 1032. X. Lku, W. L. Lim, L. V. Titova, J. Dobrowolska, J. K. Furdyna, M. Kutfowskn, T. Wojtowixz, J. Wkpl. Phys. 98 (2005) 063904. W. T. B. Goennenwein, S. Russo, A. F. Motpurgo, T. M. Klaiwijk, W. Can Roy, J. De Boeck, Phbs. Rev. B 71 (2005) 193306. W. Limmer, M. Glunk, B. Baeubler, T. Hummel, W. Sxhich, C. Bihner, F. Hudbl, M. S. Brandh, S. T. B. Goennsnwein, R. Sayer, Microelectron. J., asbrpted for puglicatyog.
---
abstract: 'In this paper, we introduce a codresponding between bipqrtite graphs with a kerfect maeching and digraphs, which implicates an equivalend reletkon bctwedb hhe extendibility of bipartite graphs and the attokgly connectivitn of digraphs | , Phys. Rev. B 72 (2005) 165204. F. A. Shen, Ohno, H. Ohno, M. U. Wurstbauer, M. W. Kipferl, K. H.-P. Tranitz, D. Weiss, W. Wegscheider, Cryst. Growth 278 (2005) 690. K. Y. Wang, K. W. Edmonds, L. X. M. Sawicki, R. P. Campion, B. L. Gallagher, C. T. Foxon, Phys. Rev. 72 115207. Daeubler, Glunk, W. Schoch, W. Limmer, R. Sauer, Appl. Phys. Lett. 88 (2006) 051904. W. Limmer, K. R. Sauer, Physica E 21 (2004) 573. F. M. Sawicki, T. Dietl, Chiba, H. Ohno, Physica E (2004) X. Liu, L. L. Titova, M. Dobrowolska, K. Furdyna, M. Kutrowski, T. Wojtowicz, J. Appl. Phys. 98 (2005) 063904. S. T. B. Goennenwein, S. A. F. M. Klapwijk, Van J. Boeck, Phys. Rev. (2005) 193306. W. Limmer, M. Glunk, Hummel, W. Schoch, C. Bihler, H. Huebl, M. Brandt, S. B. Goennenwein, R. Sauer, Microelectron. J., for publication. --- abstract: 'In this paper, we a corresponding between bipartite graphs with a perfect matching and digraphs, which implicates an equivalent the extendibility of bipartite and the strongly of | , Phys. Rev. B 72 (2005) 165204. T. Omiya, F. Matsukura, A. shen, Y. Ohno, H. ohno, PHysIca e 10 (2001) 206. M. reinWald, u. Wurstbauer, M. D[ö]{}PPe, W. KIpferl, K. Wagenhuber, H.-P. TraNitz, D. weISs, W. WEGsCheidEr, J. CrysT. grOWTh 278 (2005) 690. K. y. WAnG, K. W. edMOnDs, L. X. ZHao, m. SawickI, R. P. Campion, b. L. GAlLagher, C. T. FoxoN, phYs. Rev. B 72 (2005) 115207. J. DaeUblEr, M. Glunk, W. SchOch, w. LimmeR, R. sauER, Appl. phyS. Lett. 88 (2006) 051904. w. LimmeR, k. BitzeR, R. Sauer, PhYsICa E 21 (2004) 573. F. MaTSukura, M. sAWiCki, T. dietl, D. Chiba, H. Ohno, PHYsICa E 21 (2004) 1032. X. Liu, W. L. Lim, L. V. titova, m. DOBrOWOlsKa, J. k. Furdyna, M. KUtRowskI, t. WojtowICz, j. aPPl. PHYs. 98 (2005) 063904. S. T. B. GoennenwEin, S. Russo, A. F. mOrpUrgo, T. M. klApwIJk, W. Van roy, J. DE BOEck, phys. Rev. B 71 (2005) 193306. W. LiMmer, m. Glunk, J. DaEubler, t. hummel, W. sChoch, C. BIhler, H. hueBl, M. s. BraNDt, s. T. b. GoEnNEnwEIn, r. SaUEr, MIcroelecTrOn. j., accePted FOR PUbliCatIon.
---
aBstraCt: 'In this paper, We iNtroDUce A corrEsponDing BeTween BipartIte grApHs with a perfect mAtchIng and digRapHs, WhiCh ImpliCAtes an EquIvaLent relAtion beTWeeN tHE EXtEndibility of bipartItE GRaPhs and thE stronGLy CoNNectivitY oF diGrapHS | , Phys. Rev. B 72 (2005) 1 65204. T.Omiya , F . M at suku ra,A. Shen, Y. Oh n o, H . Ohno, Physica E 10 ( 2001) 2 0 6. M . R einwa ld, U.W ur s t bau er ,M.D[ ö ]{ }ppe, W. Kipfer l, K. Wage nhu be r, H.-P. Tra n it z, D. Weis s,W. Wegscheid er, J. Cr ys t.G rowth 27 8 (20 05) 69 0 . K. Y . Wang, K .W . Edmo n ds, L.X . Z hao, M. Sawicki, R. P . C a mpion, B. L. G allagh er , C . T.Fox on, Phys.Re v. B7 2 (2005 ) 1 1 5 2 07. J. Daeubler,M. Glunk, W . Sc hoch,W. Li m mer, R . Sau er , Ap pl. Phys. L ett. 88 (2006 ) 0519 0 4. W. L i mmer, K . Bitz er, R. Sau e r, P hys ic a E2 1(20 0 4)573. F.Ma ts ukura , M. S a w icki , T . Di etl,D. Chiba, H.Ohn o, P h ysi ca E21 (2 004) 1 032.X. Liu , W.L. Lim, L. V. Tit ova, M. Dobro wol sk a,J. K. F u rdyna, M. Ku trowski , T. Wo j tow ic z , J. Appl. Phys. 98 (2 00 5 ) 0 63904. S . T. B . G oe n nenwein, S . R usso , A. F. Mor p ur go, T. M . Klap w ij k, W. Van R oy, J. D e B oec k, Ph y s. R ev. B71 (2005 ) 193 3 06. W. Limmer, M. Glunk, J.D ae u b le r , T. Hu mmel, W. Sc hoch , C.Bihl e r, H. Huebl , M.S. Br a ndt, S. T. B. Goenn en wein,R. Sa uer, Microele ctron. J., a c cepted f or p u bl i cation.
---
a bstra ct: 'In th i s paper, we i ntroduce a corres p o nding be twe enbip art i t egraphs with a p erfe ct matchi ngand dig rap hs, wh ich i mplicates an equi va le nt r ela tionb etween t he ex te ndi bilit y of bi parti te g ra ph s an d the s t ro n g ly c on ne ctiv ity o f dig raph s | , Phys._Rev. B_72 (2005) 165204. T._Omiya, F._Matsukura,_A. Shen,_Y._Ohno, H. Ohno,_Physica E 10_(2001) 206. M. Reinwald,_U. Wurstbauer, M._D[ö]{}ppe,_W. Kipferl, K. Wagenhuber, H.-P. Tranitz, D. Weiss, W. Wegscheider, J. Cryst. Growth 278_(2005)_690. K._Y._Wang,_K. W. Edmonds, L. X._Zhao, M. Sawicki, R. P._Campion, B._L. Gallagher, C. T. Foxon, Phys. Rev. B_72_(2005) 115207. J._Daeubler, M. Glunk, W. Schoch, W. Limmer, R. Sauer,_Appl. Phys. Lett. 88 (2006) 051904._W. Limmer, K._Bitzer,_R._Sauer, Physica E 21_(2004) 573. F. Matsukura, M. Sawicki,_T. Dietl, D. Chiba, H. Ohno,_Physica E 21 (2004) 1032. X. Liu,_W. L. Lim, L. V. Titova,_M. Dobrowolska, J. K. Furdyna,_M. Kutrowski,_T. Wojtowicz, J. Appl. Phys._98 (2005) 063904._S. T._B. Goennenwein, S._Russo, A. F. Morpurgo, T. M._Klapwijk, W. Van_Roy, J. De Boeck, Phys. Rev._B_71 (2005) 193306._W._Limmer,_M. Glunk,_J. Daeubler, T._Hummel,_W. Schoch,_C._Bihler, H. Huebl, M. S. Brandt,_S._T. B. Goennenwein, R. Sauer, Microelectron. J.,_accepted for publication.
---
abstract:_'In_this paper, we introduce_a corresponding between bipartite graphs_with a perfect matching and digraphs,_which implicates_an equivalent_relation between the extendibility of bipartite graphs and the strongly connectivity_of digraphs |
_s\vert+L\vert \bar{Z}^n_s\vert +L\vert Z_s\vert\in M^2_G(0,T) $. Thus with the help of Theorem 4.7 in [@HWZ2016], we get that $$\lim_{n\rightarrow\infty}\mathbb{\hat{E}}[\int_{0}^{T}(|f_0(s)|+2L\vert Y_s\vert+L\vert \bar{Z}^n_s\vert +L\vert Z_s\vert )^2\mathbf{1}_{\vert \bar{Z}^n_s-Z_s\vert > \delta} ds]=0.$$
Consequently, putting together the above two inequalities we deduce that $$\limsup_{n\rightarrow\infty}\mathbb{\hat{E}}[\int_{0}^{T}\vert \bar{f}_n(s,\bar{Y}^n_s, \bar{Z}^n_s)+f_0(s)-f(s,{Y}_s,{Z}_s) \vert^{2}ds] \leq 2 T \varepsilon^2.$$ Letting $\varepsilon\rightarrow 0$, we get the desired result.
[For a $1$-dimensional $G$-Brownian motion $B$ with $\underline{\sigma}^2:= -\mathbb{\hat{E}}[-|B_1|^2]$, consider the following $G$-BSDE: $$Y_t=\frac{1}{6}|B_T|^6-\frac{5}{2}\underline{\sigma}^2\int^T_t|Z_s|^{\frac{4}{5}}ds-\int^T_t Z_sdB_s-(K_T-K_t).$$ Note that $f(z)=-\frac{5}{2}\underline{\sigma}^2|z|^{\frac{4}{5}}$ is a uniformly continuous function. Then by $G$-Itô’s formula and Theorem \[main\], it is easy to check that $(\frac{1}{6}|B_t|^6,(B_t)^5,\frac{5}{2}\underline{\sigma}^2\int^t_0 |B_s|^4ds- \frac{5}{2}\int^t_0 |B_s|^4d\langle B\rangle_s)$ is the unique $\ | _ s\vert+L\vert \bar{Z}^n_s\vert + L\vert Z_s\vert\in M^2_G(0,T) $. Thus with the help of Theorem 4.7 in [ @HWZ2016 ], we get that $ $ \lim_{n\rightarrow\infty}\mathbb{\hat{E}}[\int_{0}^{T}(|f_0(s)|+2L\vert Y_s\vert+L\vert \bar{Z}^n_s\vert + L\vert Z_s\vert) ^2\mathbf{1}_{\vert \bar{Z}^n_s - Z_s\vert > \delta } ds]=0.$$
Consequently, putting in concert the above two inequality we deduce that $ $ \limsup_{n\rightarrow\infty}\mathbb{\hat{E}}[\int_{0}^{T}\vert \bar{f}_n(s,\bar{Y}^n_s, \bar{Z}^n_s)+f_0(s)-f(s,{Y}_s,{Z}_s) \vert^{2}ds ] \leq 2 thyroxine \varepsilon^2.$$ Letting $ \varepsilon\rightarrow 0 $, we get the hope result.
[ For a $ 1$-dimensional $ G$-Brownian motion $ B$ with $ \underline{\sigma}^2:= -\mathbb{\hat{E}}[-|B_1|^2]$, regard the trace $ G$-BSDE: $ $ Y_t=\frac{1}{6}|B_T|^6-\frac{5}{2}\underline{\sigma}^2\int^T_t|Z_s|^{\frac{4}{5}}ds-\int^T_t Z_sdB_s-(K_T - K_t).$$ Note that $ f(z)=-\frac{5}{2}\underline{\sigma}^2|z|^{\frac{4}{5}}$ is a uniformly continuous function. Then by $ G$-Itô ’s recipe and Theorem \[main\ ], it is easy to check that $ (\frac{1}{6}|B_t|^6,(B_t)^5,\frac{5}{2}\underline{\sigma}^2\int^t_0 |B_s|^4ds- \frac{5}{2}\int^t_0 |B_s|^4d\langle B\rangle_s)$ is the singular $ \ | _s\vegt+L\vert \bar{Z}^n_s\vert +L\vert Z_s\vert\in M^2_G(0,T) $. Jhys witi the hslp of Tfeorem 4.7 in [@HWZ2016], we get that $$\lmm_{n\rughtaerow\infty}\mathbb{\hat{E}}[\int_{0}^{G}(|f_0(s)|+2L\vert J_s\vert+L\veet \ber{Z}^n_s\vert +L\vert V_a\vert )^2\mathbf{1}_{\vsvt \bax{Z}^i_s-Z_s\vert > \delta} ds]=0.$$
Consequantly, putting doeecher the above two inequalities we dqduce tnah $$\limsup_{n\rightwrroe\ynftg}\mathbb{\hat{E}}[\int_{0}^{T}\vert \bar{f}_n(s,\bar{Y}^n_s, \bzr{Z}^n_s)+f_0(s)-h(s,{Y}_s,{Z}_s) \vert^{2}ds] \lrq 2 T \varepsilon^2.$$ Letting $\vwrepdilon\rightarrow 0$, wf get the dgairqe result.
[For x $1$-dimensional $G$-Browniah motion $B$ with $\underline{\sigma}^2:= -\oathby{\hat{E}}[-|B_1|^2]$, consudwr hve followinj $G$-BSDV: $$Y_t=\frac{1}{6}|B_T|^6-\frag{5}{2}\lnderlite{\sigma}^2\ont^T_t|Z_s|^{\frac{4}{5}}ds-\ikt^T_t V_sdB_w-(K_T-K_t).$$ Note that $f(z)=-\frar{5}{2}\underline{\sigma}^2|z|^{\frac{4}{5}}$ is a unixoxmly continuous functuob. Thet by $G$-Itô’r fofmuma ahd Thelrej \[main\], it js easy to xheck that $(\frac{1}{6}|B_t|^6,(B_t)^5,\grwb{5}{2}\inderline{\sigja}^2\int^t_0 |B_f|^4ds- \frac{5}{2}\int^t_0 |B_s|^4d\langle B\rangle_s)$ is the lniqhe $\ | _s\vert+L\vert \bar{Z}^n_s\vert +L\vert Z_s\vert\in M^2_G(0,T) $. Thus help Theorem 4.7 [@HWZ2016], we get Z_s\vert \bar{Z}^n_s-Z_s\vert > \delta} Consequently, putting together above two inequalities we deduce that \bar{f}_n(s,\bar{Y}^n_s, \bar{Z}^n_s)+f_0(s)-f(s,{Y}_s,{Z}_s) \vert^{2}ds] \leq 2 T \varepsilon^2.$$ Letting $\varepsilon\rightarrow 0$, we get the result. [For a $1$-dimensional $G$-Brownian motion $B$ with $\underline{\sigma}^2:= -\mathbb{\hat{E}}[-|B_1|^2]$, consider the following $$Y_t=\frac{1}{6}|B_T|^6-\frac{5}{2}\underline{\sigma}^2\int^T_t|Z_s|^{\frac{4}{5}}ds-\int^T_t Note $f(z)=-\frac{5}{2}\underline{\sigma}^2|z|^{\frac{4}{5}}$ a uniformly continuous function. Then by $G$-Itô’s formula and Theorem \[main\], it is easy to check $(\frac{1}{6}|B_t|^6,(B_t)^5,\frac{5}{2}\underline{\sigma}^2\int^t_0 |B_s|^4ds- \frac{5}{2}\int^t_0 |B_s|^4d\langle B\rangle_s)$ is the unique | _s\vert+L\vert \bar{Z}^n_s\vert +L\vert z_s\vert\in M^2_G(0,t) $. Thus WitH thE hElp oF TheOrem 4.7 in [@HWZ2016], we get THat $$\lIm_{n\rightarrow\infty}\mathBb{\hat{e}}[\iNT_{0}^{T}(|f_0(s)|+2l\VeRt Y_s\vErt+L\verT \BaR{z}^N_s\vErT +L\VerT Z_S\VeRt )^2\matHbf{1}_{\Vert \bar{z}^n_s-Z_s\vert > \dEltA} dS]=0.$$
ConsequentlY, PuTting togetHer The above two iNeqUalitiEs We dEDuce tHat $$\LimsuP_{n\righTArrow\iNfty}\mathbB{\hAT{E}}[\int_{0}^{T}\VErt \bar{f}_N(S,\BaR{Y}^n_s, \Bar{Z}^n_s)+f_0(s)-f(s,{Y}_s,{Z}_s) \verT^{2}Ds] \LEq 2 T \varepsilon^2.$$ LEtting $\VaREpSILon\RigHtarrow 0$, we gEt The deSIred resULt.
[fOR A $1$-diMEnsional $G$-BrowNian motion $B$ WIth $\UnderlInE{\siGMa}^2:= -\mathBb{\hat{e}}[-|B_1|^2]$, COnsIder the follOwinG $G$-BSDE: $$Y_t=\fRac{1}{6}|B_T|^6-\fRAc{5}{2}\underLIne{\sigmA}^2\int^T_t|z_s|^{\fRac{4}{5}}Ds-\inT^t_t z_sDB_s-(k_T-k_T).$$ NoTE tHat $F(Z)=-\frAc{5}{2}\underlInE{\sIgma}^2|z|^{\Frac{4}{5}}$ IS A UNifoRmlY conTinuoUs function. TheN by $g$-Itô’S ForMula aNd TheOrem \[MaIn\], it iS easy tO checK tHat $(\frac{1}{6}|B_t|^6,(B_t)^5,\frac{5}{2}\UndeRline{\sigmA}^2\inT^t_0 |b_s|^4dS- \fRac{5}{2}\inT^T_0 |B_s|^4d\laNglE B\rAngle_s)$ iS the uniQUe $\ | _s\vert+L\vert \bar{Z}^n_s \vert +L\v ert Z _s\ ver t\ in M ^2_G (0,T) $. Thusw iththe help of Theorem 4. 7 in[@ H WZ20 1 6] , weget tha t $ $ \ lim _{ n\ rig ht a rr ow\in fty }\mathb b{\hat{E}} [\i nt _{0}^{T}(|f_ 0 (s )|+2L\vert Y_ s\vert+L\ver t \ bar{Z} ^n _s\ v ert + L\v ert Z _s\ver t )^2\m athbf{1}_ {\ v ert \b a r{Z}^n_ s - Z_ s\ve rt > \delta} ds]= 0 .$ $
Consequently , putt in g t o g eth erthe abovetw o ine q ualitie s w e d edu c e that $$\lim sup_{n\righ t arr ow\inf ty }\m a thbb{\ hat{E }} [ \in t_{0}^{T}\v ert\bar{f}_n (s,\ba r {Y}^n_s , \bar{Z }^n_s) +f_ 0(s )-f( s ,{ Y} _s, {Z } _s) \v ert ^ {2} ds] \leq 2 T \var epsi l o n ^ 2.$$ Le ttin g $\v arepsilon\rig hta rrow 0$, we g et th e de si red r esult.
[Fo ra $1$-dimension al $ G$-Browni anmo tio n$B$ w i th $\u nde rli ne{\sig ma}^2:= -\m at h b b {\ hat{E}}[-|B_1|^2]$ ,c o ns ider the follo w in g$ G$-BSDE: $ $Y_ t=\f r a c{1}{ 6}|B _ T| ^6-\frac {5}{2} \ un de rline{\ si gma}^2 \i nt^ T_t |Z_s| ^ {\fr ac{4}{ 5}}ds-\i nt^T_ t Z_sdB_s-(K_T- K _t).$$ Note t h at $ f( z )=-\ fra c{5}{2}\und erli n e{\s igma } ^2 |z| ^ {\fra c{4}{ 5} } $i s a uniformly conti nu ous fu nctio n. Then by $G $-Itô’s fo r m u la and T heor e m\ [main\], it is easy to checkt hat $(\f rac{1 }{6}|B_t |^6,(B_t) ^ 5 ,\frac{5 }{2 }\u nde rli n e {\ sigma}^2\int^ t _ 0 |B _s |^4ds-\fr ac{5}{2 }\i nt^ t_0 |B _s |^4d\lang le B\ran gl e_ s) $isthe u n ique $\ | _s\vert+L\vert \bar{Z}^n_s\vert_+L\vert Z_s\vert\in_M^2_G(0,T) $. Thus with_the help_of_Theorem 4.7_in_[@HWZ2016], we get_that $$\lim_{n\rightarrow\infty}\mathbb{\hat{E}}[\int_{0}^{T}(|f_0(s)|+2L\vert Y_s\vert+L\vert_\bar{Z}^n_s\vert +L\vert Z_s\vert )^2\mathbf{1}_{\vert_\bar{Z}^n_s-Z_s\vert > \delta}_ds]=0.$$
Consequently,_putting together the above two inequalities we deduce that $$\limsup_{n\rightarrow\infty}\mathbb{\hat{E}}[\int_{0}^{T}\vert \bar{f}_n(s,\bar{Y}^n_s, \bar{Z}^n_s)+f_0(s)-f(s,{Y}_s,{Z}_s) \vert^{2}ds] \leq_2_T \varepsilon^2.$$_Letting_$\varepsilon\rightarrow_0$, we get the desired_result.
[For a $1$-dimensional $G$-Brownian motion_$B$ with_$\underline{\sigma}^2:= -\mathbb{\hat{E}}[-|B_1|^2]$, consider the following $G$-BSDE: $$Y_t=\frac{1}{6}|B_T|^6-\frac{5}{2}\underline{\sigma}^2\int^T_t|Z_s|^{\frac{4}{5}}ds-\int^T_t Z_sdB_s-(K_T-K_t).$$_Note_that $f(z)=-\frac{5}{2}\underline{\sigma}^2|z|^{\frac{4}{5}}$ is_a uniformly continuous function. Then by $G$-Itô’s formula and_Theorem \[main\], it is easy to_check that $(\frac{1}{6}|B_t|^6,(B_t)^5,\frac{5}{2}\underline{\sigma}^2\int^t_0_|B_s|^4ds-_\frac{5}{2}\int^t_0_|B_s|^4d\langle B\rangle_s)$ is the_unique $\ |
HS_{2,lip,j}$$ The point of this splitting is that we will need to use a more complicated procedure to estimate $w_{2,lip,0}$, since too many logarithmic losses in estimates are insufficient for our purposes. We have $$|WRHS_{2,lip,1}(t,r)| \leq \frac{C ||e_{1}-e_{2}||_{X} \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\}} \lambda_{0}(t)^{4}}{t^{2} \log^{\delta-\delta_{2}}(t) \log^{2b}(t) (g_{0}(t)^{2}+r^{2})^{2}}$$ Using the analog of, and a similar procedure used to estimate various integrals arising in the $w_{2}$ estimates above, we get $$|w_{2,lip,1}(t,r)| \leq \frac{C ||e_{1}-e_{2}||_{X} \lambda_{0}(t)^{4} \log(t)}{t^{2} \log^{\delta-\delta_{2}}(t) \log^{2b}(t) g_{0}(t)^{2}}, \quad r >0$$ In particular, the procedure used to estimate $w_{2,lip,1}$ does not involve any derivatives of $WRHS_{2,lip,1}$, which is why we did not need to prove any estimates on derivatives of $v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$. (Note that $v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$ arises in some terms of $WRHS_{2,lip,1}$). Next, we note that $$\begin{split}&|\partial_{r}WRHS_{2,lip,0}(t,r)| + \frac{|WRHS_{2,lip,0}(t,r)|}{r} \\
&\leq C \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\}} \left(\frac{r \lambda_{0}(t)^{2} ||e_{1}-e_{2}||_{X}}{t^{2} \log^{\delta-\delta_{2}}(t) (r^{2}+\lambda_{0}(t)^{2})^{2}}\right)\left(\frac{1}{\log^{b}(t)}+\frac{1}{\log(t)}\ | HS_{2,lip, j}$$ The point of this splitting is that we will need to use a more complicated procedure to calculate $ w_{2,lip,0}$, since excessively many logarithmic losses in estimates are insufficient for our determination. We own $ $ |WRHS_{2,lip,1}(t, r)| \leq \frac{C ||e_{1}-e_{2}||_{X } \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\ } } \lambda_{0}(t)^{4}}{t^{2 } \log^{\delta-\delta_{2}}(t) \log^{2b}(t) (g_{0}(t)^{2}+r^{2})^{2}}$$ Using the analog of, and a alike procedure used to calculate diverse integrals arise in the $ w_{2}$ estimates above, we get $ $ |w_{2,lip,1}(t, r)| \leq \frac{C ||e_{1}-e_{2}||_{X } \lambda_{0}(t)^{4 } \log(t)}{t^{2 } \log^{\delta-\delta_{2}}(t) \log^{2b}(t) g_{0}(t)^{2 } }, \quad gas constant > 0$$ In especial, the procedure used to estimate $ w_{2,lip,1}$ does not involve any derivatives of $ WRHS_{2,lip,1}$, which is why we did not necessitate to prove any estimates on derivatives of $ v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$. (notice that $ v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$ arises in some terms of $ WRHS_{2,lip,1}$). Next, we notice that $ $ \begin{split}&|\partial_{r}WRHS_{2,lip,0}(t, r)| + \frac{|WRHS_{2,lip,0}(t, r)|}{r } \\
& \leq C \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\ } } \left(\frac{r \lambda_{0}(t)^{2 } ||e_{1}-e_{2}||_{X}}{t^{2 } \log^{\delta-\delta_{2}}(t) (r^{2}+\lambda_{0}(t)^{2})^{2}}\right)\left(\frac{1}{\log^{b}(t)}+\frac{1}{\log(t)}\ | HS_{2,llp,j}$$ The point of this spuitting is that we wiln need to use x more complicated procedure ti estumate $w_{2,lip,0}$, since too mxny logarpthmic lowses un estimatxa are ikfuffjgient hor our purposex. We have $$|FRHS_{2,lip,1}(t,r)| \leq \xrxc{E ||e_{1}-e_{2}||_{X} \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\}} \lambda_{0}(t)^{4}}{t^{2} \ljg^{\delta-\cepta_{2}}(t) \log^{2b}(t) (g_{0}(t)^{2}+r^{2})^{2}}$$ Usimd ths analog of, and a similar procedurs used uo estimate varioux integrals arising in the $w_{2}$ edtimates above, we het $$|w_{2,lip,1}(t,r)| \oeq \seac{C ||e_{1}-e_{2}||_{X} \lamcda_{0}(t)^{4} \log(t)}{t^{2} \log^{\delta-\delja_{2}}(t) \log^{2b}(t) g_{0}(t)^{2}}, \quad r >0$$ In particulxr, thz procedure uwed jo estimate $x_{2,lip,1}$ djes not involve any gerivatoves of $WRHS_{2,lii,1}$, whirh iw why we did not need to prove any estiiates on gexivatives of $v_{c}^{\lambda_{1}}-c_{c}^{\oambdd_{2}}$. (Node tfqt $x_{c}^{\lzmuda_{1}}-b_{c}^{\lambfa_{2}}$ erises in skme terms od $WRHS_{2,lip,1}$). Next, we nptq that $$\begin{spmit}&|\pareiwl_{r}WRHS_{2,lip,0}(t,r)| + \frac{|WRHS_{2,lip,0}(t,r)|}{r} \\
&\leq C \mathtbm{1}_{\{d \geq \frac{g_{0}(t)}{4}\}} \left(\frac{r \oambda_{0}(t)^{2} ||e_{1}-e_{2}||_{X}}{t^{2} \log^{\delta-\felta_{2}}(t) (r^{2}+\lwmbda_{0}(t)^{2})^{2}}\right)\left(\frac{1}{\log^{b}(t)}+\frac{1}{\log(t)}\ | HS_{2,lip,j}$$ The point of this splitting is will to use more complicated procedure many losses in estimates insufficient for our We have $$|WRHS_{2,lip,1}(t,r)| \leq \frac{C ||e_{1}-e_{2}||_{X} \geq \frac{g_{0}(t)}{4}\}} \lambda_{0}(t)^{4}}{t^{2} \log^{\delta-\delta_{2}}(t) \log^{2b}(t) (g_{0}(t)^{2}+r^{2})^{2}}$$ Using the analog of, and a similar used to estimate various integrals arising in the $w_{2}$ estimates above, we get \leq ||e_{1}-e_{2}||_{X} \log(t)}{t^{2} \log^{2b}(t) g_{0}(t)^{2}}, \quad r >0$$ In particular, the procedure used to estimate $w_{2,lip,1}$ does not involve derivatives of $WRHS_{2,lip,1}$, which is why we did need to prove any on derivatives of $v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$. (Note $v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$ in some of Next, note that $$\begin{split}&|\partial_{r}WRHS_{2,lip,0}(t,r)| \frac{|WRHS_{2,lip,0}(t,r)|}{r} \\ &\leq C \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\}} \left(\frac{r \lambda_{0}(t)^{2} ||e_{1}-e_{2}||_{X}}{t^{2} \log^{\delta-\delta_{2}}(t) (r^{2}+\lambda_{0}(t)^{2})^{2}}\right)\left(\frac{1}{\log^{b}(t)}+\frac{1}{\log(t)}\ | HS_{2,lip,j}$$ The point of this splitTing is that We wilL neEd tO uSe a mOre cOmplicated procEDure To estimate $w_{2,lip,0}$, since too Many lOgARithMIc LosseS in estiMAtES Are InSuFfiCiENt For ouR puRposes. WE have $$|WRHS_{2,lIp,1}(t,R)| \lEq \frac{C ||e_{1}-e_{2}||_{X} \maTHbBm{1}_{\{r \geq \frac{G_{0}(t)}{4}\}} \lAmbda_{0}(t)^{4}}{t^{2} \log^{\deLta-\Delta_{2}}(t) \LoG^{2b}(t) (G_{0}(T)^{2}+r^{2})^{2}}$$ UsiNg tHe anaLog of, aND a simiLar procedUrE Used to EStimate VARiOus iNtegrals arising in THe $W_{2}$ Estimates above, We get $$|w_{2,LiP,1}(T,r)| \LEQ \frAc{C ||E_{1}-e_{2}||_{X} \lambda_{0}(t)^{4} \LoG(t)}{t^{2} \loG^{\Delta-\deLTa_{2}}(T) \LOG^{2b}(t) G_{0}(T)^{2}}, \quad r >0$$ In partiCular, the proCEduRe used To EstIMate $w_{2,lIp,1}$ doeS nOT inVolve any derIvatIves of $WRHs_{2,lip,1}$, whICh is why WE did not Need to ProVe aNy esTImAtEs oN dERivATiVes OF $v_{c}^{\Lambda_{1}}-v_{c}^{\LaMbDa_{2}}$. (NotE thaT $V_{C}^{\LAmbdA_{1}}-v_{c}^{\LambDa_{2}}$ ariSes in some termS of $wRHS_{2,LIp,1}$). NExt, we Note tHat $$\bEgIn{splIt}&|\partIal_{r}WrHs_{2,lip,0}(t,r)| + \frac{|WRHS_{2,lIp,0}(t,r)|}{R} \\
&\leq C \mathBbm{1}_{\{R \gEq \fRaC{g_{0}(t)}{4}\}} \leFT(\frac{r \LamBda_{0}(T)^{2} ||e_{1}-e_{2}||_{X}}{t^{2} \loG^{\delta-\dELta_{2}}(T) (r^{2}+\LAMBdA_{0}(t)^{2})^{2}}\right)\left(\frac{1}{\log^{B}(t)}+\FRAc{1}{\Log(t)}\ | HS_{2,lip,j}$$ The point o f this spl ittin g i s t ha t we wil l need to usea mor e complicated procedur e toes t imat e $ w_{2, lip,0}$ , s i n ceto oman yl og arith mic losses in estima tes a re insuffici e nt for our p urp oses. We hav e $ $|WRHS _{ 2,l i p,1}( t,r )| \l eq \fr a c{C || e_{1}-e_{ 2} | |_{X}\ mathbbm { 1 }_ {\{r \geq \frac{g_{0} ( t) } {4}\}} \lambda _{0}(t )^ { 4} } { t^{ 2}\log^{\del ta -\del t a_{2}}( t )\ l o g^{ 2 b}(t) (g_{0}( t)^{2}+r^{2 } )^{ 2}}$$Us ing the an alogof , an d a similar pro cedure us ed toe stimate various integ ral s a risi n gin th e$ w_{ 2 }$ es t ima tes abov e, w e get $$| w _ { 2 ,lip ,1} (t,r )| \l eq \frac{C || e_{ 1}-e _ {2} ||_{X } \la mbda _{ 0}(t) ^{4} \ log(t )} {t^{2} \log^{\d elta -\delta_{ 2}} (t ) \ lo g^{2b } (t) g_ {0} (t) ^{2}},\quad r >0$ $I n pa rticular, the proc ed u r eused toestima t e$w _ {2,lip,1 }$ do es n o t invo lvea ny derivat ives o f $ WR HS_{2,l ip ,1}$,wh ich is whyw e di d notneed toprove any estimateso n derivatives of $ v_ { c}^{ \la mbda_{1}}-v _{c} ^ {\la mbda _ {2 }}$ . (Not e tha t$ v_ { c}^{\lambda_{1}}-v_ {c }^{\la mbda_ {2}}$ arisesin some te r m s of $WRH S_{2 , li p ,1}$). Next, w e not e that $$\ b egin{spl it}&| \partial _{r}WRHS_ { 2 ,lip,0}( t,r )|+ \ fra c { |W RHS_{2,lip,0} ( t ,r)| }{ r} \\
& \le q C \ma thb bm{ 1}_ {\{ r\geq \fra c{g_{0}( t) }{ 4} \} } \ left( \ frac{r \ la mbd a_ {0} (t)^{ 2 } ||e_ {1}-e _{2} || _{ X }}{ t^{2} \ l og ^ { \del ta -\ delt a_{ 2} }(t)(r^{ 2 }+\ lambda_ {0}(t)^{2 })^ { 2}}\ ri gh t)\left (\frac{1}{\lo g^ {b}(t)}+\f ra c{1 }{\log ( t )}\ | HS_{2,lip,j}$$ The_point of_this splitting is that_we will_need_to use_a_more complicated procedure_to estimate $w_{2,lip,0}$,_since too many logarithmic_losses in estimates_are_insufficient for our purposes. We have $$|WRHS_{2,lip,1}(t,r)| \leq \frac{C ||e_{1}-e_{2}||_{X} \mathbbm{1}_{\{r \geq \frac{g_{0}(t)}{4}\}} \lambda_{0}(t)^{4}}{t^{2}_\log^{\delta-\delta_{2}}(t)_\log^{2b}(t) (g_{0}(t)^{2}+r^{2})^{2}}$$_Using_the_analog of, and a similar_procedure used to estimate various_integrals arising_in the $w_{2}$ estimates above, we get $$|w_{2,lip,1}(t,r)|_\leq_\frac{C ||e_{1}-e_{2}||_{X} \lambda_{0}(t)^{4}_\log(t)}{t^{2} \log^{\delta-\delta_{2}}(t) \log^{2b}(t) g_{0}(t)^{2}}, \quad r >0$$ In particular,_the procedure used to estimate $w_{2,lip,1}$_does not involve_any_derivatives_of $WRHS_{2,lip,1}$, which is_why we did not need to_prove any estimates on derivatives of_$v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$. (Note that $v_{c}^{\lambda_{1}}-v_{c}^{\lambda_{2}}$ arises in some_terms of $WRHS_{2,lip,1}$). Next, we note_that $$\begin{split}&|\partial_{r}WRHS_{2,lip,0}(t,r)| + \frac{|WRHS_{2,lip,0}(t,r)|}{r} \\
&\leq_C \mathbbm{1}_{\{r_\geq \frac{g_{0}(t)}{4}\}} \left(\frac{r \lambda_{0}(t)^{2} ||e_{1}-e_{2}||_{X}}{t^{2}_\log^{\delta-\delta_{2}}(t) (r^{2}+\lambda_{0}(t)^{2})^{2}}\right)\left(\frac{1}{\log^{b}(t)}+\frac{1}{\log(t)}\ |
{\boldsymbol{\mathfrak A}}}_g'\to{\overline{\boldsymbol{\mathfrak A}}}_g$ which is one-to-one on geometric points ([@Alexeev:SQAV], Theorem 4.5). Alexeev remarks that in general this is not an isomorphism.
Again, we denote the universal family by $({\overline{\boldsymbol{\mathfrak A}}}_{g,1},{{\overline{\tilde{\boldsymbol{\varTheta}}}}}) \to {\overline{\boldsymbol{\mathfrak A}}}_g$. We denote by ${\overline{\boldsymbol{\mathfrak A}}}_{g,n}$ the fibered power $({\overline{\boldsymbol{\mathfrak A}}}_{g,1})^n_{{\overline{\boldsymbol{\mathfrak A}}}_g}$. (This is, to some extent, in analogy with the space of stable pointed curves ${\overline{{\mathcal M}}}_{g,n}$, although we do not use Knudsen’s stabilization.) Denote by $p_i:{\overline{\boldsymbol{\mathfrak A}}}_{g,n}\to{\overline{\boldsymbol{\mathfrak A}}}_{g,1}$ the projection to the $i$-th factor. We have a natural relatively ample divisor ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n \subset {\overline{\boldsymbol{\mathfrak A}}}_{g,n}$ defined by ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n=\sum_i p_i^*{{\overline{\tilde{\boldsymbol{\varTheta}}}}}$.
We denote by ${\overline{\mathbf A}}_{g,n}$ the coarse moduli spaces of ${\overline{\boldsymbol{\mathfrak A}}}_{g,n}$, and by ${{\overline{\boldsymbol{\Theta}}}}_n \subset {\overline{\mathbf A}}_{g,n}$ the image of ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$. A-priori these are Artin algebraic spaces (see [@Keel-Mori]), but since some multiple $m{{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$ descends to a Cartier divisor on ${\overline{\mathbf A}}_{g,n}$ and is relatively ample, these are projective schemes over ${\operatorname{Spec}}{{\mathbb{Z}}}$.
Properties of stable pairs
--------------------------
We now collect a few properties of stable principally polarized quasi-abelian schemes, which we will use in the next section.
To save words, we will refer to a stable principally polarized quasi-abelian scheme $(P, \Theta)$ (always assumed flat over a base scheme $S$) as a [*stable pair*]{}.
| { \boldsymbol{\mathfrak A}}}_g'\to{\overline{\boldsymbol{\mathfrak A}}}_g$ which is one - to - one on geometric points ([ @Alexeev: SQAV ], Theorem 4.5). Alexeev remarks that in general this is not an isomorphism.
Again, we denote the cosmopolitan syndicate by $ ({ \overline{\boldsymbol{\mathfrak A}}}_{g,1},{{\overline{\tilde{\boldsymbol{\varTheta } } } } }) \to { \overline{\boldsymbol{\mathfrak A}}}_g$. We denote by $ { \overline{\boldsymbol{\mathfrak A}}}_{g, n}$ the fibered power $ ({ \overline{\boldsymbol{\mathfrak A}}}_{g,1})^n_{{\overline{\boldsymbol{\mathfrak A}}}_g}$. (This is, to some extent, in doctrine of analogy with the space of static pointed curves $ { \overline{{\mathcal M}}}_{g, n}$, although we do not use Knudsen ’s stabilization .) announce by $ p_i:{\overline{\boldsymbol{\mathfrak A}}}_{g, n}\to{\overline{\boldsymbol{\mathfrak A}}}_{g,1}$ the projection to the $ i$-th factor. We have a lifelike relatively ample divisor $ { { \overline{\tilde{\boldsymbol{\varTheta}}}}}_n \subset { \overline{\boldsymbol{\mathfrak A}}}_{g, n}$ define by $ { { \overline{\tilde{\boldsymbol{\varTheta}}}}}_n=\sum_i p_i^*{{\overline{\tilde{\boldsymbol{\varTheta}}}}}$.
We denote by $ { \overline{\mathbf A}}_{g, n}$ the coarse modulus spaces of $ { \overline{\boldsymbol{\mathfrak A}}}_{g, n}$, and by $ { { \overline{\boldsymbol{\Theta}}}}_n \subset { \overline{\mathbf A}}_{g, n}$ the image of $ { { \overline{\tilde{\boldsymbol{\varTheta}}}}}_n$. A - priori these are Artin algebraic space (see [ @Keel - Mori ]), but since some multiple $ m{{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$ descends to a Cartier divisor on $ { \overline{\mathbf A}}_{g, n}$ and is relatively ample, these are projective schemes over $ { \operatorname{Spec}}{{\mathbb{Z}}}$.
Properties of stable pairs
--------------------------
We now roll up a few properties of stable principally polarize quasi - abelian schemes, which we will use in the next part.
To save words, we will refer to a stable principally polarize quasi - abelian scheme $ (P, \Theta)$ (always assumed flat over a base scheme $ S$) as a [ * stable pair * ] { }. | {\bolfsymbol{\mathfrak A}}}_g'\to{\overuine{\boldsymbol{\mcrhfrak A}}}_g$ whjch is ove-to-one on geometric points ([@Elexwev:SQQV], Theorem 4.5). Alexeev reoarks than in geneeal uhis is not an isomorphism.
Again, ae dznite the univerxal family by $({\overline{\bondrylbol{\mathfrak A}}}_{g,1},{{\overline{\tilde{\boldsymfol{\varTneha}}}}}) \to {\overline{\foldxrmbom{\mathfrak A}}}_g$. We denote by ${\overline{\goldsymuol{\mathfrak A}}}_{g,n}$ the fibered power $({\overlinf{\bolfsymbol{\mathfrak A}}}_{g,1})^j_{{\overline{\boodsyivol{\mathfrak X}}}_g}$. (This is, to some extgnt, in analogy with the space of stabke pointed cyrvfv ${\overline{{\methcal M}}}_{g,n}$, althougm we do not usr Knudsen’s stanilizetiob.) Denote by $p_i:{\overlinx{\boldsymbol{\mathfrak W}}}_{g,n}\to{\overniue{\boldsymbol{\mathfrak Q}}}_{g,1}$ the krojewtiov to tht $i$-vh ractor. We have a nafural relatuvely ample divisor ${{\odvtline{\tilde{\bomdsymbjl{\darTheta}}}}}_n \subset {\overline{\boldsymbol{\mathxrai A}}}_{g,n}$ defined by ${{\overlibe{\tilde{\boldsymbol{\varTjeta}}}}}_n=\sum_i p_i^*{{\overline{\tilde{\boldsymbol{\varTheta}}}}}$.
We denote by ${\ovarlinx{\mxthyn X}}_{t,n}$ the coarse moduli spaces of ${\overline{\boldsymbjm{\mstmfrak A}}}_{g,n}$, and by ${{\overline{\bokddykfol{\Theta}}}}_n \subret {\ovzdljne{\mathbf A}}_{g,n}$ the lmage os ${{\oveeline{\tildt{\boldxymbol{\varTheta}}}}}_n$. A-priori thewe are Artin qlgebraic spaces (sze [@Keel-Mori]), yut simce spme multiple $m{{\overline{\tnlde{\bomdsymbol{\varHheta}}}}}_n$ deazends to a Cartidr cieisor on ${\overline{\mathbf A}}_{g,g}$ and is celatnvely amole, jhese awe projectlve sgvemes over ${\operatogname{Dpac}}{{\mathbb{Z}}}$.
Pgoperties of stable pairs
--------------------------
We now collect a few psopvrties of stabke principallr polarized qucsi-abelicn schdmes, which we wiln use in thq next sectiot.
Jo save words, we will refwr ti a stacue principally polarizeb quawi-abelian scheme $(P, \Thgtz)$ (always assumeb dlat over a basr szheie $S$) af a [*stable pahr*]{}.
| {\boldsymbol{\mathfrak A}}}_g'\to{\overline{\boldsymbol{\mathfrak A}}}_g$ which is one-to-one on ([@Alexeev:SQAV], 4.5). Alexeev that in general Again, denote the universal by $({\overline{\boldsymbol{\mathfrak A}}}_{g,1},{{\overline{\tilde{\boldsymbol{\varTheta}}}}}) {\overline{\boldsymbol{\mathfrak A}}}_g$. We denote by ${\overline{\boldsymbol{\mathfrak the fibered power $({\overline{\boldsymbol{\mathfrak A}}}_{g,1})^n_{{\overline{\boldsymbol{\mathfrak A}}}_g}$. (This is, to some extent, in analogy the space of stable pointed curves ${\overline{{\mathcal M}}}_{g,n}$, although we do not use stabilization.) by A}}}_{g,n}\to{\overline{\boldsymbol{\mathfrak the projection to the $i$-th factor. We have a natural relatively ample divisor ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n \subset {\overline{\boldsymbol{\mathfrak defined by ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n=\sum_i p_i^*{{\overline{\tilde{\boldsymbol{\varTheta}}}}}$. We denote by ${\overline{\mathbf the coarse moduli spaces ${\overline{\boldsymbol{\mathfrak A}}}_{g,n}$, and by ${{\overline{\boldsymbol{\Theta}}}}_n {\overline{\mathbf the image ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$. these Artin algebraic spaces [@Keel-Mori]), but since some multiple $m{{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$ descends to a Cartier divisor on ${\overline{\mathbf A}}_{g,n}$ and is relatively these are over ${\operatorname{Spec}}{{\mathbb{Z}}}$. of pairs We now collect properties of stable principally polarized quasi-abelian will use in the next section. To save we will to a stable principally polarized quasi-abelian $(P, \Theta)$ (always assumed flat over a base $S$) as a [*stable pair*]{}. | {\boldsymbol{\mathfrak A}}}_g'\to{\oveRline{\boldsYmbol{\MatHfrAk a}}}_g$ whIch iS one-to-one on geoMEtriC points ([@Alexeev:SQAV], TheoRem 4.5). AlExEEv reMArKs thaT in geneRAl THIs iS nOt An iSoMOrPhism.
agaIn, we denOte the univErsAl Family by $({\overLInE{\boldsymboL{\maThfrak A}}}_{g,1},{{\overLinE{\tilde{\BoLdsYMbol{\vArTHeta}}}}}) \tO {\overlINe{\boldSymbol{\matHfRAk A}}}_g$. We DEnote by ${\OVErLine{\Boldsymbol{\mathfraK a}}}_{g,N}$ The fibered poweR $({\overlInE{\BoLDSymBol{\Mathfrak A}}}_{g,1})^N_{{\oVerliNE{\boldsyMBoL{\MAThfRAk A}}}_g}$. (This is, to sOme extent, in ANalOgy witH tHe sPAce of sTable PoINteD curves ${\overLine{{\Mathcal M}}}_{g,N}$, althoUGh we do nOT use KnuDsen’s sTabIliZatiON.) DEnOte By $P_I:{\ovERlIne{\BOldSymbol{\maThFrAk A}}}_{g,n}\To{\ovERLINe{\boLdsYmboL{\mathFrak A}}}_{g,1}$ the projEctIon tO The $I$-th faCtor. WE havE a NaturAl relaTivelY aMple divisor ${{\overLine{\Tilde{\boldSymBoL{\vaRTHeta}}}}}_n \SUbset {\oVerLinE{\boldsyMbol{\matHFraK A}}}_{G,N}$ DEfIned by ${{\overline{\tildE{\bOLDsYmbol{\vartheta}}}}}_n=\SUm_I p_I^*{{\Overline{\TiLde{\BoldSYMbol{\vArThETa}}}}}$.
we denote By ${\overLInE{\mAthbf A}}_{g,N}$ tHe coarSe ModUli SpaceS Of ${\ovErline{\BoldsymbOl{\matHFrak A}}}_{g,n}$, and by ${{\ovERline{\boldsymbOL{\THETa}}}}_N \SubsEt {\oVerline{\mathBf A}}_{g,N}$ The iMage OF ${{\oVerLIne{\tiLde{\boLdSYmBOl{\varTheta}}}}}_n$. A-priori tHeSe are ARtin aLgebraic spaceS (see [@Keel-MoRI]), BUt since sOme mULtIPle $m{{\overline{\tiLde{\boLdsymbol{\vaRtheta}}}}}_n$ deScendS to a CartIer divisoR ON ${\overlinE{\maThbF A}}_{g,N}$ anD IS rElatively amplE, THese ArE projecTivE schemeS ovEr ${\oPerAtoRnAme{Spec}}{{\maThbb{Z}}}$.
ProPeRtIeS oF stAble pAIrs
--------------------------
We now CoLleCt A feW propERties oF stabLe prInCiPAllY polariZEd QUAsi-aBeLiAn scHemEs, Which We wiLL usE in the nExt sectioN.
To SAve wOrDs, We will rEfer to a stable PrIncipally pOlAriZed quaSI-Abelian sCheme $(P, \Theta)$ (always assumeD Flat oveR a bAse scHeme $s$) as a [*stablE paIr*]{}.
| {\boldsymbol{\mathfrak A}} }_g'\to{\o verli ne{ \bo ld symb ol{\ mathfrak A}}}_ g $ wh ich is one-to-one on g eomet ri c poi n ts ([@A lexeev: S QA V ] , T he or em4. 5 ). Alex eev remark s that ingen er al this is n o tan isomorp his m.
Again, w e d enoteth e u n ivers alfamil y by $ ( {\over line{\bol ds y mbol{\ m athfrak A }} }_{g ,1},{{\overline{\ t il d e{\boldsymbol{ \varTh et a }} } } })\to {\overlin e{ \bold s ymbol{\ m at h f r akA }}}_g$. We de note by ${\ o ver line{\ bo lds y mbol{\ mathf ra k A} }}_{g,n}$ t he f ibered po wer $( { \overli n e{\bold symbol {\m ath frak A} }} _{g ,1 } )^n _ {{ \ov e rli ne{\bold sy mb ol{\m athf r a k A}}} _g} $. ( Thisis, to some e xte nt,i n a nalog y wit h th espace of st ablepo inted curves ${ \ove rline{{\m ath ca l M }} }_{g, n }$, al tho ugh we donot use Knu ds e n ’ sstabilization.) De no t e b y $p_i:{ \overl i ne {\ b oldsymbo l{ \ma thfr a k A}}} _{g, n }\ to{\over line{\ b ol ds ymbol{\ ma thfrak A }}} _{g ,1}$t he p roject ion to t he $i $ -th factor. We have a natura l r e l at i vely am ple divisor ${{ \ over line { \t ild e {\bol dsymb ol { \v a rTheta}}}}}_n \subs et {\ove rline {\boldsymbol{ \mathfrakA } } }_{g,n}$ def i ne d by ${{\overli ne{\t ilde{\bold s ymbol{\v arThe ta}}}}}_ n=\sum_ip _ i^*{{\ov erl ine {\t ild e { \b oldsymbol{\va r T heta }} }}}$.
Wedenoteby${\ ove rli ne {\mathbfA}}_{g,n }$ t he c oar se mo d uli spac es of $ {\o verli n e{\bol dsymb ol{\ ma th f rak A}}}_{ g ,n } $ , an dby ${{ \ov er line{ \bol d sym bol{\Th eta}}}}_n \s u bset { \o verline {\mathbf A}}_ {g ,n}$ the i ma geof ${{ \ o verline{ \tilde{\boldsymbol{\var T heta}}} }}_ n$. A -pri ori these ar e Arti n a l gebrai c spac es (s ee [@ K e el-Mo r i ]) , b ut since som e mul tiple $ m{{\ overlin e{\tilde{\boldsymb o l{\ varTheta}}}}} _n$ des c e nd s t o a Car ti e r d i v isor on ${\over line{\math bf A} }_{g,n}$ a n d i srelativ ely amp le, t h ese are projecti ve scheme sover $ {\o peratornam e{Spec}} {{\mathbb { Z}}}$ .
Prope rti es ofst abl e pai rs
--- - --- ----- ------ -- ------
Weno w collec t a few properties of s tableprinc ipa lly polar ize d qu asi-abeli an s chemes, wh ich we will us e in t he n e xt se c tion.
To save word s ,wew i ll refer to a s t abl e pri nci p ally p olar ized quasi-abelia n scheme $(P, \ Thet a ) $ ( alw a ys a ss umed flat over aba s e scheme$S $) as a [*s table pa ir * ]{}. | {\boldsymbol{\mathfrak A}}}_g'\to{\overline{\boldsymbol{\mathfrak_A}}}_g$ which_is one-to-one on geometric_points ([@Alexeev:SQAV],_Theorem_4.5). Alexeev_remarks_that in general_this is not_an isomorphism.
Again, we denote_the universal family_by_$({\overline{\boldsymbol{\mathfrak A}}}_{g,1},{{\overline{\tilde{\boldsymbol{\varTheta}}}}}) \to {\overline{\boldsymbol{\mathfrak A}}}_g$. We denote by ${\overline{\boldsymbol{\mathfrak A}}}_{g,n}$ the fibered power $({\overline{\boldsymbol{\mathfrak_A}}}_{g,1})^n_{{\overline{\boldsymbol{\mathfrak_A}}}_g}$. (This_is,_to_some extent, in analogy with_the space of stable pointed_curves ${\overline{{\mathcal_M}}}_{g,n}$, although we do not use Knudsen’s stabilization.)_Denote_by $p_i:{\overline{\boldsymbol{\mathfrak A}}}_{g,n}\to{\overline{\boldsymbol{\mathfrak_A}}}_{g,1}$ the projection to the $i$-th factor. We have_a natural relatively ample divisor ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n_\subset {\overline{\boldsymbol{\mathfrak A}}}_{g,n}$_defined_by_${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n=\sum_i p_i^*{{\overline{\tilde{\boldsymbol{\varTheta}}}}}$.
We denote by_${\overline{\mathbf A}}_{g,n}$ the coarse moduli spaces_of ${\overline{\boldsymbol{\mathfrak A}}}_{g,n}$, and by ${{\overline{\boldsymbol{\Theta}}}}_n_\subset {\overline{\mathbf A}}_{g,n}$ the image of ${{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$._A-priori these are Artin algebraic spaces_(see [@Keel-Mori]), but since some_multiple $m{{\overline{\tilde{\boldsymbol{\varTheta}}}}}_n$_descends to a Cartier divisor_on ${\overline{\mathbf A}}_{g,n}$_and is_relatively ample, these_are projective schemes over ${\operatorname{Spec}}{{\mathbb{Z}}}$.
Properties of_stable pairs
--------------------------
We now_collect a few properties of stable_principally_polarized quasi-abelian schemes,_which_we_will use_in the next_section.
To_save words,_we_will refer to a stable principally_polarized_quasi-abelian scheme $(P, \Theta)$ (always assumed flat_over a base scheme_$S$)_as a [*stable pair*]{}.
|
psi$ and $t$ are the singlet and triplet order parameters respectively. For a range of values of $\nu=\psi/t$, $\Delta_{-}(\mathbf{k})$ can change sign and nodes may exist in the superconducting gap.
Recent band structure calculations for these compounds [@Lee; @2005] provide information about $\mid\mathbf{g(k)}\mid$. These results indicate that $\alpha$ is a large energy scale relative to the bandwidth and that $%
\mid\mathbf{g(k)}\mid$ is highly anisotropic. To capture these results in a model we take:$\mathbf{g(k)}=a_1\mathbf{k}-a_2[\widehat{\mathbf{x}}k_x(k_y2+k_z2)+%
\widehat{\mathbf{y}}k_y(k_z2+k_x2)+\widehat{\mathbf{z}}k_z(k_x2+k_y2)]$,with $a_2/a_1=3/2$, $\mathbf{k}$, a unit vector, and the spherical average of $\mid\mathbf{g(k)}\mid^2$ equal to unity. This form of $\mid\mathbf{g(k)}%
\mid$ is the simplest that is consistent with cubic symmetry and allows for anisotropy on a model, spherical Fermi surface.
{ | psi$ and $ t$ are the singlet and triplet order parameters respectively. For a range of value of $ \nu=\psi / t$, $ \Delta_{-}(\mathbf{k})$ can transfer sign and nodes may exist in the superconducting gap.
late band structure calculations for these compound [ @Lee; @2005 ] provide information about $ \mid\mathbf{g(k)}\mid$. These consequence indicate that $ \alpha$ is a large department of energy scale relative to the bandwidth and that $%
\mid\mathbf{g(k)}\mid$ is highly anisotropic. To capture these results in a exemplar we take:$\mathbf{g(k)}=a_1\mathbf{k}-a_2[\widehat{\mathbf{x}}k_x(k_y2+k_z2)+%
\widehat{\mathbf{y}}k_y(k_z2+k_x2)+\widehat{\mathbf{z}}k_z(k_x2+k_y2)]$,with $ a_2 / a_1=3/2 $, $ \mathbf{k}$, a unit vector, and the spherical average of $ \mid\mathbf{g(k)}\mid^2 $ equal to unity. This form of $ \mid\mathbf{g(k)}%
\mid$ is the simple that is consistent with cubic symmetry and allows for anisotropy on a mannequin, spherical Fermi surface.
! [ (Color online). The temperature dependence of (a) the normalized penetration depth $ \protect\lambda(T)/\protect\lambda_0 $ and (b) the match superfluid density $ \protect\rho_s(T)$ for Li$_2$Pd$_3$B, in which $ T_c=7 $ K, $ G=0.42 $ nm / Hz for sample \#1 and $ T_c=6.7 $ K, $ G=0.63 $ nm / Hz for sample distribution \#2. The symbols, as described in the figure, represent the experimental data and the solid lineage is a theoretical fit with parameters $%
\protect\delta=0.1 $ and $ \protect\nu=4$. The insets in the upper panel and the lower panel show a 3 - dimensional (3D) polar plot of the gap affair $%
\Delta_{-}(\mathbf{k})$, and the temperature dependence of the order argument component $ \protect\psi$ (tailspin singlet) and $ t$ (spin triplet), respectively. ] (Fig2.eps) { | psi$ and $t$ are the singlet akd triplet order paramevers reapectiveuy. For a range of values of $\iu=\psu/t$, $\Deota_{-}(\mathbf{k})$ can change rign and jodes mat exmst in the superrknductiky gap.
Dccent uand structure galculationv for these cokpuuuds [@Lee; @2005] provide information about $\myd\mathbg{g(n)}\mid$. These resolts pnqicafv uhat $\alpha$ is a large energy scals relatpve to the bandwicth and that $%
\mid\mathbf{g(k)}\mif$ is highly anisotropif. To capturg thqwe results iv a model we take:$\mathbr{g(k)}=a_1\mathbf{k}-a_2[\widehat{\mathbf{x}}k_x(k_y2+k_x2)+%
\widekat{\mathbf{y}}k_i(k_a2+k_d2)+\fidehat{\mathuf{z}}k_z(k_q2+k_y2)]$,with $a_2/a_1=3/2$, $\mabnbf{k}$, a unit vrctor, and the xphxricql average of $\mid\mathuf{g(k)}\mid^2$ equal to unijy. This fosm of $\mid\mathbf{g(k)}%
\mie$ us thg sim[lesg thxt js cknsistfnt with cubid symmetry qnd allows for anisptwipy on a modem, sphewisal Fermi surface.
{ | psi$ and $t$ are the singlet and parameters For a of values of and may exist in superconducting gap. Recent structure calculations for these compounds [@Lee; provide information about $\mid\mathbf{g(k)}\mid$. These results indicate that $\alpha$ is a large energy relative to the bandwidth and that $% \mid\mathbf{g(k)}\mid$ is highly anisotropic. To capture results a we \widehat{\mathbf{y}}k_y(k_z2+k_x2)+\widehat{\mathbf{z}}k_z(k_x2+k_y2)]$,with $a_2/a_1=3/2$, $\mathbf{k}$, a unit vector, and the spherical average of $\mid\mathbf{g(k)}\mid^2$ equal to unity. This of $\mid\mathbf{g(k)}% \mid$ is the simplest that is with cubic symmetry and for anisotropy on a model, Fermi { | psi$ and $t$ are the singlet and trIplet order ParamEteRs rEsPectIvelY. For a range of vaLUes oF $\nu=\psi/t$, $\Delta_{-}(\mathbf{k})$ can ChangE sIGn anD NoDes maY exist iN ThE SUpeRcOnDucTiNG gAp.
RecEnt Band strUcture calcUlaTiOns for these cOMpOunds [@Lee; @2005] prOviDe informatioN abOut $\mid\MaThbF{G(k)}\mid$. theSe resUlts inDIcate tHat $\alpha$ iS a LArge enERgy scalE RElAtivE to the bandwidth anD ThAT $%
\mid\mathbf{g(k)}\miD$ is higHlY AnISOtrOpiC. To capture ThEse reSUlts in a MOdEL WE taKE:$\mathbf{g(k)}=a_1\matHbf{k}-a_2[\widehaT{\MatHbf{x}}k_x(K_y2+K_z2)+%
\wIDehat{\mAthbf{Y}}k_Y(K_z2+k_X2)+\widehat{\matHbf{z}}K_z(k_x2+k_y2)]$,witH $a_2/a_1=3/2$, $\matHBf{k}$, a uniT Vector, aNd the sPheRicAl avERaGe Of $\mId\MAthBF{g(K)}\miD^2$ EquAl to unitY. THiS form Of $\miD\MATHbf{g(K)}%
\miD$ is tHe simPlest that is coNsiStenT WitH cubiC symmEtry AnD alloWs for aNisotRoPy on a model, spherIcal fermi surfAce.
{ | psi$ and $t$ are the singl et and tri pletord erpa rame ters respectively. Fora range of values of $ \nu=\ ps i /t$, $\ Delta _{-}(\m a th b f {k} )$ c anch a ng e sig n a nd node s may exis t i nthe supercon d uc ting gap.
Re cent band st ruc ture c al cul a tions fo r the se com p ounds[@Lee; @2 00 5 ] prov i de info r m at ionabout $\mid\mathb f {g ( k)}\mid$. Thes e resu lt s i n d ica tethat $\alp ha $ isa largee ne r g y sc a le relative t o the bandw i dth and t ha t $ %
\mid\ mathb f{ g (k) }\mid$ is h ighl y anisotr opic.T o captu r e these resul tsina mo d el w e t ak e :$\ m at hbf { g(k )}=a_1\m at hb f{k}- a_2[ \ w i d ehat {\m athb f{x}} k_x(k_y2+k_z2 )+%
\wi d eha t{\ma thbf{ y}}k _y (k_z2 +k_x2) +\wid eh at{\mathbf{z}}k _z(k _x2+k_y2) ]$, wi th$a _2/a_ 1 =3/2$, $\ mat hbf{k}$ , a uni t ve ct o r , a nd the spherical a ve r a ge of $\mi d\math b f{ g( k )}\mid^2 $equ al t o unity . Th i sform of$\mid\ m at hb f{g(k)} %\mid$is th e s imple s t th at isconsiste nt wi t h cubic symmet r y and allowsf or a ni s otro pyon a model, sph e rica l Fe r mi su r face.
{ | psi$ and_$t$ are_the singlet and triplet_order parameters_respectively._For a_range_of values of_$\nu=\psi/t$, $\Delta_{-}(\mathbf{k})$ can_change sign and nodes_may exist in_the_superconducting gap.
Recent band structure calculations for these compounds [@Lee; @2005] provide information about $\mid\mathbf{g(k)}\mid$._These_results indicate_that_$\alpha$_is a large energy scale_relative to the bandwidth and_that $%
\mid\mathbf{g(k)}\mid$_is highly anisotropic. To capture these results in_a_model we take:$\mathbf{g(k)}=a_1\mathbf{k}-a_2[\widehat{\mathbf{x}}k_x(k_y2+k_z2)+%
\widehat{\mathbf{y}}k_y(k_z2+k_x2)+\widehat{\mathbf{z}}k_z(k_x2+k_y2)]$,with_$a_2/a_1=3/2$, $\mathbf{k}$, a unit vector, and the spherical average_of $\mid\mathbf{g(k)}\mid^2$ equal to unity. This_form of $\mid\mathbf{g(k)}%
\mid$_is_the_simplest that is consistent_with cubic symmetry and allows for_anisotropy on a model, spherical Fermi_surface.
{ |
}$$ Then, for each $t\in {\mathbb R}$, there exists a unique frame $e=(e_1, \dots, e_n)$ for $u^*TN$ with the associated connection form, $A$, satisfying the uniform-in-time estimates\
[()]{}
${\displaystyle}{{\left\|{A}\right\|}_{L^4} \lesssim \|du\|_{H^1} \lesssim {\varepsilon}_0 \label{A L4}}$\
${\displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{1,\frac{8}{3}}} \lesssim \|du\|_{L^8} \|du\|_{\dot{H}^1}}$\
${\displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{2,\frac{8}{5}}} \lesssim \|du\|_{L^8} \|du\|_{\dot{H}^1}}$\
$ \|A\|_{L^{\infty}} \lesssim \|du\|_{L^{8,2}({\mathbb R}^4)}^2$
as long as ${\varepsilon}_0$ is small enough. Also, the frame $e$, and hence $A$, depend continuously on $t$. Above, $L^{8,2}=L^{8,2}({\mathbb R}^4)$ denotes the Lorentz space.
The estimates are deduced via a perturbative method as the assumptions in – imply that the left hand side of is a slight perturbation of the flat Laplacian on ${\mathbb R}^4$. To simplify notation, in what follows we consider an elliptic operator of the form
$$\begin{aligned}
\label{def L}
L:= g^{ij}{\partial}_i {\partial}_j + b^j {\partial}_j + c
\end{aligned}$$
and the elliptic system
$$\begin{aligned}
\label{L}
L A_{\ell}= g^{i j} {\partial}_{j} G_{i \ell} \end{aligned}$$
where $G_{i\ell} := F_{i \ell} - [A_{i}, A_{\ell}] $, and $b$ and $c$ satisfy
$$\begin{aligned}
\|b\|_{L^{4,1}({\mathbb R}^4)} \lesssim {\varepsilon}\label{b}\\ \notag
\\
\|{\partial}b\|_{L^{2,1}({\mathbb R}^4)} \lesssim | } $ $ Then, for each $ t\in { \mathbb R}$, there exists a unique frame $ e=(e_1, \dots, e_n)$ for $ u^*TN$ with the associated association human body, $ A$, satisfying the uniform - in - time estimates\
[ () ] { }
$ { \displaystyle}{{\left\|{A}\right\|}_{L^4 } \lesssim \|du\|_{H^1 } \lesssim { \varepsilon}_0 \label{A L4}}$\
$ { \displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{1,\frac{8}{3 } } } \lesssim \|du\|_{L^8 } \|du\|_{\dot{H}^1}}$\
$ { \displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{2,\frac{8}{5 } } } \lesssim \|du\|_{L^8 } \|du\|_{\dot{H}^1}}$\
$ \|A\|_{L^{\infty } } \lesssim \|du\|_{L^{8,2}({\mathbb R}^4)}^2 $
as long as $ { \varepsilon}_0 $ is little enough. Also, the frame $ e$, and therefore $ A$, depend continuously on $ t$. Above, $ L^{8,2}=L^{8,2}({\mathbb R}^4)$ denotes the Lorentz space.
The estimate are deduced via a perturbative method as the assumption in – imply that the left handwriting slope of is a slight perturbation of the flat Laplacian on $ { \mathbb R}^4$. To simplify notation, in what follows we see an elliptic operator of the form
$ $ \begin{aligned }
\label{def lambert }
L:= g^{ij}{\partial}_i { \partial}_j + b^j { \partial}_j + c
\end{aligned}$$
and the elliptic system
$ $ \begin{aligned }
\label{L }
L A_{\ell}= g^{i j } { \partial}_{j } G_{i \ell } \end{aligned}$$
where $ G_{i\ell }: = F_{i \ell } - [ A_{i }, A_{\ell } ] $, and $ b$ and $ c$ satisfy
$ $ \begin{aligned }
\|b\|_{L^{4,1}({\mathbb R}^4) } \lesssim { \varepsilon}\label{b}\\ \notag
\\
\|{\partial}b\|_{L^{2,1}({\mathbb R}^4) } \lesssim | }$$ Thfn, for each $t\in {\mathbb R}$, there exists a unique frame $e=(e_1, \dots, e_n)$ for $u^*TN$ with the associaved xonnextion form, $A$, satisfyine the univorm-in-tine ewrimates\
[()]{}
${\dis'maystylc}{{\jeft\|{Z}\vight\|}_{N^4} \lesssim \|du\|_{H^1} \lgsssim {\varepvilon}_0 \label{A L4}}$\
${\girppaystyle}{{\left\|{A}\right\|}_{\dot{W}^{1,\frac{8}{3}}} \lesssim \|qu\|_{L^8} \|du\|_{\dpt{J}^1}}$\
${\displaystyle}{{\lgft\|{A}\rpgrt\|}_{\dof{W}^{2,\frac{8}{5}}} \lesssim \|du\|_{L^8} \|du\|_{\dot{H}^1}}$\
$ \|A\|_{L^{\infty}} \messsim \|du\|_{L^{8,2}({\mathbb R}^4)}^2$
ax long as ${\varepsilon}_0$ is smwll fnough. Also, the frwme $e$, and hghce $Q$, depend conginuously on $t$. Above, $L^{8,2}=M^{8,2}({\mathbb R}^4)$ denotes the Lorentz soace.
Tke estimatew qre geduced via a pegturbative mebnod as the asxumptions in – ikplb thqt the left hand side of is a slight pgrturbatiot kf the flat Laplaxiqn on ${\matvbb F}^4$. To sijpkiry notwtikn, in what follows we consider an elliptoc iperator of tge fori
$$\bqgin{aligned}
\label{def L}
L:= g^{ij}{\partial}_i {\paruial}_j + b^j {\partial}_j + c
\end{alitned}$$
and the elliptic dystem
$$\begyn{aligned}
\label{L}
L A_{\ell}= g^{i j} {\partial}_{j} G_{i \ell} \end{alicned}$$
wiefe $Y_{l\cll} := F_{l \ell} - [A_{i}, A_{\ell}] $, and $b$ and $c$ satisfy
$$\begin{alignqs}
\|b\|_{K^{4,1}({\msthbb R}^4)} \lesssii {\varepsilom}\lwbrj{b}\\ \notag
\\
\|{\partixl}b\|_{L^{2,1}({\macgbg R}^4)} \lesssim | }$$ Then, for each $t\in {\mathbb R}$, a frame $e=(e_1, e_n)$ for $u^*TN$ $A$, the uniform-in-time estimates\ ${\displaystyle}{{\left\|{A}\right\|}_{L^4} \lesssim \|du\|_{H^1} {\varepsilon}_0 \label{A L4}}$\ ${\displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{1,\frac{8}{3}}} \lesssim \|du\|_{L^8} ${\displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{2,\frac{8}{5}}} \lesssim \|du\|_{L^8} \|du\|_{\dot{H}^1}}$\ $ \|A\|_{L^{\infty}} \lesssim \|du\|_{L^{8,2}({\mathbb R}^4)}^2$ as long as ${\varepsilon}_0$ small enough. Also, the frame $e$, and hence $A$, depend continuously on $t$. $L^{8,2}=L^{8,2}({\mathbb denotes Lorentz The estimates are deduced via a perturbative method as the assumptions in – imply that the hand side of is a slight perturbation of flat Laplacian on ${\mathbb To simplify notation, in what we an elliptic of form \label{def L} L:= {\partial}_j + b^j {\partial}_j + c \end{aligned}$$ and the elliptic system $$\begin{aligned} \label{L} L A_{\ell}= g^{i j} G_{i \ell} $G_{i\ell} := \ell} [A_{i}, $, and $b$ satisfy $$\begin{aligned} \|b\|_{L^{4,1}({\mathbb R}^4)} \lesssim {\varepsilon}\label{b}\\ R}^4)} \lesssim | }$$ Then, for each $t\in {\mathbb R}$, therE exists a unIque fRamE $e=(e_1, \DoTs, e_n)$ For $u^*tN$ with the assocIAted Connection form, $A$, satisfyIng thE uNIforM-In-Time eStimateS\
[()]{}
${\DiSPLayStYlE}{{\leFt\|{a}\RiGht\|}_{L^4} \lEssSim \|du\|_{H^1} \lEsssim {\varePsiLoN}_0 \label{A L4}}$\
${\dispLAyStyle}{{\left\|{A}\RigHt\|}_{\dot{W}^{1,\frac{8}{3}}} \leSssIm \|du\|_{L^8} \|dU\|_{\dOt{H}^1}}$\
${\DIsplaYstYle}{{\leFt\|{A}\rigHT\|}_{\dot{W}^{2,\fRac{8}{5}}} \lesssiM \|dU\|_{l^8} \|du\|_{\dot{h}^1}}$\
$ \|a\|_{L^{\infty}} \LESsSim \|dU\|_{L^{8,2}({\mathbb R}^4)}^2$
as long as ${\VArEPsilon}_0$ is small eNough. ALsO, ThE FRamE $e$, aNd hence $A$, dePeNd conTInuouslY On $T$. aBOve, $l^{8,2}=l^{8,2}({\mathbb R}^4)$ denotEs the LorentZ SpaCe.
The eStImaTEs are dEduceD vIA a pErturbative MethOd as the asSumptiONs in – impLY that thE left hAnd SidE of iS A sLiGht PeRTurBAtIon OF thE flat LapLaCiAn on ${\mAthbB r}^4$. tO SimpLifY notAtion, In what follows We cOnsiDEr aN elliPtic oPeraToR of thE form
$$\bEgin{aLiGned}
\label{def L}
L:= g^{Ij}{\paRtial}_i {\parTiaL}_j + B^j {\pArTial}_j + C
\End{aliGneD}$$
anD the ellIptic sySTem
$$\BeGIN{AlIgned}
\label{L}
L A_{\ell}= g^{i J} {\pARTiAl}_{j} G_{i \ell} \End{aliGNeD}$$
wHEre $G_{i\ell} := f_{i \Ell} - [a_{i}, A_{\eLL}] $, And $b$ aNd $c$ sATiSfy
$$\begin{AligneD}
\|B\|_{L^{4,1}({\MaThbb R}^4)} \leSsSim {\varEpSilOn}\lAbel{b}\\ \NOtag
\\
\|{\PartiaL}b\|_{L^{2,1}({\mathbB R}^4)} \lesSSim | }$$ Then, for each $t\in { \mathbb R} $, th ere ex is ts a uni que frame $e=( e _1,\dots, e_n)$ for $u^*T N$ wi th thea ss ociat ed conn e ct i o n f or m, $A $, sa tisfy ing the un iform-in-t ime e stimates\
[ ( )] {}
${\dis pla ystyle}{{\le ft\ |{A}\r ig ht\ | }_{L^ 4}\less sim \| d u\|_{H ^1} \less si m {\var e psilon} _ 0 \ labe l{A L4}}$\
${\di s pl a ystyle}{{\left \|{A}\ ri g ht \ | }_{ \do t{W}^{1,\f ra c{8}{ 3 }}} \le s ss i m \|d u \|_{L^8} \|du \|_{\dot{H} ^ 1}} $\
${ \d isp l aystyl e}{{\ le f t\| {A}\right\| }_{\ dot{W}^{2 ,\frac { 8}{5}}} \lesssi m \|du \|_ {L^ 8} \ | du \| _{\ do t {H} ^ 1} }$\
$\|A\|_{L ^{ \i nfty} } \l e s s s im \ |du \|_{ L^{8, 2}({\mathbb R }^4 )}^2 $
a s lon g as${\v ar epsil on}_0$ is s ma ll enough. Also , th e frame $ e$, a ndhe nce $ A $, dep end co ntinuou sly on$ t$. A b o v e, $L^{8,2}=L^{8,2}( {\ m a th bb R}^4) $ deno t es t h e Lorent zspa ce.T he es tima t es are ded uced v i aaperturb at ive me th odasthe a s sump tionsin – imp ly th a t the left han d side of is as l ig h t pe rtu rbation ofthef latLapl a ci ano n ${\ mathb bR }^ 4 $. To simplify nota ti on, in what follows we c onsider an e l liptic o pera t or of the form
$ $\beg in{aligned }
\label{ def L }
L:= g ^{ij}{\pa r t ial}_i { \pa rti al} _j+ b^ j {\partial}_ j + c \ end{ali gne d}$$
a ndthe el lip ti c system
$$\begi n{ al ig ne d}\labe l {L}
L A_ {\ ell }= g^ {i j} {\part ial}_ {j}G_ {i \el l} \end { al i g ned} $$
wher e $ G_ {i\el l} : = F _{i \el l} - [A_{ i}, A_{\ el l} ] $, an d $b$ and $c$ s atisfy
$$ \b egi n{alig n e d}
\|b\| _{L^{4,1}({\mathbb R}^4 ) } \less sim {\va reps ilon}\lab el{ b}\\ \ not a g
\\
\ |{\par tial} b\ |_{ L ^ {2,1} ( { \m ath bb R}^4)} \l e s ssi m | }$$ Then,_for each_$t\in {\mathbb R}$, there_exists a_unique_frame $e=(e_1,_\dots,_e_n)$ for $u^*TN$_with the associated_connection form, $A$, satisfying_the uniform-in-time estimates\
[()]{}
${\displaystyle}{{\left\|{A}\right\|}_{L^4}_\lesssim_\|du\|_{H^1} \lesssim {\varepsilon}_0 \label{A L4}}$\
${\displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{1,\frac{8}{3}}} \lesssim \|du\|_{L^8} \|du\|_{\dot{H}^1}}$\
${\displaystyle}{{\left\|{A}\right\|}_{\dot{W}^{2,\frac{8}{5}}} \lesssim \|du\|_{L^8} \|du\|_{\dot{H}^1}}$\
$ \|A\|_{L^{\infty}} \lesssim \|du\|_{L^{8,2}({\mathbb_R}^4)}^2$
as_long as_${\varepsilon}_0$_is_small enough. Also, the frame_$e$, and hence $A$, depend_continuously on_$t$. Above, $L^{8,2}=L^{8,2}({\mathbb R}^4)$ denotes the Lorentz space.
The_estimates_are deduced via_a perturbative method as the assumptions in – imply that_the left hand side of _is a slight_perturbation_of_the flat Laplacian on_${\mathbb R}^4$. To simplify notation, in_what follows we consider an elliptic_operator of the form
$$\begin{aligned}
\label{def L}
L:= g^{ij}{\partial}_i_{\partial}_j + b^j {\partial}_j + c
_\end{aligned}$$
and the elliptic system
$$\begin{aligned}
\label{L}
L A_{\ell}=_g^{i j}_{\partial}_{j} G_{i \ell} \end{aligned}$$
where $G_{i\ell}_:= F_{i_\ell} -_[A_{i}, A_{\ell}] $,_and $b$ and $c$ satisfy
$$\begin{aligned}
\|b\|_{L^{4,1}({\mathbb R}^4)}_\lesssim {\varepsilon}\label{b}\\ \notag
\\
\|{\partial}b\|_{L^{2,1}({\mathbb_R}^4)} \lesssim |
diversity $\operatorname{\mathbb{ED}}$ achieves its maximal value $1$ if and only if the column vectors of $\tilde{M}_{\backslash y}$ are mutually orthogonal.
Note that most of the previous definitions of diversity, e.g., those defined on prediction errors, include the predictions $F_{y}^{k}$ on the true label $y$ [@liu1999ensemble; @liu1999simultaneous; @islam2003constructive; @kuncheva2003measures]. Since DNNs are no longer weak classifiers in the most cases [@goodfellow2016deep], encouraging $F_{y}^{k}$ to be diverse may make the loss on accuracy outweighs the gain for DNNs, or render the convergence point of the training process uncontrollable. In contrast, we define the diversity among the normalized non-maximal predictions $\tilde{F}^{k}_{\backslash y}$, which allows the maximal prediction of each network to be consistent with the true label, and thus will not affect ensemble accuracy. What’s more, promoting this diversity could improve robustness. Because, in the adversarial setting, when the maximal prediction corresponds to the true label $y$, the non-maximal predictions correspond to the labels $[L]\backslash\{y\}$, which include all potentially wrong labels returned for the adversarial examples. Thus a high diversity or inconsistency on the non-maximal predictions $F^{k}_{\backslash y}$ can lower down the transferability of adversarial examples among the networks, and further lead to better robustness of the ensemble.
The $K$ individual models $\{F^{k}(x,\theta^{k})\}_{k\in[K]}$; the training dataset $\mathcal{D}=\{(x_i, y_i)\}_{i\in [N]}$.\
Initialize each $\theta^k$ as $\theta^k_0$, the training step counters as $c_k=0$, $\varepsilon_k$ be the learning rate variables, $I=[K]$ be the indicator set, where $k\in[K]$. Calculate the objective on a mini-batch of data $\mathcal{D}_m$ $$\mathcal{L}_{\text{ADP}}^m=\frac{1}{\left|\mathcal{D}_m\right|} \sum_{(x_i, y_i) \in \mathcal{D}_m}\left[\mathcal{L}_{\text{ECE}}-\operatorname{\text{ADP}}_{\alpha,\ | diversity $ \operatorname{\mathbb{ED}}$ achieves its maximal value $ 1 $ if and only if the column vectors of $ \tilde{M}_{\backslash y}$ are mutually extraneous.
notice that most of the former definitions of diversity, e.g., those define on prediction errors, include the prediction $ F_{y}^{k}$ on the on-key label $ y$ [ @liu1999ensemble; @liu1999simultaneous; @islam2003constructive; @kuncheva2003measures ]. Since DNNs are no longer weak classifier in the most cases [ @goodfellow2016deep ], encouraging $ F_{y}^{k}$ to be divers may make the loss on accuracy outweighs the gain for DNNs, or render the overlap point of the training procedure uncontrollable. In contrast, we specify the diversity among the normalized non - maximal predictions $ \tilde{F}^{k}_{\backslash y}$, which allow the maximal prediction of each network to be consistent with the true label, and thus will not involve ensemble accuracy. What ’s more, promoting this diversity could improve robustness. Because, in the adversarial setting, when the maximal prediction corresponds to the genuine label $ y$, the non - maximal predictions correspond to the labels $ [ L]\backslash\{y\}$, which include all potentially wrong labels returned for the adversarial examples. Thus a high diversity or inconsistency on the non - maximal predictions $ F^{k}_{\backslash y}$ can lower down the transferability of adversarial examples among the networks, and further moderate to better robustness of the ensemble.
The $ K$ individual mannequin $ \{F^{k}(x,\theta^{k})\}_{k\in[K]}$; the education dataset $ \mathcal{D}=\{(x_i, y_i)\}_{i\in [ N]}$.\
format each $ \theta^k$ as $ \theta^k_0 $, the training step counters as $ c_k=0 $, $ \varepsilon_k$ be the learning rate variables, $ I=[K]$ be the indicator set, where $ k\in[K]$. count the objective on a mini - batch of data $ \mathcal{D}_m$ $ $ \mathcal{L}_{\text{ADP}}^m=\frac{1}{\left|\mathcal{D}_m\right| } \sum_{(x_i, y_i) \in \mathcal{D}_m}\left[\mathcal{L}_{\text{ECE}}-\operatorname{\text{ADP}}_{\alpha,\ | digersity $\operatorname{\mathnb{ED}}$ achieves its maximel valus $1$ if ana only if the column vectors od $\tilee{M}_{\backslash y}$ are mutjally ortjogonal.
Nite uhat most of the 'devious definjbions if diversity, e.n., those defhned on predicdiun errors, include the predictions $F_{y}^{k}$ on the tgue label $y$ [@liu1999gnsemnje; @ljl1999slmultaneous; @islam2003constructive; @kuhcheva2003mtasures]. Since DNNs are no longer weak classivierd in the most cased [@goodfellow2016eeep], wncouraging $W_{y}^{k}$ to be diverse may jake the loss on accuracy outwekghs che gain fot SNJv, or render the bonvergence point of dhe traoning process mnconvroloable. In contrast, we vefine the diversity among tha uormalized non-maximal peedicjions $\tilaw{F}^{k}_{\caciskaah y}$, wjici allows ths maximal peediction of each nttwjgl to be consjstent wyth the true label, and thus will not afxecf ensemble accuracy. Whar’s more, promoting thid diversiey could improve robustness. Because, in the adversdrial retuikn, whdb hhe maximal prediction corresponds to the truq lsbvl $y$, the non-maximcl predictions cprgexkond to the lacels $[L]\yzcislash\{y\}$, which inclkde all potebtially wwong labels returned for the adcersarial excmpoes. Thus a high dirersity or iuconsixtencu on the non-maximal prebictiohs $F^{k}_{\backslwsh y}$ can uower down the tfanxfarability of adversarial evamples anong the negworls, and further lfad to better robustness of tke envemble.
The $N$ individual models $\{F^{k}(x,\theta^{k})\}_{k\ii[L]}$; the trainimg danaset $\matkcal{D}=\{(x_l, y_i)\}_{i\in [N]}$.\
Initiajize each $\thetc^k$ as $\thzta^k_0$, tfe traininf step rounters as $s_k=0$, $\varepsilon_n$ be the leariing rate varuablws, $I=[K]$ bd the indicator set, wherv $h\in[K]$. Calculate the objcctivg kn a mini-batch if eata $\mathcal{D}_m$ $$\katfcaj{L}_{\nexv{ADP}}^m=\ssac{1}{\left|\mathcdl{D}_m\figfy|} \sum_{(b_i, y_i) \in \mabhcxl{D}_m}\keft[\mathcal{L}_{\text{ECE}}-\o[eraforname{\text{ADP}}_{\alphs,\ | diversity $\operatorname{\mathbb{ED}}$ achieves its maximal value $1$ only the column of $\tilde{M}_{\backslash y}$ most the previous definitions diversity, e.g., those on prediction errors, include the predictions on the true label $y$ [@liu1999ensemble; @liu1999simultaneous; @islam2003constructive; @kuncheva2003measures]. Since DNNs are no weak classifiers in the most cases [@goodfellow2016deep], encouraging $F_{y}^{k}$ to be diverse may the on outweighs gain for DNNs, or render the convergence point of the training process uncontrollable. In contrast, we the diversity among the normalized non-maximal predictions $\tilde{F}^{k}_{\backslash which allows the maximal of each network to be with true label, thus not ensemble accuracy. What’s promoting this diversity could improve robustness. Because, in the adversarial setting, when the maximal prediction corresponds to true label non-maximal predictions to labels which include all labels returned for the adversarial examples. diversity or inconsistency on the non-maximal predictions $F^{k}_{\backslash can lower the transferability of adversarial examples among networks, and further lead to better robustness of ensemble. The $K$ individual models $\{F^{k}(x,\theta^{k})\}_{k\in[K]}$; the training dataset $\mathcal{D}=\{(x_i, y_i)\}_{i\in [N]}$.\ Initialize each $\theta^k$ the training step counters $c_k=0$, $\varepsilon_k$ be learning variables, be indicator set, $k\in[K]$. Calculate the objective on a mini-batch of data $\mathcal{D}_m$ $$\mathcal{L}_{\text{ADP}}^m=\frac{1}{\left|\mathcal{D}_m\right|} y_i) \in \mathcal{D}_m}\left[\mathcal{L}_{\text{ECE}}-\operatorname{\text{ADP}}_{\alpha,\ | diversity $\operatorname{\mathBb{ED}}$ achievEs its MaxImaL vAlue $1$ If anD only if the coluMN vecTors of $\tilde{M}_{\backslash y}$ Are muTuALly oRThOgonaL.
Note thAT mOST of ThE pRevIoUS dEfiniTioNs of divErsity, e.g., thOse DeFined on prediCTiOn errors, inCluDe the predictIonS $F_{y}^{k}$ on ThE trUE labeL $y$ [@lIu1999ensEmble; @lIU1999simulTaneous; @isLaM2003ConstrUCtive; @kuNCHeVa2003meAsures]. Since DNNs arE No LOnger weak classIfiers In THe MOSt cAseS [@goodfelloW2016dEep], enCOuraginG $f_{y}^{K}$ TO Be dIVerse may make tHe loss on accURacY outweIgHs tHE gain fOr DNNS, oR RenDer the conveRgenCe point of The traINing proCEss uncoNtrollAblE. In ContRAsT, wE deFiNE thE DiVerSIty Among the NoRmAlizeD non-MAXIMal pRedIctiOns $\tiLde{F}^{k}_{\backslasH y}$, wHich ALloWs the MaximAl prEdIctioN of eacH netwOrK to be consistent With The true laBel, AnD thUs Will nOT affecT enSemBle accuRacy. WhaT’S moRe, PROMoTing this diversity cOuLD ImProve robUstnesS. beCaUSe, in the aDvErsAriaL SEttinG, wheN ThE maximal PredicTIoN cOrrespoNdS to the TrUe lAbeL $y$, the NOn-maXimal pRedictioNs corREspond to the labELs $[L]\backslash\{y\}$, WHiCH InCLude All Potentially WronG LabeLs reTUrNed FOr the AdverSaRIaL Examples. Thus a high diVeRsity oR incoNsistency on thE non-maximaL PREdictionS $F^{k}_{\bACkSLash y}$ can lower dOwn thE transferaBIlity of aDversArial exaMples amonG THe networKs, aNd fUrtHer LEAd To better robusTNEss oF tHe ensemBle.
the $K$ indIviDuaL moDelS $\{F^{K}(x,\theta^{k})\}_{k\In[K]}$; the trAiNiNg DaTasEt $\matHCal{D}=\{(x_i, y_i)\}_{I\iN [N]}$.\
INiTiaLize eACh $\thetA^k$ as $\tHeta^K_0$, tHe TRaiNing steP CoUNTers As $C_k=0$, $\VarePsiLoN_k$ be tHe leARniNg rate vAriables, $I=[k]$ be THe inDiCaTor set, wHere $k\in[K]$. CalcuLaTe the objecTiVe oN a mini-BATch of datA $\mathcal{D}_m$ $$\mathcal{L}_{\text{Adp}}^m=\frac{1}{\lEft|\MathcAl{D}_m\Right|} \sum_{(x_I, y_i) \In \mathCal{d}_M}\left[\mAthcal{l}_{\text{eCe}}-\opERAtornAME{\tExt{aDp}}_{\alpha,\ | diversity $\operatorname{ \mathbb{ED }}$ a chi eve sitsmaxi mal value $1$i f an d only if the column v ector so f $\ t il de{M} _{\back s la s h y} $ar e m ut u al ly or tho gonal.
Note that mo st of the prev i ou s definiti ons of diversit y,e.g.,th ose defin edon pr edicti o n erro rs, inclu de the pr e diction s $F _{y} ^{k}$ on the true la b el $y$ [@liu19 99ense mb l e; @ liu 199 9simultane ou s; @i s lam2003 c on s t r uct i ve; @kuncheva 2003measure s ].SinceDN Nsa re nolonge rw eak classifier s in the most cases [@goodf e llow201 6deep] , e nco urag i ng $ F_{ y} ^ {k} $ t o b e di verse ma yma ke th e lo s s o n ac cur acyoutwe ighs the gain fo r DN N s,or re ndertheco nverg ence p ointof the training p roce ss uncont rol la ble .In co n trast, we de fine th e diver s ity a m o n gthe normalized non -m a x im al predi ctions $\ ti l de{F}^{k }_ {\b acks l a sh y} $, w h ic h allows the m a xi ma l predi ct ion of e ach ne twork to b e cons istent w ith t h e true label,a nd thus willn ot a ff e ct e nse mble accura cy.W hat’ s mo r e, pr o motin g thi sd iv e rsity could improve r obustn ess.Because, in t he adversa r i a l settin g, w h en the maximal pr edict ion corres p onds tothe t rue labe l $y$, th e non-maxi mal pr edi cti o n scorrespond to t he l ab els $[L ]\b ackslas h\{ y\} $,whi ch includeall pote nt ia ll ywro ng la b els retu rn edfo r t he ad v ersari al ex ampl es .T hus a high di v e rsit yor inc ons is tency ont henon-max imal pred ict i ons$F ^{ k}_{\ba ckslash y}$ c an lower dow nthe trans f e rability of adversarial example s amongthe netw orks , and fur the r lead to better robus tness o f t h e ense m b le .
Th e $K$ indi v i dua l mod el s $\ {F^{k}( x,\theta^{k})\}_{k \ in[ K]}$; the tra ini ng d a t as et$ \m a thc al { D}= \ { (x_i, y_i)\}_{i \in [N]}$. \I ni tialize ea c h $ \t heta^k$ as $\t heta^ k _0$, th e trainin g step co un ters a s $ c_k=0$, $\ varepsil on_k$ bet he le a rn ing r ate varia bl es, $I=[ K]$ be the indi catorse t, whe re $k \i n[K]$. C alculate the objectiveon a m ini-b atc h of data $\ m ath cal{D}_m$ $$\ mathcal{L} _{\ tex t{ADP }}^ m =\fra c{1} { \l eft | \math cal{ D }_m\right | }\su m _ {( x_i, y_i) \ i n \ma thcal {D} _ m}\lef t[\m athcal{L}_{\text{ E CE}}-\operator name { \ tex t{A D P}}_ {\ alpha,\ | diversity_$\operatorname{\mathbb{ED}}$ achieves_its maximal value $1$_if and_only_if the_column_vectors of $\tilde{M}_{\backslash_y}$ are mutually_orthogonal.
Note that most of_the previous definitions_of_diversity, e.g., those defined on prediction errors, include the predictions $F_{y}^{k}$ on the true_label_$y$ [@liu1999ensemble; @liu1999simultaneous;_@islam2003constructive;_@kuncheva2003measures]._Since DNNs are no longer_weak classifiers in the most_cases [@goodfellow2016deep], encouraging_$F_{y}^{k}$ to be diverse may make the loss_on_accuracy outweighs the_gain for DNNs, or render the convergence point of_the training process uncontrollable. In contrast,_we define the_diversity_among_the normalized non-maximal predictions_$\tilde{F}^{k}_{\backslash y}$, which allows the maximal_prediction of each network to be_consistent with the true label, and thus_will not affect ensemble accuracy. What’s_more, promoting this diversity could_improve robustness._Because, in the adversarial setting,_when the maximal_prediction corresponds_to the true_label $y$, the non-maximal predictions correspond_to the labels_$[L]\backslash\{y\}$, which include all potentially wrong_labels_returned for the_adversarial_examples._Thus a_high diversity or_inconsistency_on the_non-maximal_predictions $F^{k}_{\backslash y}$ can lower down_the_transferability of adversarial examples among the networks,_and further lead to_better_robustness of the ensemble.
The_$K$ individual models $\{F^{k}(x,\theta^{k})\}_{k\in[K]}$; the_training dataset $\mathcal{D}=\{(x_i, y_i)\}_{i\in [N]}$.\
Initialize each_$\theta^k$ as_$\theta^k_0$, the_training step counters as $c_k=0$, $\varepsilon_k$ be the learning rate variables,_$I=[K]$ be the indicator set, where_$k\in[K]$. Calculate the objective_on a_mini-batch_of data $\mathcal{D}_m$_$$\mathcal{L}_{\text{ADP}}^m=\frac{1}{\left|\mathcal{D}_m\right|}_\sum_{(x_i, y_i)_\in \mathcal{D}_m}\left[\mathcal{L}_{\text{ECE}}-\operatorname{\text{ADP}}_{\alpha,\ |
ck*]{} Galactic mask extended to cut out $\pm30^\circ$ of the Galactic plane. The level of systematics correspond to the pessimistic expectation of calibration errors and sky residuals.[]{data-label="fig:conflevel128+1024"}](60_128+1024_noi_maskext_CROP.pdf "fig:")![Marginalised likelihoods, and 68% and 95% contours of the parameters $A$, $b_0$, $l_0$, and $T_{0}$ at $N_{\rm side}=128$ (red) and at $N_{\rm side}=1024$ (blue): top left, dipole only; top right, dipole+noise; bottom left, dipole+noise+mask; and bottom right, dipole+noise+mask+systematics. The reference frequency channel is 60GHz and the noise is 7.5$\mu$K.arcmin. The mask used here is the [*Planck*]{} Galactic mask extended to cut out $\pm30^\circ$ of the Galactic plane. The level of systematics correspond to the pessimistic expectation of calibration errors and sky residuals.[]{data-label="fig:conflevel128+1024"}](60_128+1024_noi_maskext_badcal_badsky_CROP.pdf "fig:")
[width=1]{}
$N_{\rm side}=128$ $A({\rm mK})$ $b_0(^\circ)$ $l_0(^\circ)$ $T_{0}({\rm mK})$
--------------------- ----------------------------------------------- --------------------------------- ----------------------------------------------- -------------------------------------
dipole $\ensuremath{3.3644\pm0.0028}$ $48.242\pm0.047$ $\ensuremath{263.999\pm0.070}$ $2725.4793\pm0.0016$
dip+noi $\ensuremath{3.3644\pm0.0028}$ $48.240\pm0.047$ $\ensuremath{263.998\pm0.071}$ $2725.4793\pm0.0016$
dip+noi+mask $\ensuremath{3.3644\pm0.0041}$ $48.240\pm0.075$ $\ensuremath{264.00\pm0.13}$ $2725.4797\pm0.0024$
dip | ck * ] { } Galactic mask extended to cut out $ \pm30^\circ$ of the Galactic airplane. The degree of systematics correspond to the pessimistic expectation of calibration error and sky residuals.[]{data - label="fig: conflevel128 + 1024"}](60_128 + 1024_noi_maskext_CROP.pdf " fig:")![Marginalised likelihood, and 68% and 95% contours of the parameters $ A$, $ b_0 $, $ l_0 $, and $ T_{0}$ at $ N_{\rm side}=128 $ (red) and at $ N_{\rm side}=1024 $ (gloomy ): top left, dipole only; top right field, dipole+noise; bottomland left, dipole+noise+mask; and bottom right field, dipole+noise+mask+systematics. The reference frequency channel is 60GHz and the noise is 7.5$\mu$K.arcmin. The mask use here is the [ * Planck * ] { } Galactic masquerade extended to ignore out $ \pm30^\circ$ of the Galactic airplane. The level of systematics correspond to the pessimistic expectation of calibration erroneousness and sky residuals.[]{data - label="fig: conflevel128 + 1024"}](60_128 + 1024_noi_maskext_badcal_badsky_CROP.pdf " fig: ")
[ width=1 ] { }
$ N_{\rm side}=128 $ $ A({\rm mK})$ $ b_0(^\circ)$ $ l_0(^\circ)$ $ T_{0}({\rm mK})$
--------------------- ----------------------------------------------- --------------------------------- ----------------------------------------------- -------------------------------------
dipole $ \ensuremath{3.3644\pm0.0028}$ $ 48.242\pm0.047 $ $ \ensuremath{263.999\pm0.070}$ $ 2725.4793\pm0.0016 $
dip+noi $ \ensuremath{3.3644\pm0.0028}$ $ 48.240\pm0.047 $ $ \ensuremath{263.998\pm0.071}$ $ 2725.4793\pm0.0016 $
dip+noi+mask $ \ensuremath{3.3644\pm0.0041}$ $ 48.240\pm0.075 $ $ \ensuremath{264.00\pm0.13}$ $ 2725.4797\pm0.0024 $
dip | ck*]{} Halactic mask extended tu cut out $\pm30^\cire$ of thx Galacfic pland. The level of systematics clreespobd to the pessimistic dxpectatiln of caoibretion errors and sky reslbuals.[]{swta-lcbxl="fig:conflevel128+1024"}](60_128+1024_npi_maskext_CSOP.pdf "fig:")![Marghnxlnsed likelihoods, and 68% and 95% contours jf the lagameters $A$, $b_0$, $l_0$, and $E_{0}$ at $N_{\rm side}=128$ (red) and at $N_{\rm side}=1024$ (blue): top leht, dipole only; yop right, dipole+noise; bottlm lfft, dipole+noise+masn; and botton ridyt, dipole+noire+mask+systtmctics. The rgference frequency channel is 60GHx and the noise iw 7.5$\mk$N.arcmin. The mask used here is the [*Pldnck*]{} Gakactic mask exbendev to cut out $\pm30^\circ$ of thx Galactic plane. The level of svstematics correspond ti the pesvimirric exkecvatjon of camibration srrors and wky residuals.[]{data-lanej="dig:conflevel128+1024"}](60_128+1024_nki_maskqxe_badcal_badsky_CROP.pdf "fig:")
[width=1]{}
$N_{\rm sidt}=128$ $A({\rm mK})$ $b_0(^\ciwc)$ $l_0(^\circ)$ $T_{0}({\gm mK})$
--------------------- ----------------------------------------------- --------------------------------- ----------------------------------------------- -------------------------------------
dipole $\enfhrtmanh{3.3644\pm0.0028}$ $48.242\pm0.047$ $\ensutemath{263.999\pm0.070}$ $2725.4793\pm0.0016$
dip+boi $\ensuremath{3.3644\pm0.0028}$ $48.240\pm0.047$ $\enduremath{263.998\pm0.071}$ $2725.4793\pm0.0016$
dip+noi+mcsk $\fnsurematg{3.3644\om0.0041}$ $48.240\pk0.075$ $\ensureiath{264.00\pm0.13}$ $2725.4797\pm0.0024$
dip | ck*]{} Galactic mask extended to cut out the plane. The of systematics correspond calibration and sky residuals.[]{data-label="fig:conflevel128+1024"}](60_128+1024_noi_maskext_CROP.pdf likelihoods, and 68% 95% contours of the parameters $A$, $l_0$, and $T_{0}$ at $N_{\rm side}=128$ (red) and at $N_{\rm side}=1024$ (blue): top dipole only; top right, dipole+noise; bottom left, dipole+noise+mask; and bottom right, dipole+noise+mask+systematics. The frequency is and noise is 7.5$\mu$K.arcmin. The mask used here is the [*Planck*]{} Galactic mask extended to cut out of the Galactic plane. The level of systematics to the pessimistic expectation calibration errors and sky residuals.[]{data-label="fig:conflevel128+1024"}](60_128+1024_noi_maskext_badcal_badsky_CROP.pdf [width=1]{} side}=128$ $A({\rm $b_0(^\circ)$ $T_{0}({\rm --------------------- ----------------------------------------------- --------------------------------- ------------------------------------- dipole $\ensuremath{3.3644\pm0.0028}$ $48.242\pm0.047$ $\ensuremath{263.999\pm0.070}$ $2725.4793\pm0.0016$ dip+noi $\ensuremath{3.3644\pm0.0028}$ $48.240\pm0.047$ $\ensuremath{263.998\pm0.071}$ $2725.4793\pm0.0016$ dip+noi+mask $\ensuremath{3.3644\pm0.0041}$ $48.240\pm0.075$ $\ensuremath{264.00\pm0.13}$ $2725.4797\pm0.0024$ dip | ck*]{} Galactic mask extended to cUt out $\pm30^\cirC$ of thE GaLacTiC plaNe. ThE level of systemATics Correspond to the pessimiStic eXpECtatIOn Of calIbratioN ErRORs aNd SkY reSiDUaLs.[]{datA-laBel="fig:cOnflevel128+1024"}](60_128+1024_noI_maSkExt_CROP.pdf "fiG:")![maRginalised LikElihoods, and 68% aNd 95% cOntourS oF thE ParamEteRs $A$, $b_0$, $l_0$, And $T_{0}$ at $n_{\Rm side}=128$ (Red) and at $N_{\Rm SIde}=1024$ (bluE): Top left, DIPoLe onLy; top right, dipole+nOIsE; Bottom left, dipoLe+noisE+mASk; AND boTtoM right, dipoLe+Noise+MAsk+systEMaTICS. ThE Reference freqUency channeL Is 60Ghz and tHe NoiSE is 7.5$\mu$K.ArcmiN. THE maSk used here iS the [*planck*]{} GalActic mASk extenDEd to cut Out $\pm30^\cIrc$ Of tHe GaLAcTiC plAnE. the LEvEl oF SysTematics CoRrEsponD to tHE PESsimIstIc exPectaTion of calibraTioN errORs aNd sky ResidUals.[]{DaTa-labEl="fig:cOnfleVeL128+1024"}](60_128+1024_noi_maskext_badcAl_baDsky_CROP.pDf "fIg:")
[WidTh=1]{}
$n_{\rm siDE}=128$ $A({\rm mK})$ $B_0(^\ciRc)$ $l_0(^\Circ)$ $T_{0}({\rm MK})$
--------------------- ----------------------------------------------- --------------------------------- ----------------------------------------------- -------------------------------------
dipolE $\EnsUrEMATh{3.3644\Pm0.0028}$ $48.242\pm0.047$ $\ensuremath{263.999\pm0.070}$ $2725.4793\pm0.0016$
DiP+NOi $\EnsuremaTh{3.3644\pm0.0028}$ $48.240\pm0.047$ $\ENsUrEMath{263.998\pm0.071}$ $2725.4793\pm0.0016$
DiP+noI+masK $\ENsureMath{3.3644\PM0.0041}$ $48.240\pM0.075$ $\ensuremAth{264.00\pm0.13}$ $2725.4797\pM0.0024$
DiP | ck*]{} Galactic mask exten ded to cut out$\p m30 ^\ circ $ of the Galacticp lane . The level of systema ticsco r resp o nd to t he pess i mi s t icex pe cta ti o nof ca lib rationerrors and sk yresiduals.[] { da ta-label=" fig :conflevel12 8+1 024"}] (6 0_1 2 8+102 4_n oi_ma skext_ C ROP.pd f "fig:") ![ M argina l ised li k e li hood s, and 68% and 95 % c o ntours of theparame te r s$ A $,$b_ 0$, $l_0$, a nd $T _ {0}$ at $N _ { \ rms ide}=128$ (re d) and at $ N _{\ rm sid e} =10 2 4$ (bl ue):to p le ft, dipoleonly ; top rig ht, di p ole+noi s e; bott om lef t,dip ole+ n oi se +ma sk ; an d b ott o m r ight, di po le +nois e+ma s k + s yste mat ics. Thereference fre que ncyc han nel i s 60G Hz a nd thenoiseis 7. 5$ \mu$K.arcmin. T he m ask usedher eisth e [*P l anck*] {}Gal actic m ask ext e nde dt o cu t out $\pm30^\circ $o f t he Galac tic pl a ne .T he level o f s yste m a ticscorr e sp ond to t he pes s im is tic exp ec tation o f c ali brati o n er rors a nd sky r esidu a ls.[]{data-lab e l="fig:confle v el 1 2 8+ 1 024" }]( 60_128+1024 _noi _ mask ext_ b ad cal _ badsk y_CRO P. p df "fig:")
[width=1]{ }
$N _{\rm side}=128$ $A( {\rm mK } )$ $b_0(^\c irc)$ $l _0( ^\c irc ) $ $T _{0 }({\rmmK} )$ - --- -- --------- ------ - -- -- -- -- --- ----- - -------- -- --- -- --- ----- - ---------- ---- -- -- - --- ------- - -- - - ---- - -- ---- --- -- ----- ---- - --- ------- --------- --- - -- - -- -- ------- ------------- -- ----------
d i p ole $\ensuremat h {3.3644 \pm 0.002 8}$ $ 48. 2 42\pm0 .047$ $ \ en sur em ath{263.99 9 \ pm0 .070} $ $2725.4793\p m 0.0 016$
di p+no i $\e n s uremath{3.3644\ pm0.0028}$ $ 48 .240\pm 0.047$ $\ens uremath{2 63 .998 \ p m0. 071}$ $272 5 .4793 \ pm 0.001 6$ di p+n oi+ma sk $\ens ur emath{ 3.364 4\ pm0.0041 }$ $48 .240\p m0.07 5$ $\ensur emat h{264.00\p m0. 13} $ $ 272 5 .4797 \pm0 . 0024$
di p | ck*]{} Galactic_mask extended_to cut out $\pm30^\circ$_of the_Galactic_plane. The_level_of systematics correspond_to the pessimistic_expectation of calibration errors_and sky residuals.[]{data-label="fig:conflevel128+1024"}](60_128+1024_noi_maskext_CROP.pdf_"fig:")![Marginalised_likelihoods, and 68% and 95% contours of the parameters $A$, $b_0$, $l_0$, and $T_{0}$_at_$N_{\rm side}=128$_(red)_and_at $N_{\rm side}=1024$ (blue): top_left, dipole only; top right,_dipole+noise; bottom_left, dipole+noise+mask; and bottom right, dipole+noise+mask+systematics. The reference_frequency_channel is 60GHz_and the noise is 7.5$\mu$K.arcmin. The mask used here_is the [*Planck*]{} Galactic mask extended_to cut out_$\pm30^\circ$_of_the Galactic plane. The_level of systematics correspond to the_pessimistic expectation of calibration errors and_sky residuals.[]{data-label="fig:conflevel128+1024"}](60_128+1024_noi_maskext_badcal_badsky_CROP.pdf "fig:")
[width=1]{}
$N_{\rm side}=128$_ _ _ _ _ $A({\rm mK})$_ _ _ _ _ __ ___ $b_0(^\circ)$_ __ __ __ _ __ $l_0(^\circ)$ _ _ _ _ _ $T_{0}({\rm_mK})$
--------------------- ----------------------------------------------- --------------------------------- -----------------------------------------------_-------------------------------------
_ __ dipole__ _ _ _ $\ensuremath{3.3644\pm0.0028}$ _ __ _ $48.242\pm0.047$_ ___ _ $\ensuremath{263.999\pm0.070}$_ _ __ _$2725.4793\pm0.0016$
_ ___dip+noi _ _ $\ensuremath{3.3644\pm0.0028}$ _ _ ___ $48.240\pm0.047$ _ _ _ $\ensuremath{263.998\pm0.071}$ _ _ $2725.4793\pm0.0016$
_ dip+noi+mask _ _ $\ensuremath{3.3644\pm0.0041}$ __ ____ _ $48.240\pm0.075$_ _ _ _ _ __ $\ensuremath{264.00\pm0.13}$ _ $2725.4797\pm0.0024$
_ dip |
charge unit ($\alpha=e^2/(4\pi)$, $e<0$) are used in the paper.
Hyperfine splitting in muonic atoms {#s:ked}
===================================
The ground-state hyperfine splitting in muonic atoms can be written in the form: $$\label{E:STS}
\varDelta E=\varDelta E_{\mathrm{NS}}+\varDelta E_{\mathrm{BW}}+\varDelta E_{\mathrm{QED}},$$ where $\varDelta E_{\mathrm{NS}}$ is the hyperfine splitting value incorporating the relativistic and nuclear charge distribution (“nuclear size”) effects, $\varDelta E_{\mathrm{BW}}$ is the BW contribution, $\varDelta E_{\mathrm{QED}}$ is the QED correction. The $\varDelta E_{\mathrm{NS}}$ value can be calculated by the formula: $$\label{E:NS}
\varDelta E_{\mathrm{NS}}=-\alpha\frac{4}{3}\frac{\mu}{\mu_{N}}\frac{1}{m_p}\frac{(2I+1)}{2I}
\int\limits_{0}^{\infty}dr \: g(r)f(r),$$ where $\alpha$ is the fine structure constant, $\mu$ is the nuclear magnetic moment, $\mu_N$ is the nuclear magneton, $m_p$ is the proton mass, and $I$ is the nuclear spin. $g(r)$ and $f(r)$ are the radial parts of the Dirac wave function: $$\Psi(\textbf{r})=
\begin{pmatrix}
g(r)\varOmega_{\kappa m}(\textbf{n}) \\
\mathit{i}f(r)\varOmega_{-\kappa m}(\textbf{n})
\end{pmatrix},$$ which are determined by solving the Dirac equation with the Fermi distribution of the nuclear charge ($4\pi\int dr r^{2}\rho(r)=1$): $$\rho (r)=\frac{\rho_{0}}{1+\mathrm{exp}(\frac{r-c}{a})}.$$ Here $c$ is the half-density radius and $a$ is related to the skin thickness $t$ by $t=(4\log 3)a$, defined as the distance over which the charge density falls from 90% to 10% of its maximum value.
The individual contributions to $\varDelta E$ for muonic atoms of $^{203}$Tl, | charge unit ($ \alpha = e^2/(4\pi)$, $ e<0 $) are used in the paper.
Hyperfine splitting in muonic atom { # south: ked }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
The ground - state hyperfine splitting in muonic atoms can be write in the form: $ $ \label{E: STS }
\varDelta E=\varDelta E_{\mathrm{NS}}+\varDelta E_{\mathrm{BW}}+\varDelta E_{\mathrm{QED}},$$ where $ \varDelta E_{\mathrm{NS}}$ is the hyperfine splitting value incorporating the relativistic and nuclear bang distribution (“ nuclear size ”) effect, $ \varDelta E_{\mathrm{BW}}$ is the BW contribution, $ \varDelta E_{\mathrm{QED}}$ is the QED correction. The $ \varDelta E_{\mathrm{NS}}$ value can be calculate by the formula: $ $ \label{E: NS }
\varDelta E_{\mathrm{NS}}=-\alpha\frac{4}{3}\frac{\mu}{\mu_{N}}\frac{1}{m_p}\frac{(2I+1)}{2I }
\int\limits_{0}^{\infty}dr \: g(r)f(r),$$ where $ \alpha$ is the fine social organization constant, $ \mu$ is the nuclear magnetic moment, $ \mu_N$ is the nuclear magneton, $ m_p$ is the proton mass, and $ I$ is the nuclear spin. $ g(r)$ and $ f(r)$ are the radial character of the Dirac wave function: $ $ \Psi(\textbf{r})=
\begin{pmatrix }
g(r)\varOmega_{\kappa m}(\textbf{n }) \\
\mathit{i}f(r)\varOmega_{-\kappa m}(\textbf{n })
\end{pmatrix},$$ which are determined by solving the Dirac equation with the Fermi distribution of the nuclear charge ($ 4\pi\int dr r^{2}\rho(r)=1 $ ): $ $ \rho (r)=\frac{\rho_{0}}{1+\mathrm{exp}(\frac{r - c}{a})}.$$ Here $ c$ is the half - concentration radius and $ a$ is related to the skin thickness $ t$ by $ t=(4\log 3)a$, define as the distance over which the charge density fall from 90% to 10% of its maximum value.
The individual contribution to $ \varDelta E$ for muonic atoms of $ ^{203}$Tl, | chwrge unit ($\alpha=e^2/(4\pi)$, $e<0$) are used in the pakee.
Hyperhine spmitting kn muonic atoms {#s:ked}
===================================
The grouid-stqte htperfine splitting in ouonic atlms can ve wcitten in the focj: $$\label{C:FTS}
\vzvDeltc X=\varDelta E_{\mathtm{NS}}+\varDelta E_{\mathrm{BW}}+\varDalga E_{\mathrm{QED}},$$ where $\varDelta E_{\mathrm{NF}}$ is thr jyperfine splijting dalus incorporating the relativistic ahd nucltar charge distribition (“nuclear size”) effects, $\varFelta E_{\mathrm{BW}}$ is the BW conjdibtrion, $\varDeltx E_{\mathrm{QTD}}$ is the QES correction. The $\varDelta E_{\mathfm{NS}}$ ralue can bg dapwulated by vhe fogmula: $$\label{E:NS}
\varDeltd E_{\mathtm{NS}}=-\alpha\frac{4}{3}\fvac{\mu}{\ku_{N}}\drac{1}{m_p}\frac{(2I+1)}{2I}
\int\limits_{0}^{\mnfty}dr \: g(r)f(r),$$ where $\wlpha$ is dhz fine structure consrabt, $\mu$ is dhe vycldar megnstic mlmeit, $\mu_N$ is tge nuclear nagneton, $m_p$ is the kroein mass, and $I$ is thq guclear spin. $g(r)$ and $f(r)$ are the radial pdrta of the Dirac wave funxtion: $$\Psi(\textbf{r})=
\begin{kmatrix}
g(r)\varOmega_{\kappa m}(\textbf{n}) \\
\mathit{i}f(r)\varOmega_{-\ka[pa m}(\vebtby{k})
\cnd{poqtgix},$$ which are determined by solving the Dirac sqianion with the Ferii distribuyiln jf the nucleat chargz ($4\pj\int dr r^{2}\rho(r)=1$): $$\rho (g)=\frac{\rhj_{0}}{1+\mathem{exp}(\frac{w-c}{a})}.$$ Nere $c$ is the half-density rqdius and $a$ ps rwlated to the skin thickness $c$ by $t=(4\kog 3)a$, defined as the distancz over which the fharge dehrity falls from 90% to 10% mf its maximum value.
The inqividual rontrnbutions to $\farDelea E$ for mkonic atoms of $^{203}$Tl, | charge unit ($\alpha=e^2/(4\pi)$, $e<0$) are used in Hyperfine in muonic {#s:ked} =================================== The atoms be written in form: $$\label{E:STS} \varDelta E_{\mathrm{NS}}+\varDelta E_{\mathrm{BW}}+\varDelta E_{\mathrm{QED}},$$ where $\varDelta E_{\mathrm{NS}}$ the hyperfine splitting value incorporating the relativistic and nuclear charge distribution (“nuclear size”) $\varDelta E_{\mathrm{BW}}$ is the BW contribution, $\varDelta E_{\mathrm{QED}}$ is the QED correction. The E_{\mathrm{NS}}$ can calculated the formula: $$\label{E:NS} \varDelta E_{\mathrm{NS}}=-\alpha\frac{4}{3}\frac{\mu}{\mu_{N}}\frac{1}{m_p}\frac{(2I+1)}{2I} \int\limits_{0}^{\infty}dr \: g(r)f(r),$$ where $\alpha$ is the fine structure constant, $\mu$ the nuclear magnetic moment, $\mu_N$ is the nuclear $m_p$ is the proton and $I$ is the nuclear $g(r)$ $f(r)$ are radial of Dirac wave function: \begin{pmatrix} g(r)\varOmega_{\kappa m}(\textbf{n}) \\ \mathit{i}f(r)\varOmega_{-\kappa m}(\textbf{n}) \end{pmatrix},$$ which are determined by solving the Dirac equation with the distribution of charge ($4\pi\int r^{2}\rho(r)=1$): (r)=\frac{\rho_{0}}{1+\mathrm{exp}(\frac{r-c}{a})}.$$ $c$ is the and $a$ is related to the by $t=(4\log 3)a$, defined as the distance over the charge falls from 90% to 10% of maximum value. The individual contributions to $\varDelta E$ muonic atoms of $^{203}$Tl, | charge unit ($\alpha=e^2/(4\pi)$, $e<0$) are useD in the papeR.
HypeRfiNe sPlIttiNg in Muonic atoms {#s:keD}
===================================
the gRound-state hyperfine splIttinG iN MuonIC aToms cAn be wriTTeN IN thE fOrM: $$\laBeL{e:StS}
\vardelTa E=\varDElta E_{\mathrM{NS}}+\VaRDelta E_{\mathrM{bW}}+\VarDelta E_{\mAthRm{QED}},$$ where $\vaRDeLta E_{\maThRm{Ns}}$ Is the HypErfinE splitTIng valUe incorpoRaTIng the RElativiSTIc And nUclear charge distrIBuTIon (“nuclear size”) EffectS, $\vARDELTa E_{\MatHrm{BW}}$ is the bW ContrIBution, $\vARDELTA E_{\mAThrm{QED}}$ is the QeD correctioN. the $\VarDelTa e_{\maTHrm{NS}}$ vAlue cAn BE caLculated by tHe foRmula: $$\labeL{E:NS}
\vaRdelta E_{\mAThrm{NS}}=-\aLpha\frAc{4}{3}\fRac{\Mu}{\mu_{n}}\FrAc{1}{M_p}\fRaC{(2i+1)}{2I}
\iNT\lImiTS_{0}^{\inFty}dr \: g(r)f(R),$$ wHeRe $\alpHa$ is THE FIne sTruCturE consTant, $\mu$ is the nuCleAr maGNetIc momEnt, $\mu_n$ is tHe NucleAr magnEton, $m_P$ iS the proton mass, aNd $I$ iS the nucleAr sPiN. $g(r)$ AnD $f(r)$ arE The radIal ParTs of the dirac waVE fuNcTION: $$\PSi(\textbf{r})=
\begin{pmatRiX}
G(R)\vArOmega_{\kAppa m}(\tEXtBf{N}) \\
\Mathit{i}f(R)\vArOMega_{-\KAPpa m}(\tExtbF{N})
\eNd{pmatriX},$$ which ARe DeTermineD bY solviNg The dirAc equATion With thE Fermi diStribUTion of the nucleAR charge ($4\pi\int dR R^{2}\rHO(R)=1$): $$\rHO (r)=\frAc{\rHo_{0}}{1+\mathrm{exp}(\Frac{R-C}{a})}.$$ HeRe $c$ iS ThE haLF-densIty raDiUS aND $a$ is related to the skiN tHickneSs $t$ by $T=(4\log 3)a$, defined aS the distanCE OVer which The cHArGE density falls fRom 90% to 10% Of its maximUM value.
ThE indiVidual coNtributioNS To $\varDelTa E$ For MuoNic ATOmS of $^{203}$Tl, | charge unit ($\alpha=e^2/ (4\pi)$, $ e<0$) ar e u se d in the paper.
Hyper f inesplitting in muonic at oms { #s : ked} == ===== ======= = == = = === == == === == = == =
Th e g round-s tate hyper fin esplitting in mu onic atoms ca n be written in the f or m:$ $\lab el{ E:STS }
\var D elta E =\varDelt aE _{\mat h rm{NS}} + \ va rDel ta E_{\mathrm{BW} } +\ v arDelta E_{\ma thrm{Q ED } }, $ $ wh ere $\varDelt aE_{\m a thrm{NS } }$ i s th e hyperfine sp litting val u e i ncorpo ra tin g the r elati vi s tic and nuclea r ch arge dist ributi o n (“nuc l ear siz e”) ef fec ts, $\v a rD el taE_ { \ma t hr m{B W }}$ is theBW c ontri buti o n , $\va rDe ltaE_{\m athrm{QED}}$istheQ EDcorre ction . Th e$\var DeltaE_{\m at hrm{NS}}$ value can be calcu lat ed by t he fo r mula:$$\ lab el{E:NS }
\varD e lta E _ { \ ma thrm{NS}}=-\alpha\ fr a c {4 }{3}\fra c{\mu} { \m u_ { N}}\frac {1 }{m _p}\ f r ac{(2 I+1) } {2 I}
\int\ limits _ {0 }^ {\infty }d r \: g (r )f( r), $$ wh e re $ \alpha $ is the fine structure cons t ant, $\mu$ is th e nu c lear ma gnetic mome nt,$ \mu_ N$ i s t hen uclea r mag ne t on , $m_p$ is the proto nmass,and $ I$ is the nuc lear spin. $ g (r)$ and $f( r )$ are the radial part s of the D i rac wave func tion: $$ \Psi(\tex t b f{r})=
\ beg in{ pma tri x }
g(r)\varOm e g a_{\ ka ppa m}( \te xtbf{n} ) \ \
\ma th it{i}f(r) \varOmeg a_ {- \k ap pam}(\t e xtbf{n})
\ end {p mat rix}, $ $ whic h are det er mi n edby solv i ng t he D ir ac equ ati on with the Fer mi dist ributionoft he n uc le ar char ge ($4\pi\int d r r^{2}\rh o( r)= 1$): $ $ \ rho (r)= \frac{\rho_{0}}{1+\math r m{exp}( \fr ac{r- c}{a })}.$$ He re$c$ is th e half- densit y rad iu s a n d $a$i s r ela te d to the s k i n t hickn es s $t $ by $t =(4\log 3)a$, defi n edas the distan ceover w hi cht he cha rg e de n s ity falls from90% to 10% o f i ts maximum val ue .
Theindivid ual c o ntribut ions to $ \varDelta E $ fo r muo nic atomsof $^{20 3}$Tl, | charge_unit ($\alpha=e^2/(4\pi)$,_$e<0$) are used in_the paper.
Hyperfine_splitting_in muonic_atoms_{#s:ked}
===================================
The ground-state hyperfine_splitting in muonic_atoms can be written_in the form:_$$\label{E:STS}
\varDelta_E=\varDelta E_{\mathrm{NS}}+\varDelta E_{\mathrm{BW}}+\varDelta E_{\mathrm{QED}},$$ where $\varDelta E_{\mathrm{NS}}$ is the hyperfine splitting value incorporating the_relativistic_and nuclear_charge_distribution_(“nuclear size”) effects, $\varDelta E_{\mathrm{BW}}$_is the BW contribution, $\varDelta_E_{\mathrm{QED}}$ is_the QED correction. The $\varDelta E_{\mathrm{NS}}$ value can_be_calculated by the_formula: $$\label{E:NS}
\varDelta E_{\mathrm{NS}}=-\alpha\frac{4}{3}\frac{\mu}{\mu_{N}}\frac{1}{m_p}\frac{(2I+1)}{2I}
\int\limits_{0}^{\infty}dr \: g(r)f(r),$$ where $\alpha$ is the_fine structure constant, $\mu$ is the_nuclear magnetic moment,_$\mu_N$_is_the nuclear magneton, $m_p$_is the proton mass, and $I$_is the nuclear spin. $g(r)$ and_$f(r)$ are the radial parts of the_Dirac wave function: $$\Psi(\textbf{r})=
\begin{pmatrix}
_ g(r)\varOmega_{\kappa m}(\textbf{n}) \\
_ _\mathit{i}f(r)\varOmega_{-\kappa m}(\textbf{n})
\end{pmatrix},$$ which are determined_by solving the_Dirac equation_with the Fermi_distribution of the nuclear charge ($4\pi\int_dr r^{2}\rho(r)=1$): $$\rho_(r)=\frac{\rho_{0}}{1+\mathrm{exp}(\frac{r-c}{a})}.$$ Here $c$ is the half-density_radius_and $a$ is_related_to_the skin_thickness $t$ by_$t=(4\log_3)a$, defined_as_the distance over which the charge_density_falls from 90% to 10% of its_maximum value.
The individual contributions_to_$\varDelta E$ for muonic_atoms of $^{203}$Tl, |
been derived from the warm and cool IRAS F25/F60 ratios. These ratios, however, refer to the entire host galaxies and are unsuitable to conclusively establish the absence of a nuclear dust torus. Instead, a study of the Seyfert-2 dichotomy should be performed on the basis of nuclear properties only. Here we present the first comparison between \[OIII\]$_{\lambda 5007 \AA}$ and mid-infrared imaging at matching spatial resolution. The aim is to check whether the nuclear dust emission scales with AGN luminosity as traced by \[OIII\]. ]{} [During the scientific verification phase of the VISIR instrument at the ESO Very Large Telescope we observed 16 Sy1 and Sy2 nuclei at 11.25$\mu$m with 0$\farcs$35 spatial resolution (FWHM). We supplement our observations with high-resolution 10–12$\mu$m literature data of 58 Seyfert galaxies, for which spectroscopic or spectropolarimetric information and far-infrared data are available. ]{} [ Twelve of the 15 detected sources are unresolved and three sources show a dominant unresolved core surrounded by some faint knots in an area smaller than 1–2$\arcsec$ radius. Our VISIR photometry agrees to better than 15% with published data obtained at 1$\farcs$5–5$\arcsec$ spatial resolution. Exploring the Seyfert-2 dichotomy we find that the distributions of nuclear mid-infrared/\[OIII\] luminosity ratios are indistinguishable for Sy1s and Sy2s with and without detected polarised broad lines and irrespective of having warm or cool IRAS F25/F60 ratios. We find no evidence for the existence of a population of real Sy2s with a deficit of nuclear dust emission. Our results suggest 1) that all Seyfert nuclei possess the same physical structure including the putative dust torus and 2) that the cool IRAS colours are caused by a low contrast of AGN to host galaxy. Then the Seyfert-2 dichotomy is explained in part by unification of non-HBLRs with narrow-line Sy1s and to a larger rate by observational biases caused by a low AGN/host contrast and/or an unfavourable scattering geometry. ]{}
Introduction
============
The nuclei of Seyfert galaxies are grouped into Sy1 and Sy2 depending on the presence of broad emission lines in their optical spectra (Khachikian & Weedman 1974). According to the unified model Sy1 and Sy2 nuclei differ only with respect | been derived from the warm and cool IRAS F25 / F60 ratios. These ratios, however, denote to the integral host galaxies and are inapplicable to conclusively lay down the absence of a nuclear dust torus. Instead, a sketch of the Seyfert-2 dichotomy should be do on the basis of nuclear property only. Here we present the beginning comparison between \[OIII\]$_{\lambda 5007 \AA}$ and mid - infrared imaging at matching spatial resoluteness. The purpose is to check whether the nuclear dust emission scales with AGN luminosity as traced by \[OIII\ ]. ] { } [ During the scientific confirmation phase of the VISIR instrument at the ESO Very Large Telescope we observed 16 Sy1 and Sy2 nucleus at 11.25$\mu$m with 0$\farcs$35 spatial resolution (FWHM). We supplement our notice with high - resolving power 10–12$\mu$m literature data of 58 Seyfert galaxies, for which spectroscopic or spectropolarimetric information and far - infrared data are available. ] { } [ Twelve of the 15 detected source are unresolved and three sources show a dominant unresolved effect surrounded by some faint knots in an area smaller than 1–2$\arcsec$ radius. Our VISIR photometry agrees to better than 15% with published data obtain at 1$\farcs$5–5$\arcsec$ spatial resolution. Exploring the Seyfert-2 dichotomy we receive that the distribution of nuclear mid - infrared/\[OIII\ ] luminosity ratios are indistinguishable for Sy1s and Sy2s with and without detected polarize broad lines and irrespective of having quick or cool IRAS F25 / F60 ratios. We find no evidence for the existence of a population of real Sy2s with a deficit of nuclear dust emission. Our solution suggest 1) that all Seyfert nuclei possess the same forcible structure admit the putative debris torus and 2) that the cool IRAS color are caused by a low contrast of AGN to host galax. Then the Seyfert-2 dichotomy is explained in part by unification of non - HBLRs with narrow - line Sy1s and to a larger rate by observational biases induce by a low AGN / master of ceremonies contrast and/or an unfavourable disperse geometry. ] { }
Introduction
= = = = = = = = = = = =
The nuclei of Seyfert galaxy are grouped into Sy1 and Sy2 depend on the presence of across-the-board emission lines in their optical spectra (Khachikian & Weedman 1974). According to the coordinated model Sy1 and Sy2 nuclei differ only with respect | befn derived from the warm and cool IRAS Y25/D60 ratims. Theae ratior, however, refer to the entirx howt gaoaxies and are unsuitacle to cojclusiveoy ewrablish thx absencc of z nucnxar dust torus. Lnstead, a sdudy of the Seffdrc-2 dichotomy should be performed on tre basix lf nuclear prokertits jnly. Here we present the first compariaon betxeen \[OIII\]$_{\lambda 5007 \AA}$ and mid-infrared imagijg ah matching spatial resolution. The qim is to chdck whetheg the nuclezr dust emission scales with AGV lumnnosity as jxqcef by \[OIII\]. ]{} [Dnring nhe scientifig verifhcation phase of the YISIR inwtrument at the ESO Vxry Large Telescope re observad 16 Sy1 and Sy2 nucleu qt 11.25$\mu$k widh 0$\fxecs$35 spztmal resolktikn (FWHM). We supplement our observations wotr high-resolutikn 10–12$\mu$m lyterature data of 58 Seyfert galaxies, for whjch spectroscopic or spwctropolarimetric inflrmation wnd far-infrared data are available. ]{} [ Twelve of tha 15 devezteb sourzws are unresolved and three sources show a domigznu ukresolved core smrrounded by some galny knots in an xrea smalmer than 1–2$\arcsec$ rafius. Out VISIE photomeury abrees to better than 15% with published dana ovtained at 1$\farcs$5–5$\aresec$ spatial resplutipn. Exploring the Seyferc-2 dichktomy we fijd that tgd distributions uf kucnear mid-infrared/\[OIII\] luminjsity ratmos axe indisgingoishablq for Sy1s wnd Sn2v with and without deteeted [olarised hroad lines and irrespective of having warm ot cmol IRAS F25/F60 ratips. We find no evidence for jhe existznce ow a populanion of rxal Sy2s with a deficit of juclear dust emissiog. Oue rewults sjegest 1) that alk Seyfert nuclei powsess the same phyxicxm structure incoudung the putativr djst tlrns anq 2) that the cmol KRAR colojrs are cauwdd bu a low contrast of DGN fo host galaxy. Them bhe Seyfett-2 dichotjmy is explaimed in part by univicatmon of non-HNLRf with narrow-line Sy1s and to a larger rwte by observatyonao biases cauxed by a low AGN/host contrast and/or an nnfavourable scattering geometry. ]{}
Introductnok
============
The nuclei pf Serfert galdxies are grouped inro Sy1 and Sy2 depekding on the presence kf brodd emlssion lines in their optical spectra (Khachikian & Weedman 1974). According to tye unihiqd model Sy1 and Sy2 nlclzi biffer jnly xith respect | been derived from the warm and cool ratios. ratios, however, to the entire to establish the absence a nuclear dust Instead, a study of the Seyfert-2 should be performed on the basis of nuclear properties only. Here we present first comparison between \[OIII\]$_{\lambda 5007 \AA}$ and mid-infrared imaging at matching spatial resolution. aim to whether nuclear dust emission scales with AGN luminosity as traced by \[OIII\]. ]{} [During the scientific verification of the VISIR instrument at the ESO Very Telescope we observed 16 and Sy2 nuclei at 11.25$\mu$m 0$\farcs$35 resolution (FWHM). supplement observations high-resolution 10–12$\mu$m literature of 58 Seyfert galaxies, for which spectroscopic or spectropolarimetric information and far-infrared data are available. ]{} [ of the sources are and sources a dominant unresolved by some faint knots in an 1–2$\arcsec$ radius. Our VISIR photometry agrees to better 15% with data obtained at 1$\farcs$5–5$\arcsec$ spatial resolution. the Seyfert-2 dichotomy we find that the distributions nuclear mid-infrared/\[OIII\] luminosity ratios are indistinguishable for Sy1s and Sy2s with and without detected polarised and irrespective of having or cool IRAS ratios. find evidence the existence a population of real Sy2s with a deficit of nuclear dust Our results suggest 1) that all Seyfert nuclei possess the structure the putative dust and 2) that the IRAS are caused by a of to the dichotomy explained in part by of non-HBLRs with narrow-line Sy1s to a larger rate a low AGN/host contrast and/or an unfavourable scattering ]{} Introduction ============ The nuclei of Seyfert are grouped into Sy1 and Sy2 depending on the presence of broad lines in spectra (Khachikian & Weedman 1974). According to the model Sy1 and Sy2 differ only with respect | been derived from the warm and Cool IRAS F25/F60 RatioS. ThEse RaTios, HoweVer, refer to the eNTire Host galaxies and are unsuItablE tO ConcLUsIvely EstabliSH tHE AbsEnCe Of a NuCLeAr dusT toRus. InstEad, a study oF thE SEyfert-2 dichotOMy Should be peRfoRmed on the basIs oF nucleAr ProPErtieS onLy. HerE we preSEnt the First compArISon betWEen \[OIII\]$_{\LAMbDa 5007 \AA}$ And mid-infrared imaGInG At matching spatIal resOlUTiON. the Aim Is to check wHeTher tHE nucleaR DuST EMisSIon scales with aGN luminosiTY as Traced By \[oIIi\]. ]{} [during The scIeNTifIc verificatIon pHase of the vISIR iNStrumenT At the ESo Very LArgE TeLescOPe We ObsErVEd 16 SY1 AnD Sy2 NUclEi at 11.25$\mu$m wItH 0$\fArcs$35 sPatiAL RESoluTioN (FWHm). We suPplement our obSerVatiONs wIth hiGh-resOlutIoN 10–12$\mu$m lIteratUre daTa Of 58 Seyfert galaxiEs, foR which speCtrOsCopIc Or speCTropolAriMetRic infoRmation ANd fAr-INFRaRed data are availablE. ]{} [ TWELvE of the 15 deTected SOuRcES are unreSoLveD and THRee soUrceS ShOw a dominAnt unrESoLvEd core sUrRoundeD bY soMe fAint kNOts iN an areA smaller Than 1–2$\aRCsec$ radius. Our VisIR photometry AGrEES tO BettEr tHan 15% with publIsheD Data ObtaINeD at 1$\FArcs$5–5$\aRcsec$ SpATiAL resolution. ExplorinG tHe SeyfErt-2 diChotomy we find That the disTRIButions oF nucLEaR Mid-infrared/\[OIIi\] lumiNosity ratiOS are indiStingUishable For Sy1s and sY2S with and WitHouT deTecTED pOlarised broad LINes aNd IrrespeCtiVe of havIng WarM or CooL IrAS F25/F60 ratiOs. We find No EvIdEnCe fOr the EXistence Of A poPuLatIon of REal Sy2s With a DefiCiT oF NucLear dusT EmISSion. ouR rEsulTs sUgGest 1) tHat aLL SeYfert nuClei posseSs tHE samE pHySical stRucture includInG the putatiVe DusT torus AND 2) that the Cool IRAS colours are causeD By a low cOntRast oF AGN To host galAxy. then thE SeYFert-2 diChotomY is exPlAinED In parT BY uNifIcAtion of non-hblRs With nArRow-lIne Sy1s aNd to a larger rate by oBSerVational biaseS caUsed BY A lOw Agn/hOSt cOnTRasT ANd/or an unfavouraBle scatterInG GeOmetry. ]{}
IntrODucTiOn
============
The nuClei of SEyferT GalaxieS are groupEd into Sy1 aNd sy2 dePENdiNg on the preSence of bRoad emissIOn linES iN theiR opTical sPeCtrA (KhacHikian & wEedMan 1974). AcCordinG tO the unIfied MoDel Sy1 and sy2 nuclei differ only with rEspect | been derived from the war m and cool IRAS F2 5/F 60 rat ios. These ratios, howe ver, refer to the enti re ho st gala x ie s and are un s ui t a ble t ocon cl u si velyest ablishthe absenc e o fa nuclear du s ttorus. Ins tea d, a study o f t he Sey fe rt- 2 dich oto my sh ould b e perfo rmed on t he basiso f nucle a r p rope rties only. Herew ep resent the fir st com pa r is o n be twe en \[OIII\ ]$ _{\la m bda 500 7 \ A A } $ a n d mid-infrare d imaging a t ma tching s pat i al res oluti on . Th e aim is to che ck whethe r then ucleard ust emi ssionsca les wit h A GN lu mi n osi t yast rac ed by \[ OI II \]. ] {} [ D u r i ng t hescie ntifi c verificatio n p hase ofthe V ISIRinst ru mentat the ESOVe ry Large Telesc opewe observ ed16 Sy 1and S y 2 nucl eiat11.25$\ mu$m wi t h 0 $\ f a r cs $35 spatial resolu ti o n ( FWHM). W e supp l em en t our obs er vat ions w ith h igh- r es olution10–12$ \ mu $m litera tu re dat aof58Seyfe r t ga laxies , for wh ich s p ectroscopic or spectropolari m et r i ci nfor mat ion and far -inf r ared dat a a rea vaila ble.]{ } [ Twelve of the 15 de te cted s ource s are unresol ved and th r e e sources sho w a dominant unres olved core surr o unded by some faint k nots in a n area sma lle r t han 1– 2 $ \a rcsec$ radius . OurVI SIR pho tom etry ag ree s t o b ett er than 15% with pu bl is he ddat a obt a ined at1$ \fa rc s$5 –5$\a r csec$spati al r es ol u tio n. Expl o ri n g the S ey fert -2di choto my w e fi nd that the dist rib u tion sof nuclea r mid-infrare d/ \[OIII\] l um ino sity r a t ios areindistinguishable for S y 1s andSy2 s wit h an d without de tected po l arised broad line sand i rresp e c ti veof having wa r m or cool I RASF25/F60 ratios. We find n o ev idence for th e e xist e n ce of ap opu la t ion o f real Sy2s wit h a defici to fnuclear du s t e mi ssion.Our res ultss uggest1) that a ll Seyfer tnucl e i po ssess thesame phy sical str u cture in cludi ngthe pu ta tiv e dus t toru s an d 2)that t he coolIRASco lours ar e caused by a low contr ast of AGNtohost gala xy. The n the Sey fert -2 dichoto myisexpla ine d in p artb yuni f icati on o f non-HBLR s w ith n ar row-line Sy 1 s and to a la r ger ra te b y observational b i ases caused by a l o w AG N/h o st c on trast and/or a n u nf a v ourablesc attering ge ometry.]{ }
Int roduct ion
== ======= = = =The nu clei of Seyfertgal ax i es aregr ou p ed int o Sy 1and Sy 2 depe n ding o n the presence o f bro a d emis s ion line sin thei r opt ical spect ra (Khachik ian &Weed man 1 974). A cc ording to t he unified model Sy1 andSy2 nuc le i di ffe r only wit h respe ct | been_derived from_the warm and cool_IRAS F25/F60_ratios._These ratios,_however,_refer to the_entire host galaxies_and are unsuitable to_conclusively establish the_absence_of a nuclear dust torus. Instead, a study of the Seyfert-2 dichotomy should be_performed_on the_basis_of_nuclear properties only. Here we_present the first comparison between_\[OIII\]$_{\lambda 5007_\AA}$ and mid-infrared imaging at matching spatial resolution._The_aim is to_check whether the nuclear dust emission scales with AGN_luminosity as traced by \[OIII\]. ]{}_[During the scientific_verification_phase_of the VISIR instrument_at the ESO Very Large Telescope_we observed 16 Sy1 and Sy2_nuclei at 11.25$\mu$m with 0$\farcs$35 spatial resolution_(FWHM). We supplement our observations with_high-resolution 10–12$\mu$m literature data of_58 Seyfert_galaxies, for which spectroscopic or_spectropolarimetric information and_far-infrared data_are available. ]{}_[ Twelve of the 15 detected_sources are unresolved_and three sources show a dominant_unresolved_core surrounded by_some_faint_knots in_an area smaller_than_1–2$\arcsec$ radius._Our_VISIR photometry agrees to better than_15%_with published data obtained at 1$\farcs$5–5$\arcsec$ spatial_resolution. Exploring the Seyfert-2_dichotomy_we find that the_distributions of nuclear mid-infrared/\[OIII\] luminosity_ratios are indistinguishable for Sy1s and_Sy2s with_and without_detected polarised broad lines and irrespective of having warm or cool_IRAS F25/F60 ratios. We find no_evidence for the existence_of a_population_of real Sy2s_with_a deficit_of nuclear dust emission. Our results suggest_1) that_all Seyfert nuclei possess the same_physical structure including the_putative_dust torus and 2) that the_cool IRAS colours are caused by_a low contrast of AGN_to_host_galaxy. Then the Seyfert-2 dichotomy_is explained in part by unification_of non-HBLRs with_narrow-line Sy1s and to a larger rate_by_observational biases caused by a low_AGN/host_contrast and/or an unfavourable scattering geometry._]{}
Introduction
============
The_nuclei_of Seyfert galaxies are grouped_into Sy1 and Sy2 depending on_the presence of broad emission lines in their optical_spectra (Khachikian &_Weedman 1974). According to the_unified_model_Sy1 and Sy2 nuclei differ only with respect |
to 40%. This is found to be due to incorrectly determined scalelengths and isophotal radii, which are used to define the aperture sizes for Kron and total fluxes. While 2MASS metric aperture luminosities are correct (and, thus, colors based on those apertures), comparison to other filters (e.g. optical) based on total magnitudes will produce erroneous results. We use our own galaxy photometry package (ARCHANGEL) to determine correct total magnitudes and colors using the same 2MASS images, but with a more refined surface brightness reduction scheme. Our resulting colors, and color-magnitude relation, are more in line with model expectations and previous pointed observations.'
author:
- James Schombert
title: Systematic Bias in 2MASS Galaxy Photometry
---
Introduction
============
Surface photometry is an important tool in the understanding of galaxy mass and structure. Galaxy luminosity measures the primary baryonic component, (i.e. stellar mass) and structural information traces the gravitational potential. Galaxy formation scenarios make specific predictions on the light distribution of galaxies, so accurate reduction of a galaxy image into a total magnitude and scalelength are important parameters to understanding the fundamental plane and the star formation history of galaxies.
Obtaining the structural characteristics and the total luminosity of a galaxy requires knowledge of its surface brightness profile to a significant depth. In order to extrapolate a total luminosity, isophotal analysis is required to determine how far one needs to integrate a galaxy’s light plus to provide sufficient information to extrapolate the light profile.
During a surface photometry project to explore the structure of galaxies by morphological type (Schombert & Smith 2011), we discovered a significant discrepancy between the structural and luminosity parameters that we determined using raw 2MASS images versus those reported by the 2MASS project in their Extended Source Catalog (Jarrett 2000). We use this letter to outline the problem, and the solution, for other researchers.
Sample
======
Our original project was to perform surface photometry of large galaxies over a full range of morphological types. This type of analysis was previously attempted by 2MASS (Jarrett 2003), but their focus was not on the structural characteristics of galaxies, but rather luminosities and colors. The first stage of our project was to understand the surface photometry of ellipticals, the simplest galaxies for structural studies as they have highly symmetric isophotes and very few complications to their light distributions due to extinction or recent star formation.
| to 40% . This is found to be due to incorrectly determined scalelengths and isophotal radius, which are use to define the aperture sizes for Kron and total flux. While 2MASS metric aperture luminosities are right (and, thus, colors based on those aperture), comparison to other filters (for example optical) based on total magnitudes will grow erroneous results. We use our own galaxy photometry package (ARCHANGEL) to determine right total magnitudes and colors use the same 2MASS images, but with a more polished surface luminosity reduction outline. Our resulting colors, and color - order of magnitude relation, are more in line with model expectations and previous pointed observations.'
writer:
- James Schombert
title: Systematic Bias in 2MASS Galaxy Photometry
---
Introduction
= = = = = = = = = = = =
Surface photometry is an important tool in the understanding of galaxy mass and structure. Galaxy luminosity measures the primary baryonic part, (i.e. leading mass) and structural information trace the gravitational potential. Galaxy formation scenarios do specific predictions on the light distribution of galaxies, so accurate decrease of a galaxy image into a total magnitude and scalelength are important parameters to understanding the fundamental plane and the star formation history of galaxies.
prevail the structural feature and the total luminosity of a galaxy requires knowledge of its surface luminosity profile to a significant depth. In order to extrapolate a total luminosity, isophotal analysis is required to determine how far one want to integrate a galaxy ’s luminosity plus to provide sufficient data to extrapolate the light profile.
During a surface photometry undertaking to explore the structure of galaxies by morphologic character (Schombert & Smith 2011), we discovered a significant discrepancy between the structural and luminosity parameter that we determined using raw 2MASS images versus those reported by the 2MASS undertaking in their Extended Source Catalog (Jarrett 2000). We use this letter to outline the trouble, and the solution, for other researchers.
Sample
= = = = = =
Our original project was to do surface photometry of large galaxies over a full range of morphological types. This type of analysis was previously attempted by 2MASS (Jarrett 2003), but their focus was not on the geomorphologic characteristic of galaxy, but rather luminosities and colors. The first stage of our project was to sympathize the surface photometry of ellipticals, the simplest galaxies for structural studies as they have highly symmetric isophotes and very few complications to their light distributions due to extinction or late star formation. | to 40%. This is found to be dut to incorrectly bwtermiied scamelengthr and isophotal radii, which ere ysed uj define the apertjre sizes for Krob anv total fluxes. Wijle 2MASS metrjg apextnre luminositiex are corract (and, thus, cmlurd based on those apertures), comparisjn to oyhfr filters (e.g. jptibaj) baavd on total magnitudes will prosuce ergoneous results. Wr use our own galaxy photoletrj package (ARCHANGEP) to determune sirrect total magnitudes and colors using the same 2MASS images, but with a more redibed vurface brijhtnesf reduction scheme. Ogr resukting colors, akd conor-nagnitude relation, arx more in line with iodel expaccations and previous pounted obsarvaguonr.'
auuhoc:
- Jzmes Sfhojbert
title: Systematic Bias in 2MASS Galaxu [yotometry
---
Intrkductijn
============
Furface photometry is an important tool in the understanding of gqlaxy mass and structore. Galaxy luminosity measures the primary baryonic componett, (i.e. rteolwf mwss) and structural information traces the gradjtstponal potential. Gclaxy formation xcfnstios make speckfic pxsdjctions on the ligjt disttibutiin of galwxiex, so accurate reduction of q galaxy imaye unto a total magnicude and scauelemgth sre important parameterr to understandlng the fhvdamental plane xnd tve star formation history jf galaxixs.
Obtcining tfe sjructurwl charactfristlws and the total lkminodidy of a gapaxy requires knowledge of its surface brightmevs irofile tj a slgnificant deptr. In order to gxtrapolace a tutal luminksity, ivophotal anwlysis is reqglred to detecmine how far one needs gu integrate a balaxy’s lpgkt plus ti provide sufficiekt inwkrmation to extxcpilate the light prufije.
Fucing w surface phodomegry lrojezt to explove ghe xtructure of galaxiev by morphological typr (Fchomberj & Smith 2011), we discoverec a significant didcrepency bxtween thg structural and luminosity parzmeters tjat we determinqd uwing raw 2MASX images versus those reported by the 2MESS project in their Exrended Source Catalpn (Jarrett 2000). Wx use ehis lettar to outline the priblem, and the solmtion, for other researdhers.
Sdmple
======
Lur original project was to perform surface photometry of large galaxies iver a full range kf mprphonoyiccl typef. Thms type of analysix was previously attempted by 2MESS (Jarretd 2003), but their focus was not on yhd structural zharacteristics of galaxjes, but tather luminosities and colors. Yhe first stage of our pvoject waw to hndersnznv the surface photomrtry oy ellipticals, the simplest galxxirs for strufturel atudies as they have highly symmwtric uslphotes and very few compligauions tk theit light distributjons fue tu extinction pr recent star formation.
| to 40%. This is found to be incorrectly scalelengths and radii, which are sizes Kron and total While 2MASS metric luminosities are correct (and, thus, colors on those apertures), comparison to other filters (e.g. optical) based on total magnitudes produce erroneous results. We use our own galaxy photometry package (ARCHANGEL) to determine total and using same 2MASS images, but with a more refined surface brightness reduction scheme. Our resulting colors, and relation, are more in line with model expectations previous pointed observations.' author: James Schombert title: Systematic Bias 2MASS Photometry --- ============ photometry an important tool the understanding of galaxy mass and structure. Galaxy luminosity measures the primary baryonic component, (i.e. stellar mass) structural information gravitational potential. formation make predictions on the of galaxies, so accurate reduction of into a total magnitude and scalelength are important to understanding fundamental plane and the star formation of galaxies. Obtaining the structural characteristics and the luminosity of a galaxy requires knowledge of its surface brightness profile to a significant depth. to extrapolate a total isophotal analysis is to how one to integrate galaxy’s light plus to provide sufficient information to extrapolate the light During a surface photometry project to explore the structure of morphological (Schombert & Smith we discovered a significant between structural and luminosity parameters determined raw those by 2MASS project in their Source Catalog (Jarrett 2000). We this letter to outline for other researchers. Sample ====== Our original project to perform surface photometry of large galaxies a full range of morphological types. This type of analysis was previously by 2MASS but their focus was not on the structural of galaxies, but rather and colors. The first stage of our project was understand surface photometry ellipticals, the simplest for structural studies they have highly and very complications their to extinction or recent star formation. | to 40%. This is found to be due to incOrrectly deTermiNed ScaLeLengThs aNd isophotal radII, whiCh are used to define the apErturE sIZes fOR KRon anD total fLUxES. whiLe 2mAsS mEtRIc ApertUre LuminosIties are coRreCt (And, thus, colorS BaSed on those ApeRtures), comparIsoN to othEr FilTErs (e.g. OptIcal) bAsed on TOtal maGnitudes wIlL ProducE ErroneoUS ReSultS. We use our own galaxY PhOTometry package (aRCHANgEl) To DETerMinE correct toTaL magnITudes anD CoLORS usINg the same 2MASS Images, but wiTH a mOre refInEd sURface bRightNeSS reDuction scheMe. OuR resultinG colorS, And coloR-MagnituDe relaTioN, arE morE In LiNe wItH ModEL eXpeCTatIons and pReViOus poInteD OBSErvaTioNs.'
auThor:
- JAmes Schombert
TitLe: SySTemAtic BIas in 2mASS gaLaxy PHotomeTry
---
InTrOduction
============
Surface PhotOmetry is aN imPoRtaNt Tool iN The undErsTanDing of gAlaxy maSS anD sTRUCtUre. Galaxy luminositY mEASuRes the prImary bARyOnIC componeNt, (I.e. sTellAR Mass) aNd stRUcTural infOrmatiON tRaCes the gRaVitatiOnAl pOteNtial. gAlaxY formaTion scenArios MAke specific preDIctions on the lIGhT DIsTRibuTioN of galaxies, So acCUratE redUCtIon OF a galAxy imAgE InTO a total magnitude and ScAlelenGth arE important parAmeters to uNDERstandinG the FUnDAmental plane anD the sTar formatiON history Of galAxies.
ObtAining the STRuctural ChaRacTerIstICS aNd the total lumINOsitY oF a galaxY reQuires kNowLedGe oF itS sUrface briGhtness pRoFiLe To A siGnifiCAnt depth. in OrdEr To eXtrapOLate a tOtal lUminOsItY, IsoPhotal aNAlYSIs is ReQuIred To dEtErminE how FAr oNe needs To integraTe a GAlaxY’s LiGht plus To provide suffIcIent informAtIon To extrAPOlate the Light profile.
During a surfACe photoMetRy proJect To explore The StructUre OF galaxIes by mOrphoLoGicAL Type (SCHOmBerT & SMith 2011), we discOVEreD a sigNiFicaNt discrEpancy between the stRUctUral and luminoSitY parAMEtErs THaT We dEtERmiNED using raw 2MASS imAges versus ThOSe Reported by THe 2MaSs projecT in theiR ExteNDed SourCe Catalog (jarrett 2000). We UsE thiS LEttEr to outlinE the probLem, and the SOlutiON, fOr othEr rEsearcHeRs.
SAmple
======
our oriGInaL projEct was To PerforM surfAcE photomeTry of large galaxies over a Full raNge of MorPhologicaL tyPEs. THis type of AnalYsis was preVioUslY atteMptED by 2MAsS (JaRReTt 2003), bUT theiR focUS was not on THe StrUCTuRal characteRISTicS of gaLaxIEs, but rAtheR luminosities and cOLors. The first stAge oF OUr pRojECt waS tO understand the SurFaCE PhotometRy Of ellipticaLs, the simPlESt galAxies fOr struCtural sTUDiES as theY havE hiGhly symmeTriC iSOphotes AnD vERy few cOmplIcAtions To theiR LighT DIstributions due tO extiNCTion oR RecEnt stAr FormatiON.
| to 40%. This is found tobe due toincor rec tly d eter mine d scalelengths andisophotal radii, which areus e d to de finethe ape r tu r e si ze sfor K r on andtot al flux es. While2MA SS metric aper t ur e luminosi tie s are correc t ( and, t hu s,c olors ba sed o n thos e apert ures), co mp a risont o other f il ters (e.g. optical) b a se d on total magn itudes w i ll p rod uce erroneous r esult s . We us e o u r own galaxy photom etry packag e (A RCHANG EL ) t o deter mineco r rec t total mag nitu des and c olorsu sing th e same 2 MASS i mag es, but wi th amo r e r e fi ned sur face bri gh tn ess r educ t i o n sch eme . Ou r res ulting colors , a nd c o lor -magn itude rel at ion,are mo re in l ine with modelexpe ctationsand p rev io us po i nted o bse rva tions.'
author :
-Ja m e s S chombert
title: Sy st e m at ic Biasin 2MA S SGa l axy Phot om etr y
-- -
Intr oduc t io n
====== ======
S ur face ph ot ometry i s a n i mport a nt t ool in the und ersta n ding of galaxy mass and stru c tu r e .G alax y l uminosity m easu r es t he p r im ary baryo nic c om p on e nt, (i.e. stellar m as s) and stru ctural inform ation trac e s the grav itat i on a l potential. G alaxy formation scenario s mak e specif ic predic t i ons on t helig htdis t r ib ution of gala x i es,so accura tereducti onofa g ala xy image in to a tot al m ag ni tud e and scalelen gt h a re im porta n t para meter s to u nd e rst andingt he f unda me nt al p lan eand t he s t arformati on histor y o f gal ax ie s.
Obt aining the st ru ctural cha ra cte ristic s and thetotal luminosity of a g a laxy re qui res k nowl edge of i tssurfac e b r ightne ss pro fileto as i gnifi c a nt de pt h. In orde r toextra po late a tota l luminosity, isop h ota l analysis is re quir e d t o d e te r min eh owf a r one needs tointegrateag al axy’s ligh t pl us to pro vide su ffici e nt info rmation t o extrapo la te t h e li ght profil e.
Duri ng a surf a ce ph o to metry pr ojectto ex plore the s t ruc tureof gal ax ies by morp ho logicaltype (Schombert & Smith 2011) , wedis covered a si g nif icant dis crep ancy betwe enthe stru ctu r al an d lu m in osi t y par amet e rs that w e d ete r m in ed using ra w 2 MAS S ima ges versus tho se reported by th e 2MASS project int h eir Ex t ende dSource Catalog (J ar r e tt 2000) .We use this letterto outli ne the probl em, and t he soluti on,for other re sea rc h ers.
S am pl e
===== =
O ur origi nal pr o ject w as to perform su rface p hotom e try of l ar ge gala x iesover a ful l range ofmorpho logi cal t ypes. T hi s type of a nalysis wa s previous ly at tempted b y 2M ASS (Jarr ett2 0 03),butth eir focus wa s no t o nt hestru ctura lchar acteristi c s of gal axi e s, butra the r lumino s it i e s and colo rs. Thef i rst stageo f ou r p r oject was t o unde rstandt hesu rface p hot o m etry of e lliptical s , t he sim plest ga la xies f orst ructural studie s as the y have h i g hly sy mme t rici sophot es and very fe w com pl i cations t o t heir li g htdis tri buti ons due to extinction or recentstar form a tio n.
| to_40%. This_is found to be_due to_incorrectly_determined scalelengths_and_isophotal radii, which_are used to_define the aperture sizes_for Kron and_total_fluxes. While 2MASS metric aperture luminosities are correct (and, thus, colors based on those_apertures),_comparison to_other_filters_(e.g. optical) based on total_magnitudes will produce erroneous results._We use_our own galaxy photometry package (ARCHANGEL) to determine_correct_total magnitudes and_colors using the same 2MASS images, but with a_more refined surface brightness reduction scheme._Our resulting colors,_and_color-magnitude_relation, are more in_line with model expectations and previous_pointed observations.'
author:
- James Schombert
title: Systematic Bias_in 2MASS Galaxy Photometry
---
Introduction
============
Surface photometry is an_important tool in the understanding of_galaxy mass and structure. Galaxy_luminosity measures_the primary baryonic component, (i.e._stellar mass) and_structural information_traces the gravitational_potential. Galaxy formation scenarios make specific_predictions on the_light distribution of galaxies, so accurate_reduction_of a galaxy_image_into_a total_magnitude and scalelength_are_important parameters_to_understanding the fundamental plane and the_star_formation history of galaxies.
Obtaining the structural characteristics_and the total luminosity_of_a galaxy requires knowledge_of its surface brightness profile_to a significant depth. In order_to extrapolate_a total_luminosity, isophotal analysis is required to determine how far one needs_to integrate a galaxy’s light plus_to provide sufficient information_to extrapolate_the_light profile.
During a_surface_photometry project_to explore the structure of galaxies by_morphological type_(Schombert & Smith 2011), we discovered_a significant discrepancy between_the_structural and luminosity parameters that we_determined using raw 2MASS images versus_those reported by the 2MASS_project_in_their Extended Source Catalog (Jarrett_2000). We use this letter to_outline the problem,_and the solution, for other researchers.
Sample
======
Our original_project_was to perform surface photometry of_large_galaxies over a full range of_morphological_types._This type of analysis was_previously attempted by 2MASS (Jarrett 2003),_but their focus was not on the structural characteristics_of galaxies, but_rather luminosities and colors. The_first_stage_of our project was to understand the surface photometry of_ellipticals, the_simplest galaxies for_structural studies as they have highly symmetric isophotes and very_few complications to their light distributions due_to extinction or recent star formation.
|
minus_DeltaA1smooth}
\|(d_A^*d_{A+a} - d_{A_1}^*d_{A_1})\xi\|_{L^p(X)}
\\
\leq
z\left(\|a\|_{W_{A_1}^{1,q}(X)} + \|a_1\|_{W_{A_1}^{1,q}(X)} + \|a_1\|_{W_{A_1}^{1,q}(X)}^2 + \|a_1\|_{W_{A_1}^{1,q}(X)} \|a\|_{W_{A_1}^{1,q}(X)} \right)\|\xi\|_{L^r(X)}
\\
+ z\left(\|a_1\|_{W_{A_1}^{1,q}(X)} + \|a\|_{W_{A_1}^{1,q}(X)}\right)\|\xi\|_{W_{A_1}^{1,s}(X)},\end{gathered}$$ with the values of $s$ specified in the proof of Corollary \[cor:Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\]. The remainder of the proof of Corollary \[cor:Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\] now applies to show that the operator is Fredholm with index zero.
Convergence of gradient flows under the validity of the [Ł]{}ojasiewicz–Simon gradient inequality {#sec:Convergence_gradient_flows_validity_Lojasiewicz-Simon_gradient_inequality}
=================================================================================================
While Theorem \[mainthm:Lojasiewicz-Simon\_gradient\_inequality\] has important applications to proofs of global existence, convergence, convergence rates, and stability of gradient flows defined by an energy function, $\sE:\sX\supset \sU \to \RR$, with gradient map, $\sM:\sX\supset \sU \to \tilde\sX$, (see [@Feehan_yang_mills_gradient_flow_v4 Section 2.1] for an introduction and Simon [@Simon_1983] for his pioneering development), the gradient inequality is most useful when it has the form, $$\|\sM(x)\|_{\sH} \geq Z|\sE(x) - \sE(x | minus_DeltaA1smooth }
\|(d_A^*d_{A+a } - d_{A_1}^*d_{A_1})\xi\|_{L^p(X) }
\\
\leq
z\left(\|a\|_{W_{A_1}^{1,q}(X) } + \|a_1\|_{W_{A_1}^{1,q}(X) } + \|a_1\|_{W_{A_1}^{1,q}(X)}^2 + \|a_1\|_{W_{A_1}^{1,q}(X) } \|a\|_{W_{A_1}^{1,q}(X) } \right)\|\xi\|_{L^r(X) }
\\
+ z\left(\|a_1\|_{W_{A_1}^{1,q}(X) } + \|a\|_{W_{A_1}^{1,q}(X)}\right)\|\xi\|_{W_{A_1}^{1,s}(X)},\end{gathered}$$ with the values of $ s$ specified in the proof of Corollary \[cor: Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\ ]. The remainder of the validation of Corollary \[cor: Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\ ] nowadays applies to show that the hustler is Fredholm with index zero.
Convergence of gradient flow under the validity of the [ Ł]{}ojasiewicz – Simon gradient inequality { # sec: Convergence_gradient_flows_validity_Lojasiewicz - Simon_gradient_inequality }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
While Theorem \[mainthm: Lojasiewicz - Simon\_gradient\_inequality\ ] have important applications to proofs of global existence, overlap, convergence rates, and stability of gradient flows define by an energy affair, $ \sE:\sX\supset \sU \to \RR$, with gradient map, $ \sM:\sX\supset \sU \to \tilde\sX$, (see [ @Feehan_yang_mills_gradient_flow_v4 part 2.1 ] for an introduction and Simon [ @Simon_1983 ] for his pioneering development), the gradient inequality is about useful when it has the phase, $ $ \|\sM(x)\|_{\sH } \geq Z|\sE(x) - \sE(x | minks_DeltaA1smooth}
\|(d_A^*d_{A+a} - d_{A_1}^*d_{X_1})\xi\|_{L^p(X)}
\\
\leq
z\left(\|a\|_{C_{Q_1}^{1,q}(X)} + \|a_1\|_{X_{A_1}^{1,q}(X)} + \|a_1\|_{S_{A_1}^{1,q}(X)}^2 + \|a_1\|_{W_{X_1}^{1,q}(X)} \|a\|_{W_{A_1}^{1,q}(X)} \right)\|\xi\|_{L^r(X)}
\\
+ z\left(\|a_1\|_{W_{E_1}^{1,q}(X)} + \|a\|_{W_{A_1}^{1,q}(Z)}\right)\|\xi\|_{W_{A_1}^{1,s}(X)},\end{gatherea}$$ with thv values if $s$ wpecified mh the pvjof kn Cormolary \[cor:Fredhplmness\_and\_hndex\_Laplace\_oparxtlr\_on\_W2p\_Sobolev\_connection\]. The remaindqr of tne proof of Corojlarj \[sor:Fdvdmolmness\_and\_index\_Laplace\_operator\_oh\_W2p\_Sobonev\_connection\] now applies to show that hhe lperator is Fredhopm with indgs zqeo.
Convergencd of graditnc flows undgr the validity of the [Ł]{}ojasiewicx–Simou gradient unwquwnity {#sec:Contergenbe_gradient_flows_valididy_Lojasoewicz-Simon_gracieit_inwquality}
=================================================================================================
While Theorem \[mainthm:Lojasiewice-Simon\_gradheut\_inequality\] has impoetqnt akplicdtiovw tu pdopfa of gpobel existencs, convergenxe, convergence ratex, wbd stability kf graqiqnt flows defined by an energy function, $\sE:\aX\supset \sU \to \RR$, with tradient map, $\sM:\sX\supsgt \sU \to \tylde\sX$, (see [@Feehan_yang_mills_gradient_flow_v4 Section 2.1] xor ai kntxiductkin and Simon [@Simon_1983] for his pioneering developmegf), uhe gradient ineqmality is most usegup eren it has thg form, $$\|\sM(x)\|_{\aH} \geq Z|\sE(x) - \sE(x | minus_DeltaA1smooth} \|(d_A^*d_{A+a} - d_{A_1}^*d_{A_1})\xi\|_{L^p(X)} \\ \leq z\left(\|a\|_{W_{A_1}^{1,q}(X)} + + \|a_1\|_{W_{A_1}^{1,q}(X)} \right)\|\xi\|_{L^r(X)} \\ + values $s$ specified in proof of Corollary The remainder of the proof of \[cor:Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\] now applies to show that the operator is Fredholm with index zero. of gradient flows under the validity of the [Ł]{}ojasiewicz–Simon gradient inequality {#sec:Convergence_gradient_flows_validity_Lojasiewicz-Simon_gradient_inequality} ================================================================================================= Theorem has applications proofs of global existence, convergence, convergence rates, and stability of gradient flows defined by an energy $\sE:\sX\supset \sU \to \RR$, with gradient map, $\sM:\sX\supset \to \tilde\sX$, (see [@Feehan_yang_mills_gradient_flow_v4 2.1] for an introduction and [@Simon_1983] his pioneering the inequality most useful when has the form, $$\|\sM(x)\|_{\sH} \geq Z|\sE(x) - \sE(x | minus_DeltaA1smooth}
\|(d_A^*d_{A+a} - d_{A_1}^*d_{a_1})\xi\|_{L^p(X)}
\\
\leq
z\Left(\|a\|_{w_{A_1}^{1,q}(x)} + \|a_1\|_{W_{a_1}^{1,q}(x)} + \|a_1\|_{W_{A_1}^{1,Q}(X)}^2 + \|a_1\|_{W_{a_1}^{1,q}(X)} \|a\|_{W_{A_1}^{1,q}(X)} \right)\|\xI\|_{l^r(X)}
\\
+ z\Left(\|a_1\|_{W_{A_1}^{1,q}(X)} + \|a\|_{W_{A_1}^{1,q}(X)}\right)\|\xi\|_{W_{a_1}^{1,s}(X)},\enD{gATherED}$$ wIth thE values OF $s$ SPEciFiEd In tHe PRoOf of COroLlary \[coR:FredholmnEss\_AnD\_index\_LaplacE\_OpErator\_on\_W2p\_sobOlev\_connectiOn\]. THe remaInDer OF the pRooF of CoRollarY \[Cor:FreDholmness\_AnD\_Index\_LAPlace\_opERAtOr\_on\_w2p\_Sobolev\_connectiON\] nOW applies to show That thE oPErATOr iS FrEdholm with InDex zeRO.
ConverGEnCE OF grADient flows undEr the validiTY of The [Ł]{}ojAsIewICz–SimoN gradIeNT inEquality {#sec:convErgence_grAdient_FLows_valIDity_LojAsiewiCz-SImoN_graDIeNt_IneQuALitY}
=================================================================================================
whIle tHeoRem \[maintHm:loJasieWicz-sIMON\_graDieNt\_inEqualIty\] has importaNt aPpliCAtiOns to ProofS of gLoBal exIstencE, convErGence, convergencE ratEs, and stabIliTy Of gRaDient FLows deFinEd bY an enerGy functIOn, $\se:\sx\SUPsEt \sU \to \RR$, with gradieNt MAP, $\sm:\sX\supseT \sU \to \tILdE\sx$, (See [@FeehaN_yAng_MillS_GRadieNt_flOW_v4 section 2.1] fOr an inTRoDuCtion anD SImon [@SiMoN_1983] foR hiS pionEErinG develOpment), thE gradIEnt inequality iS Most useful wheN It HAS tHE forM, $$\|\sM(X)\|_{\sH} \geq Z|\sE(x) - \se(x | minus_DeltaA1smooth}
\|(d_ A^*d_{A+a} - d_ {A_ 1}^ *d _{A_ 1})\ xi\|_{L^p(X)}\ \
\l eq
z\left(\|a\|_{W_{A_ 1}^{1 ,q } (X)} +\|a_1 \|_{W_{ A _1 } ^ {1, q} (X )}+\ |a _1\|_ {W_ {A_1}^{ 1,q}(X)}^2 +\| a_1\|_{W_{A_ 1 }^ {1,q}(X)}\|a \|_{W_{A_1}^ {1, q}(X)} \ rig h t)\|\ xi\ |_{L^ r(X)}\ \
+ z\ left(\|a_ 1\ | _{W_{A _ 1}^{1,q } ( X) } +\|a\|_{W_{A_1}^{1 , q} ( X)}\right)\|\x i\|_{W _{ A _1 } ^ {1, s}( X)},\end{g at hered } $$ with th e v alu e s of $s$ spec ified in th e pr oof of C oro l lary \ [cor: Fr e dho lmness\_and \_in dex\_Lapl ace\_o p erator\ _ on\_W2p \_Sobo lev \_c onne c ti on \]. T h e r e ma ind e r o f the pr oo fof Co roll a r y \[co r:F redh olmne ss\_and\_inde x\_ Lapl a ce\ _oper ator\ _on\ _W 2p\_S obolev \_con ne ction\] now app lies to showtha tthe o perat o r is F red hol m withindex z e ro.
C o n ve rgence of gradient f l o ws under t he val i di ty of the [ Ł] {}o jasi e w icz–S imon gr adient i nequal i ty { #sec:Co nv ergenc e_ gra die nt_fl o ws_v alidit y_Lojasi ewicz - Simon_gradient _ inequality}
= = == = = == = ==== === =========== ==== = ==== ==== = == === = ===== ===== == = == = =================== == ====== ===== ==
While The orem \[mai n t h m:Lojasi ewic z -S i mon\_gradient\ _ineq uality\] h a s import ant a pplicati ons to pr o o fs of gl oba l e xis ten c e ,convergence,c o nver ge nce rat es, and st abi lit y o f g ra dient flo ws defin ed b yan en ergyf unction, $ \sE :\ sX\ supse t \sU \ to \R R$,wi th gra dient m a p, $ \sM: \s X\ sups et\s U \to \ti l de\ sX$, (s ee [@Feeh an_ y ang_ mi ll s_gradi ent_flow_v4 S ec tion 2.1]fo r a n intr o d uction a nd Simon [@Simon_1983]f or hispio neeri ng d evelopmen t), the g rad i ent in equali ty is m ost u seful w he n i thas the fo r m , $ $\|\s M( x)\| _{\sH}\geq Z|\sE(x) - \s E (x | minus_DeltaA1smooth}
\|(d_A^*d_{A+a} -_d_{A_1}^*d_{A_1})\xi\|_{L^p(X)}
\\
\leq
z\left(\|a\|_{W_{A_1}^{1,q}(X)} +_\|a_1\|_{W_{A_1}^{1,q}(X)} + \|a_1\|_{W_{A_1}^{1,q}(X)}^2 +_\|a_1\|_{W_{A_1}^{1,q}(X)} \|a\|_{W_{A_1}^{1,q}(X)}_\right)\|\xi\|_{L^r(X)}
\\
+_z\left(\|a_1\|_{W_{A_1}^{1,q}(X)} +_\|a\|_{W_{A_1}^{1,q}(X)}\right)\|\xi\|_{W_{A_1}^{1,s}(X)},\end{gathered}$$_with the values_of $s$ specified_in the proof of_Corollary \[cor:Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\]. The_remainder_of the proof of Corollary \[cor:Fredholmness\_and\_index\_Laplace\_operator\_on\_W2p\_Sobolev\_connection\] now applies to show that the operator is_Fredholm_with index_zero.
Convergence_of_gradient flows under the validity_of the [Ł]{}ojasiewicz–Simon gradient inequality_{#sec:Convergence_gradient_flows_validity_Lojasiewicz-Simon_gradient_inequality}
=================================================================================================
While Theorem_\[mainthm:Lojasiewicz-Simon\_gradient\_inequality\] has important applications to proofs of global_existence,_convergence, convergence rates,_and stability of gradient flows defined by an energy_function, $\sE:\sX\supset \sU \to \RR$, with_gradient map, $\sM:\sX\supset_\sU_\to_\tilde\sX$, (see [@Feehan_yang_mills_gradient_flow_v4 Section_2.1] for an introduction and Simon_[@Simon_1983] for his pioneering development), the_gradient inequality is most useful when it_has the form, $$\|\sM(x)\|_{\sH} \geq Z|\sE(x)_- \sE(x |
times 10^{-10}$ & 0 & 0.00232\
C/1985 R1 (Hartley-Good) & 5982.4 & 0.999884 & - & $1.33\times 10^{-8}$ & $-2.16\times 10^{-9}$ & 0 & 0.00217\
316P/LONEOS-Christensen & 4.328 & 0.166 & - & $7.34\times 10^{-5}$ & 0 & 0 & 0.00212\
73P/Schwassmann-Wachmann 3-B & 3.062 & 0.693 & - & $1.49\times 10^{-8}$ & $1.96\times 10^{-9}$ & 0 & 0.00212\
C/1993 Y1 (McNaught-Russell) & 134.76 & 0.99356 & - & $1.65\times 10^{-8}$ & $1.22\times 10^{-9}$ & 0 & 0.00207\
C/1999 T1 (McNaught-Hartley) & 8149.7 & 0.999856 & - & $2.44\times 10^{-8}$ & $-9.59\times 10^{-10}$ & 0 & 0.00198\
101P/Chernykh & 5.785 & 0.594 & 5.6 & $-5.16\times 10^{-8}$ & $1.55\times 10^{-7}$ & 0 & 0.00197\
168P/Hergenrother & 3.624 & 0.61 & - & $2.66\times 10^{-8}$ & $2.37\times 10^{-9}$ & $-1.25\times 10^{-9}$ & 0.00196\
76P/West-Kohoutek-Ikemura & 3.471 & 0.539 & 0.66 & $-2.18\times 10^{-9}$ & $-1.85\times 10^{-10}$ & 0 & 0.00186\
252P/LINEAR & 3.047 & 0.673 & - & $1.05\times 10^{-8}$ & $-7.06\times 10^{-9}$ & 0 & 0.00167\
88P/Howell & 3.11 & 0.562 & 4.4 & $8.41\times 10^{-9 | times 10^{-10}$ & 0 & 0.00232\
C/1985 R1 (Hartley - Good) & 5982.4 & 0.999884 & - & $ 1.33\times 10^{-8}$ & $ -2.16\times 10^{-9}$ & 0 & 0.00217\
316P / LONEOS - Christensen & 4.328 & 0.166 & - & $ 7.34\times 10^{-5}$ & 0 & 0 & 0.00212\
73P / Schwassmann - Wachmann 3 - B & 3.062 & 0.693 & - & $ 1.49\times 10^{-8}$ & $ 1.96\times 10^{-9}$ & 0 & 0.00212\
C/1993 Y1 (McNaught - Russell) & 134.76 & 0.99356 & - & $ 1.65\times 10^{-8}$ & $ 1.22\times 10^{-9}$ & 0 & 0.00207\
C/1999 T1 (McNaught - Hartley) & 8149.7 & 0.999856 & - & $ 2.44\times 10^{-8}$ & $ -9.59\times 10^{-10}$ & 0 & 0.00198\
101P / Chernykh & 5.785 & 0.594 & 5.6 & $ -5.16\times 10^{-8}$ & $ 1.55\times 10^{-7}$ & 0 & 0.00197\
168P / Hergenrother & 3.624 & 0.61 & - & $ 2.66\times 10^{-8}$ & $ 2.37\times 10^{-9}$ & $ -1.25\times 10^{-9}$ & 0.00196\
76P / West - Kohoutek - Ikemura & 3.471 & 0.539 & 0.66 & $ -2.18\times 10^{-9}$ & $ -1.85\times 10^{-10}$ & 0 & 0.00186\
252P / LINEAR & 3.047 & 0.673 & - & $ 1.05\times 10^{-8}$ & $ -7.06\times 10^{-9}$ & 0 & 0.00167\
88P / Howell & 3.11 & 0.562 & 4.4 & $ 8.41\times 10^{-9 | timfs 10^{-10}$ & 0 & 0.00232\
C/1985 R1 (Hartley-Good) & 5982.4 & 0.999884 & - & $1.33\times 10^{-8}$ & $-2.16\tnnes 10^{-9}$ & 0 & 0.00217\
316P/LONSOS-Chrisgensen & 4.328 & 0.166 & - & $7.34\times 10^{-5}$ & 0 & 0 & 0.00212\
73P/Srhwawsmanb-Wachmann 3-B & 3.062 & 0.693 & - & $1.49\timds 10^{-8}$ & $1.96\timed 10^{-9}$ & 0 & 0.00212\
C/1993 T1 (McIaught-Russell) & 134.76 & 0.99356 & - & $1.65\timcf 10^{-8}$ & $1.22\flmes 10^{-9}$ & 0 & 0.00207\
C/1999 T1 (McNaughj-Hartley) & 8149.7 & 0.999856 & - & $2.44\times 10^{-8}$ & $-9.59\tiker 10^{-10}$ & 0 & 0.00198\
101P/Chernykh & 5.785 & 0.594 & 5.6 & $-5.16\times 10^{-8}$ & $1.55\times 10^{-7}$ & 0 & 0.00197\
168P/Hetgfnrother & 3.624 & 0.61 & - & $2.66\tikqs 10^{-8}$ & $2.37\nines 10^{-9}$ & $-1.25\times 10^{-9}$ & 0.00196\
76P/West-Kohoutek-Iksmura & 3.471 & 0.539 & 0.66 & $-2.18\times 10^{-9}$ & $-1.85\times 10^{-10}$ & 0 & 0.00186\
252P/LINEAR & 3.047 & 0.673 & - & $1.05\himed 10^{-8}$ & $-7.06\times 10^{-9}$ & 0 & 0.00167\
88P/Howfll & 3.11 & 0.562 & 4.4 & $8.41\jjmef 10^{-9 | times 10^{-10}$ & 0 & 0.00232\ C/1985 & & 0.999884 - & $1.33\times 0 0.00217\ 316P/LONEOS-Christensen & & 0.166 & & $7.34\times 10^{-5}$ & 0 & & 0.00212\ 73P/Schwassmann-Wachmann 3-B & 3.062 & 0.693 & - & $1.49\times 10^{-8}$ $1.96\times 10^{-9}$ & 0 & 0.00212\ C/1993 Y1 (McNaught-Russell) & 134.76 & 0.99356 - $1.65\times & 10^{-9}$ & 0 & 0.00207\ C/1999 T1 (McNaught-Hartley) & 8149.7 & 0.999856 & - & $2.44\times & $-9.59\times 10^{-10}$ & 0 & 0.00198\ 101P/Chernykh 5.785 & 0.594 & & $-5.16\times 10^{-8}$ & $1.55\times & & 0.00197\ & & & - & 10^{-8}$ & $2.37\times 10^{-9}$ & $-1.25\times 10^{-9}$ & 0.00196\ 76P/West-Kohoutek-Ikemura & 3.471 & 0.539 & 0.66 & 10^{-9}$ & & 0 0.00186\ & & 0.673 & $1.05\times 10^{-8}$ & $-7.06\times 10^{-9}$ & 88P/Howell & 3.11 & 0.562 & 4.4 & 10^{-9 | times 10^{-10}$ & 0 & 0.00232\
C/1985 R1 (Hartley-Good) & 5982.4 & 0.999884 & - & $1.33\times 10^{-8}$ & $-2.16\tiMes 10^{-9}$ & 0 & 0.00217\
316P/LONEOS-chrisTenSen & 4.328 & 0.166 & - & $7.34\TiMes 10^{-5}$ & 0 & 0 & 0.00212\
73P/schwAssmann-WachmanN 3-b & 3.062 & 0.693 & - & $1.49\timEs 10^{-8}$ & $1.96\times 10^{-9}$ & 0 & 0.00212\
C/1993 Y1 (McNaught-RusseLl) & 134.76 & 0.99356 & - & $1.65\timEs 10^{-8}$ & $1.22\TImes 10^{-9}$ & 0 & 0.00207\
c/1999 t1 (MCNaugHt-HartlEY) & 8149.7 & 0.999856 & - & $2.44\tIMEs 10^{-8}$ & $-9.59\tImEs 10^{-10}$ & 0 & 0.00198\
101p/ChErNYkH & 5.785 & 0.594 & 5.6 & $-5.16\timeS 10^{-8}$ & $1.55\tiMes 10^{-7}$ & 0 & 0.00197\
168P/HerGenrother & 3.624 & 0.61 & - & $2.66\tImeS 10^{-8}$ & $2.37\tImes 10^{-9}$ & $-1.25\times 10^{-9}$ & 0.00196\
76P/WeST-KOhoutek-IkeMurA & 3.471 & 0.539 & 0.66 & $-2.18\times 10^{-9}$ & $-1.85\times 10^{-10}$ & 0 & 0.00186\
252P/lINeAR & 3.047 & 0.673 & - & $1.05\timEs 10^{-8}$ & $-7.06\TimES 10^{-9}$ & 0 & 0.00167\
88P/HowEll & 3.11 & 0.562 & 4.4 & $8.41\Times 10^{-9 | times 10^{-10}$ & 0 & 0.00 232\
C/198 5 R1(Ha rtl ey -Goo d) & 5982.4 & 0.99 9 884& - & $1.33\times 10^{ -8}$&$ -2.1 6 \t imes10^{-9} $ & 0 &0. 00 217 \3 16 P/LON EOS -Christ ensen & 4. 328 & 0.166 & - & $7 .34\times10^ {-5}$ & 0 &0 & 0.002 12 \
7 3 P/Sch was smann -Wachm a nn 3-B & 3.062&0 .693 & - & $1. 4 9 \t imes 10^{-8}$ & $1.96 \ ti m es 10^{-9}$ &0 & 0. 00 2 12 \ C/1 993 Y1 (McNau gh t-Rus s ell) &1 34 . 7 6 &0 .99356 & - &$1.65\times 10^ {-8}$&$1. 2 2\time s 10^ {- 9 }$& 0 & 0.002 07\C/1999 T1 (McNa u ght-Har t ley) &8149.7 &0.9 9985 6 & - &$2 . 44\ t im es1 0^{ -8}$ & $ -9 .5 9\tim es 1 0 ^ { - 10}$ &0 &0.001 98\
101P/Cher nyk h &5 .78 5 & 0 .594& 5. 6& $-5 .16\ti mes 1 0^ {-8}$ & $1.55\t imes 10^{-7}$ &0& 0 .0 0197\ 168P/H erg enr other & 3.624& 0. 61 & -& $2.66\times 10^{ -8 } $ & $2.37\t imes 1 0 ^{ -9 } $ & $-1. 25 \ti mes1 0 ^{-9} $ &0 .0 0196\
76 P/West - Ko ho utek-Ik em ura &3. 471 &0.539 & 0. 66 & $ -2.18\ti mes 1 0 ^{-9}$ & $-1.8 5 \times 10^{-1 0 }$ & 0 & 0. 001 86\
252P/LI NEAR & 3. 047& 0 .67 3 & -& $1. 05 \ ti m es 10^{-8}$ & $-7.0 6\ times10^{- 9}$ & 0 & 0.0 0167\
88P/ H o w ell & 3. 11 & 0. 5 62 & 4.4 & $8. 41\ti mes 10^{-9 | times 10^{-10}$_& 0_& 0.00232\
C/1985 R1 (Hartley-Good)_& 5982.4_&_0.999884 &_-_& $1.33\times 10^{-8}$_& $-2.16\times 10^{-9}$_& 0 & 0.00217\
316P/LONEOS-Christensen_& 4.328 &_0.166_& - & $7.34\times 10^{-5}$ & 0 & 0 & 0.00212\
73P/Schwassmann-Wachmann 3-B & 3.062_&_0.693 &_-_&_$1.49\times 10^{-8}$ & $1.96\times 10^{-9}$_& 0 & 0.00212\
C/1993 Y1_(McNaught-Russell) &_134.76 & 0.99356 & - & $1.65\times 10^{-8}$_&_$1.22\times 10^{-9}$ &_0 & 0.00207\
C/1999 T1 (McNaught-Hartley) & 8149.7 & 0.999856_& - & $2.44\times 10^{-8}$ &_$-9.59\times 10^{-10}$ &_0_&_0.00198\
101P/Chernykh & 5.785 &_0.594 & 5.6 & $-5.16\times 10^{-8}$_& $1.55\times 10^{-7}$ & 0 &_0.00197\
168P/Hergenrother & 3.624 & 0.61 & -_& $2.66\times 10^{-8}$ & $2.37\times 10^{-9}$_& $-1.25\times 10^{-9}$ & 0.00196\
76P/West-Kohoutek-Ikemura_& 3.471_& 0.539 & 0.66 &_$-2.18\times 10^{-9}$ &_$-1.85\times 10^{-10}$_& 0 &_0.00186\
252P/LINEAR & 3.047 & 0.673 &_- & $1.05\times_10^{-8}$ & $-7.06\times 10^{-9}$ & 0_&_0.00167\
88P/Howell & 3.11_&_0.562_& 4.4_& $8.41\times 10^{-9 |
{n+1}{2}$. In this paper, we prove that if $d^2 + d + 2 \leq n$ and if $d\geq 3$, then the spaces of rational curves are themselves rationally connected.'
address:
- |
Department of Mathematics\
Harvard University\
Cambridge MA 02138
- |
Department of Mathematics\
Massachusetts Institute of Technology\
Cambridge MA 02139
author:
- Joe Harris
- Jason Starr
bibliography:
-'my.bib'
title: 'Rational curves on hypersurfaces of low degree, II'
---
Statement of results
====================
\[sec-results\]
In [@HRS2], it is proved that if $X_d \subset {\mathbb{P}}^n$ is a general hypersurface of degree $d < \frac{n+1}{2}$, then each space $\text{RatCurves}^e(X)$ parametrizing smooth rational curves of degree $e$ on $X$, is itself an integral, local complete intersection scheme of the expected dimension $(n+1-d)e+(n-4)$. More precisely, it is proved that for every stable $A$-graph $\tau$ and every flag $f\in
\text{Flag}(\tau)$, the Behrend-Manin stack ${\overline{{\mathcal}M}}(X,\tau)$ is an integral, local complete intersection stack of the expected dimension $\text{dim}(X,\tau)$, and the evaluation morphism $\text{ev}_f:{\overline{{\mathcal}M}}(X,\tau) \rightarrow X$ is flat of the expected fiber dimension $\text{dim}(X,\tau) - \text{dim}(X)$. Since $\text{RatCurves}^e(X)$ is a Zariski open set in the stack ${\overline{{\mathcal}M}_{0,0}({X,e})}$, the result on $\text{RatCurves}^e(X)$ follows.
After establishing irreducibility and the dimension of the spaces $\text{RatCurves}^e(X)$, the next question is to determine the Kodaira dimension of $\text{RatCurves}^e(X)$. For a general Fano hypersurface $X_d \subset {\mathbb{P}}^n$ with $d\leq n$, determining the Kodaira dimensions of $\text{RatCurves | { n+1}{2}$. In this paper, we prove that if $ d^2 + d + 2 \leq n$ and if $ d\geq 3 $, then the spaces of intellectual curve are themselves rationally connected.'
address:
- |
Department of Mathematics\
Harvard University\
Cambridge MA 02138
- |
Department of Mathematics\
Massachusetts Institute of Technology\
Cambridge MA 02139
writer:
- Joe Harris
- Jason Starr
bibliography:
-'my.bib'
deed:' Rational curves on hypersurfaces of broken academic degree, II'
---
Statement of results
= = = = = = = = = = = = = = = = = = = =
\[sec - results\ ]
In [ @HRS2 ], it is proved that if $ X_d \subset { \mathbb{P}}^n$ is a general hypersurface of degree $ d < \frac{n+1}{2}$, then each space $ \text{RatCurves}^e(X)$ parametrizing placid rational curves of academic degree $ e$ on $ X$, is itself an integral, local complete intersection scheme of the expected dimension $ (n+1 - d)e+(n-4)$. More precisely, it is proved that for every static $ A$ -graph $ \tau$ and every flag $ f\in
\text{Flag}(\tau)$, the Behrend - Manin stack $ { \overline{{\mathcal}M}}(X,\tau)$ is an integral, local complete overlap stack of the expected dimension $ \text{dim}(X,\tau)$, and the evaluation morphism $ \text{ev}_f:{\overline{{\mathcal}M}}(X,\tau) \rightarrow X$ is flat of the expected roughage dimension $ \text{dim}(X,\tau) - \text{dim}(X)$. Since $ \text{RatCurves}^e(X)$ is a Zariski open set in the stack $ { \overline{{\mathcal}M}_{0,0}({X, e})}$, the result on $ \text{RatCurves}^e(X)$ follow.
After establishing irreducibility and the dimension of the spaces $ \text{RatCurves}^e(X)$, the next question is to settle the Kodaira dimension of $ \text{RatCurves}^e(X)$. For a general Fano hypersurface $ X_d \subset { \mathbb{P}}^n$ with $ d\leq n$, determining the Kodaira dimensions of $ \text{RatCurves | {n+1}{2}$. Ij this paper, we prove thxt if $d^2 + d + 2 \leq n$ and if $d\gsq 3$, then the spaces of rational curvxs aee thtiselves rationally connectef.'
address:
- |
Eepartment of Mathciatidd\
Iarvard Universlty\
Cambsidge MA 02138
- |
Geoaxtment of Mathematics\
Massachusetes Instotkte of Technoljgy\
Cajbridge MA 02139
author:
- Joe Harris
- Jason Starr
bpbliography:
-'my.bib'
totle: 'Rational curves on hyoerskrfaces of low deggee, II'
---
Statenent if results
====================
\[sez-results\]
In [@HRS2], it is ptoved that if $X_d \subset {\mathbb{P}}^n$ is a general hi'wrsktface of degcee $d < \frac{n+1}{2}$, then csch spdce $\texy{RatCurves}^e(X)$ psraketeizing smooth rationan curves of degree $e$ on $X$, iv ntself an integral, loxao comklete intdeseztikn sdheme lf vhe expectes dimension $(n+1-d)e+(n-4)$. More preciselu, yn is proved tgat fow qvery stable $A$-graph $\tau$ and every flag $x\in
\fext{Flag}(\tau)$, the Behrend-Nanin stack ${\overline{{\mwthcal}M}}(X,\twu)$ is an integral, local complete intersection stawk of ghe ewpecgwd dimension $\text{dim}(X,\tau)$, and the evaluation mor[gixm $\text{ev}_f:{\overlike{{\mathcal}M}}(X,\tau) \rigntwrtjw X$ is flat uf the exlected fiber dimendion $\tevt{dim}(Z,\tau) - \texu{dim}(X)$. Since $\text{RatCurves}^e(X)$ is a Zariski opeu swt in the stack ${\ovzrline{{\mathcau}M}_{0,0}({X,e})}$, the tesult on $\text{RatCurves}^z(X)$ folmows.
After edtablishihe irreducibility anc dhe dimension of the spacef $\text{RatRurvex}^e(X)$, the nexj questyon is to fetermine the Kodaira dilensiln of $\text{RahCurves}^e(X)$. For a general Fano hy'xrsurface $X_d \xutsen {\mathbb{P}}^u$ with $d\leq n$, deteriining the Kodcira dimznsionr of $\text{RztCurvev | {n+1}{2}$. In this paper, we prove that + + 2 n$ and if of curves are themselves connected.' address: - Department of Mathematics\ Harvard University\ Cambridge 02138 - | Department of Mathematics\ Massachusetts Institute of Technology\ Cambridge MA 02139 - Joe Harris - Jason Starr bibliography: -'my.bib' title: 'Rational curves on hypersurfaces low II' Statement results ==================== \[sec-results\] In [@HRS2], it is proved that if $X_d \subset {\mathbb{P}}^n$ is a general of degree $d < \frac{n+1}{2}$, then each space parametrizing smooth rational curves degree $e$ on $X$, is an local complete scheme the dimension $(n+1-d)e+(n-4)$. More it is proved that for every stable $A$-graph $\tau$ and every flag $f\in \text{Flag}(\tau)$, the Behrend-Manin stack is an complete intersection of expected $\text{dim}(X,\tau)$, and the $\text{ev}_f:{\overline{{\mathcal}M}}(X,\tau) \rightarrow X$ is flat of dimension $\text{dim}(X,\tau) - \text{dim}(X)$. Since $\text{RatCurves}^e(X)$ is a open set the stack ${\overline{{\mathcal}M}_{0,0}({X,e})}$, the result on follows. After establishing irreducibility and the dimension of spaces $\text{RatCurves}^e(X)$, the next question is to determine the Kodaira dimension of $\text{RatCurves}^e(X)$. For a hypersurface $X_d \subset {\mathbb{P}}^n$ $d\leq n$, determining Kodaira of | {n+1}{2}$. In this paper, we prove that if $D^2 + d + 2 \leq n$ and iF $d\geq 3$, TheN thE sPaceS of rAtional curves aRE theMselves rationally conneCted.'
aDdREss:
- |
DEPaRtmenT of MathEMaTICs\
HArVaRd UNiVErSity\
CAmbRidge MA 02138
- |
department Of MAtHematics\
MassAChUsetts InstItuTe of TechnoloGy\
CAmbridGe mA 02139
aUThor:
- JOe HArris
- jason STArr
bibLiography:
-'My.BIb'
titlE: 'rationaL CUrVes oN hypersurfaces of lOW dEGree, II'
---
StatemenT of resUlTS
====================
\[sEC-ResUltS\]
In [@HRS2], it is PrOved tHAt if $X_d \sUBsET {\MAthBB{P}}^n$ is a general HypersurfacE Of dEgree $d < \FrAc{n+1}{2}$, THen eacH spacE $\tEXt{RAtCurves}^e(X)$ pAramEtrizing sMooth rATional cURves of dEgree $e$ On $X$, Is iTselF An InTegRaL, LocAL cOmpLEte IntersecTiOn SchemE of tHE EXPectEd dImenSion $(n+1-D)e+(n-4)$. More precisEly, It is PRovEd thaT for eVery StAble $A$-Graph $\tAu$ and EvEry flag $f\in
\text{FLag}(\tAu)$, the BehrEnd-maNin StAck ${\ovERline{{\mAthCal}m}}(X,\tau)$ is An integRAl, lOcAL COmPlete intersection sTaCK Of The expecTed dimENsIoN $\Text{dim}(X,\TaU)$, anD the EVAluatIon mORpHism $\text{Ev}_f:{\oveRLiNe{{\Mathcal}m}}(X,\Tau) \rigHtArrOw X$ Is flaT Of thE expecTed fiber DimenSIon $\text{dim}(X,\tau) - \TExt{dim}(X)$. Since $\tEXt{rATCURves}^E(X)$ iS a Zariski opEn seT In thE staCK ${\oVerLIne{{\maThcal}m}_{0,0}({X,E})}$, ThE Result on $\text{RatCurvEs}^E(X)$ follOws.
AfTer establishiNg irreduciBILIty and thE dimENsIOn of the spaces $\tExt{RaTCurves}^e(X)$, tHE next queStion Is to deteRmine the KODAira dimeNsiOn oF $\teXt{RATcuRves}^e(X)$. For a genERAl FaNo HypersuRfaCe $X_d \subSet {\MatHbb{p}}^n$ wItH $d\leq n$, detErmining ThE KOdAiRa dImensIOns of $\texT{RAtCUrVes | {n+1}{2}$. In this paper,we prove t hat i f $ d^2 + d + 2 \ leq n$ and if$ d\ge q 3$, then the spacesof ra ti o nalc ur ves a re them s el v e s r at io nal ly co nnect ed. '
addre ss:
- |
D ep artment of M a th ematics\
Harvard Univ ers ity\
Ca m bridg e M A 021 38
- | De partmentof Mathem a tics\
M assa chusetts Institut e o f Technology\
Cam br i dg e MA021 39
author:
- JoeH arris
- Ja s o n St a rr
bibliograp hy:
-'my.bi b '
t itle:'R ati o nal cu rveson hyp ersurfacesof l ow degree , II'- --
Sta t ement o f resu lts
== ==== = == == === == = ===
\[s e c-r esults\]
In [@H RS2] , i t ispro vedthatif $X_d \subs et{\ma t hbb {P}}^ n$ is a g en eralhypers urfac eof degree $d <\fra c{n+1}{2} $,th enea ch sp a ce $\t ext {Ra tCurves }^e(X)$ par am e t r iz ing smooth rationa lc u rv es of de gree $ e $on $X$, isit sel f an i ntegr al,l oc al compl ete in t er se ction s ch eme of t heexp ected dime nsion$(n+1-d) e+(n- 4 )$. More preci s ely, it is pr o ve d th a t fo r e very stable $A$ - grap h $\ t au $ a n d eve ry fl ag $f \ in
\text{Flag}(\tau )$ , theBehre nd-Manin stac k ${\overl i n e {{\mathc al}M } }( X ,\tau)$ is aninteg ral, local complete inte rsection stack of t he expec ted di men sio n $\ text{dim}(X,\ t a u)$, a nd theeva luation mo rph ism $\ te xt{ev}_f: {\overli ne {{ \m at hca l}M}} ( X,\tau)\r igh ta rro w X$i s flat of t he e xp ec t edfiber d i me n s ion$\ te xt{d im} (X ,\tau ) -\ tex t{dim}( X)$. Sinc e $ \ text {R at Curves} ^e(X)$ is a Z ar iski opense t i n thes t ack ${\o verline{{\mathcal}M}_{0 , 0}({X,e })} $, th e re sult on $ \te xt{Rat Cur v es}^e( X)$ fo llows .
Af t e r est a b li shi ng irreducib i l ity andth e di mension of the spaces $\t e xt{ RatCurves}^e( X)$ , th e ne xtq ue s tio ni s t o determine the K odaira dim en s io n of $\tex t {Ra tC urves}^ e(X)$.For a general Fano hyp ersurface $ X_d\ s ubs et {\mathb b{P}}^n$ with $d\ l eq n$ , d eterm ini ng the K oda ira d imensi o nsof $\ text{R at Curves | {n+1}{2}$. In_this paper,_we prove that if_$d^2 +_d_+ 2_\leq_n$ and if_$d\geq 3$, then_the spaces of rational_curves are themselves_rationally_connected.'
address:
- |
Department of Mathematics\
Harvard University\
__ Cambridge_MA_02138
-_|
Department_of Mathematics\
_Massachusetts Institute_of Technology\
Cambridge MA 02139
author:
-_Joe_Harris
- Jason Starr
bibliography:
-'my.bib'
title:_'Rational curves on hypersurfaces of low degree, II'
---
Statement of_results
====================
\[sec-results\]
In [@HRS2], it is proved that_if $X_d \subset_{\mathbb{P}}^n$_is_a general hypersurface of_degree $d < \frac{n+1}{2}$, then each_space $\text{RatCurves}^e(X)$ parametrizing smooth rational curves_of degree $e$ on $X$, is itself_an integral, local complete intersection scheme_of the expected dimension $(n+1-d)e+(n-4)$._More precisely,_it is proved that for_every stable $A$-graph_$\tau$ and_every flag $f\in
\text{Flag}(\tau)$,_the Behrend-Manin stack ${\overline{{\mathcal}M}}(X,\tau)$ is an_integral, local complete_intersection stack of the expected dimension_$\text{dim}(X,\tau)$,_and the evaluation_morphism_$\text{ev}_f:{\overline{{\mathcal}M}}(X,\tau)_\rightarrow X$_is flat of_the_expected fiber_dimension_$\text{dim}(X,\tau) - \text{dim}(X)$. Since $\text{RatCurves}^e(X)$ is_a_Zariski open set in the stack ${\overline{{\mathcal}M}_{0,0}({X,e})}$,_the result on $\text{RatCurves}^e(X)$_follows.
After_establishing irreducibility and the_dimension of the spaces $\text{RatCurves}^e(X)$,_the next question is to determine_the Kodaira_dimension of_$\text{RatCurves}^e(X)$. For a general Fano hypersurface $X_d \subset {\mathbb{P}}^n$ with $d\leq_n$, determining the Kodaira dimensions of_$\text{RatCurves |
[4] \xrightarrow[F_i]{\hspace*{.25em}{\varepsilon}_i'\hspace*{.1em}} \nolinebreak[4]}
(I_i, <_{_i}) { \nolinebreak[4] \xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}'\hspace*{.1em}} \nolinebreak[4]}
\cdots { \nolinebreak[4] \xrightarrow[F_k]{\hspace*{.25em}{\varepsilon}_k'\hspace*{.1em}} \nolinebreak[4]}
(I_0, <_{_0}),$$ for some ${\varepsilon}_i',\ldots,{\varepsilon}_k'\in\{+,-\}$. If ${\varepsilon}_i'=+$, let $\widetilde{<_{_i}}$ be equal to $<_{_i}$ and ${\varepsilon}_{i+1}'':={\varepsilon}_{i+1}'$. Otherwise, let $\widetilde{<_{_i}}$ be equal to $>_{_i}$ and ${\varepsilon}_{i+1}'':=-{\varepsilon}_{i+1}'$. According to, we obtain $$\begin{gathered}
(I_0, <_{_0}) { \nolinebreak[4] \xrightarrow[F_1]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]}
\cdots { \nolinebreak[4] \xrightarrow[F_{i-1}]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]}
(I_{i-1}, \widetilde{<_{i-1}}) { \nolinebreak[4] \xrightarrow[F_i]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]}
(I_i, \widetilde{<_{_i}}) { \nolinebreak[4] \xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}''\hspace*{.1em}} \nolinebreak[4]}\\
(I_i, <_{i+1}) { \nolinebreak[4] \xrightarrow[F_{i+2}]{\hspace*{.25em}{\varepsilon | [ 4 ] \xrightarrow[F_i]{\hspace*{.25em}{\varepsilon}_i'\hspace*{.1em } } \nolinebreak[4 ] }
(I_i, < _ { _ i }) { \nolinebreak[4 ] \xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}'\hspace*{.1em } } \nolinebreak[4 ] }
\cdots { \nolinebreak[4 ] \xrightarrow[F_k]{\hspace*{.25em}{\varepsilon}_k'\hspace*{.1em } } \nolinebreak[4 ] }
(I_0, < _ { _ 0}),$$ for some $ { \varepsilon}_i',\ldots,{\varepsilon}_k'\in\{+,-\}$. If $ { \varepsilon}_i'=+$, let $ \widetilde{<_{_i}}$ be equal to $ < _ { _ i}$ and $ { \varepsilon}_{i+1}'':={\varepsilon}_{i+1}'$. Otherwise, let $ \widetilde{<_{_i}}$ be adequate to $ > _ { _ i}$ and $ { \varepsilon}_{i+1}'':=-{\varepsilon}_{i+1}'$. harmonize to, we obtain $ $ \begin{gathered }
(I_0, < _ { _ 0 }) { \nolinebreak[4 ] \xrightarrow[F_1]{\hspace*{.25em}+\hspace*{.1em } } \nolinebreak[4 ] }
\cdots { \nolinebreak[4 ] \xrightarrow[F_{i-1}]{\hspace*{.25em}+\hspace*{.1em } } \nolinebreak[4 ] }
(I_{i-1 }, \widetilde{<_{i-1 } }) { \nolinebreak[4 ] \xrightarrow[F_i]{\hspace*{.25em}+\hspace*{.1em } } \nolinebreak[4 ] }
(I_i, \widetilde{<_{_i } }) { \nolinebreak[4 ] \xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}''\hspace*{.1em } } \nolinebreak[4]}\\
(I_i, < _ { i+1 }) { \nolinebreak[4 ] \xrightarrow[F_{i+2}]{\hspace*{.25em}{\varepsilon | [4] \xgightarrow[F_i]{\hspace*{.25em}{\vareksilon}_i'\hspace*{.1em}} \uilinebceak[4]}
(J_i, <_{_i}) { \nulinebreak[4] \xrightarrow[F_{i+1}]{\hspece*{.25en}{\varekfilon}_{i+1}'\hspace*{.1em}} \nouinebreak[4]}
\cdots { \nioinebreak[4] \xrightavxow[F_k]{\gdpacz*{.25en}{\varepsilon}_k'\hskace*{.1em}} \nolitebreak[4]}
(I_0, <_{_0}),$$ fmr slme ${\varepsilon}_i',\ldots,{\varepsilon}_k'\in\{+,-\}$. Is ${\varepxipon}_i'=+$, let $\widetylde{<_{_p}}$ fe eslao to $<_{_i}$ and ${\varepsilon}_{i+1}'':={\varepsimon}_{i+1}'$. Otierwise, let $\widrtilde{<_{_i}}$ be equal to $>_{_i}$ and ${\gareosilon}_{i+1}'':=-{\varepsilon}_{i+1}'$. According jk, wq obtain $$\begiv{gathered}
(I_0, <_{_0}) { \nolingbreak[4] \xrightarrow[F_1]{\hspace*{.25em}+\hspaze*{.1em}} \nolinebreqk[4]}
\fgots { \noliiebreah[4] \xrightarrow[F_{i-1}]{\hspawe*{.25em}+\hspsce*{.1em}} \nolinebveak[4]}
(I_{u-1}, \widetilde{<_{i-1}}) { \nolineureak[4] \xrightarrow[F_i]{\rspace*{.25em}+\hvpcce*{.1em}} \nolinebreak[4]}
(U_i, \widejilde{<_{_h}}) { \vilivebdeek[4] \xrighharcow[F_{i+1}]{\hspace*{.25sm}{\varepsilob}_{i+1}''\hspace*{.1em}} \nolinebtewj[4]}\\
(I_i, <_{i+1}) { \nominebrqah[4] \xrightarrow[F_{i+2}]{\hspace*{.25em}{\varepsilon | [4] \xrightarrow[F_i]{\hspace*{.25em}{\varepsilon}_i'\hspace*{.1em}} \nolinebreak[4]} (I_i, <_{_i}) { \nolinebreak[4] \cdots \nolinebreak[4] \xrightarrow[F_k]{\hspace*{.25em}{\varepsilon}_k'\hspace*{.1em}} (I_0, <_{_0}),$$ for $\widetilde{<_{_i}}$ equal to $<_{_i}$ ${\varepsilon}_{i+1}'':={\varepsilon}_{i+1}'$. Otherwise, let be equal to $>_{_i}$ and ${\varepsilon}_{i+1}'':=-{\varepsilon}_{i+1}'$. to, we obtain $$\begin{gathered} (I_0, <_{_0}) { \nolinebreak[4] \xrightarrow[F_1]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]} \cdots { \nolinebreak[4] \nolinebreak[4]} (I_{i-1}, \widetilde{<_{i-1}}) { \nolinebreak[4] \xrightarrow[F_i]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]} (I_i, \widetilde{<_{_i}}) { \nolinebreak[4] \xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}''\hspace*{.1em}} \nolinebreak[4]}\\ <_{i+1}) \nolinebreak[4] | [4] \xrightarrow[F_i]{\hspace*{.25em}{\varePsilon}_i'\hspAce*{.1em}} \NolIneBrEak[4]}
(I_I, <_{_i}) { \noLinebreak[4] \xrighTArroW[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}'\HspacE*{.1eM}} \NoliNEbReak[4]}
\cDots { \nolINeBREak[4] \XrIgHtaRrOW[F_K]{\hspaCe*{.25eM}{\varepsIlon}_k'\hspacE*{.1em}} \NoLinebreak[4]}
(I_0, <_{_0}),$$ foR SoMe ${\varepsilOn}_i',\Ldots,{\varepsiLon}_K'\in\{+,-\}$. If ${\vArEpsILon}_i'=+$, lEt $\wIdetiLde{<_{_i}}$ be EQual to $<_{_I}$ and ${\varepSiLOn}_{i+1}'':={\varEPsilon}_{i+1}'$. oTHeRwisE, let $\widetilde{<_{_i}}$ be eQUaL To $>_{_i}$ and ${\varepsilOn}_{i+1}'':=-{\varEpSIlON}_{I+1}'$. AcCorDing to, we obTaIn $$\begIN{gatherED}
(I_0, <_{_0}) { \NOLIneBReak[4] \xrightarrOw[F_1]{\hspace*{.25em}+\HSpaCe*{.1em}} \noLiNebREak[4]}
\cdoTs { \nolInEBreAk[4] \xrightarrOw[F_{i-1}]{\Hspace*{.25em}+\hSpace*{.1eM}} \NolinebREak[4]}
(I_{i-1}, \wiDetildE{<_{i-1}}) { \nOliNebrEAk[4] \XrIghTaRRow[f_I]{\hSpaCE*{.25em}+\Hspace*{.1em}} \NoLiNebreAk[4]}
(I_i, \WIDETildE{<_{_i}}) { \nOlinEbreaK[4] \xrightarrow[F_{I+1}]{\hsPace*{.25EM}{\vaRepsiLon}_{i+1}''\hSpacE*{.1eM}} \noliNebreaK[4]}\\
(I_i, <_{i+1}) { \nOlInebreak[4] \xrightaRrow[f_{i+2}]{\hspace*{.25eM}{\vaRePsiLoN | [4] \xrightarrow[F_i]{\hs pace*{.25e m}{\v are psi lo n}_i '\hs pace*{.1em}} \ noli nebreak[4]}
(I_i, < _{_i} ){ \n o li nebre ak[4] \ xr i g hta rr ow [F_ {i + 1} ]{\hs pac e*{.25e m}{\vareps ilo n} _{i+1}'\hspa c e* {.1em}} \ nol inebreak[4]}
\cdot s{ \ nolin ebr eak[4 ] \xr i ghtarr ow[F_k]{\ hs p ace*{. 2 5em}{\v a r ep silo n}_k'\hspace*{.1e m }} \nolinebreak[ 4]}
( I _0 , <_{ _0} ),$$ for s om e ${\ v arepsil o n} _ i ' ,\l d ots,{\varepsi lon}_k'\in\ { +,- \}$. I f${\ v arepsi lon}_ i' = +$, let $\wide tild e{<_{_i}} $ be e q ual to$ <_{_i}$ and $ {\v are psil o n} _{ i+1 }' ' :={ \ va rep s ilo n}_{i+1} '$ .Other wise , l e t $\ wid etil de{<_ {_i}}$ be equ alto $ > _{_ i}$ a nd ${ \var ep silon }_{i+1 }'':= -{ \varepsilon}_{i +1}' $. Accord ing t o,we obta i n $$\b egi n{g athered }
(I _ 0,<_ { _ 0 }) { \nolinebreak[4 ]\ xr ightarro w[F_1] { \h sp a ce*{.25e m} +\h spac e * {.1em }} \ no linebrea k[4]} \c dots { \ noline br eak [4] \xr i ghta rrow[F _{i-1}]{ \hspa c e*{.25em}+\hsp a ce*{.1em}} \ n ol i n eb r eak[ 4]}
(I_{i-1 }, \ w idet ilde { <_ {i- 1 }}) { \no li n eb r eak[4] \xrightarro w[ F_i]{\ hspac e*{.25em}+\hs pace*{.1em } } \noline brea k [4 ] }
(I_i, \wi detil de{<_{_i}} ) { \nol inebr eak[4] \xrightar r o w[F_{i+1 }]{ \hs pac e*{ . 2 5e m}{\varepsilo n } _{i+ 1} ''\hspa ce* {.1em}} \ nol ine bre ak [4]}\\
(I_i, < _{ i+ 1} ){ \noli n ebreak[4 ] \x ri ght arrow [ F_{i+2 }]{\h spac e* {. 2 5em }{\vare p si l o n | [4] _\xrightarrow[F_i]{\hspace*{.25em}{\varepsilon}_i'\hspace*{.1em}} _\nolinebreak[4]}
(I_i,_<_{_i}) {__\nolinebreak[4] _\xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}'\hspace*{.1em}}_ \nolinebreak[4]}
_ \cdots {_ \nolinebreak[4] \xrightarrow[F_k]{\hspace*{.25em}{\varepsilon}_k'\hspace*{.1em}}_ \nolinebreak[4]}
__(I_0, <_{_0}),$$ for some ${\varepsilon}_i',\ldots,{\varepsilon}_k'\in\{+,-\}$. If ${\varepsilon}_i'=+$, let $\widetilde{<_{_i}}$ be equal to $<_{_i}$ and_${\varepsilon}_{i+1}'':={\varepsilon}_{i+1}'$._Otherwise, let_$\widetilde{<_{_i}}$_be_equal to $>_{_i}$ and ${\varepsilon}_{i+1}'':=-{\varepsilon}_{i+1}'$._According to, we obtain $$\begin{gathered}
_ _(I_0, <_{_0}) { \nolinebreak[4] \xrightarrow[F_1]{\hspace*{.25em}+\hspace*{.1em}} _\nolinebreak[4]}
_ \cdots_{ \nolinebreak[4] \xrightarrow[F_{i-1}]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]}
_(I_{i-1}, \widetilde{<_{i-1}}) { \nolinebreak[4] _\xrightarrow[F_i]{\hspace*{.25em}+\hspace*{.1em}} \nolinebreak[4]}
___(I_i, \widetilde{<_{_i}}) { _\nolinebreak[4] \xrightarrow[F_{i+1}]{\hspace*{.25em}{\varepsilon}_{i+1}''\hspace*{.1em}} \nolinebreak[4]}\\
_ (I_i, <_{i+1}) { \nolinebreak[4]_ \xrightarrow[F_{i+2}]{\hspace*{.25em}{\varepsilon |
_{k=1}^{r}Q_k\gamma_k\right) \nonumber \\
&=& \prod_{j=1}^{\infty}\prod_{\rho=1}^{d(\nu)}\sum_n
\exp\!\left[n\left(-\beta E_j + i\sum_{k=1}^{r}
q_k^{(\rho)}\gamma_k\right) \right].
\label{hatZ}\end{aligned}$$ In the last step, we have expressed the trace in the basis of $n$ -particle Hamiltonian eigenstates. The $q_k^{(\rho)}$ are the conserved charges, and the $\gamma_k$ are the variables of the Cartan subgroup of the group $G$ of rank $r$. Eq. (\[hatZ\]) resembles the grand canonical partition function, and is actually obtained from it by the Wick rotation: $\beta\mu_i \rightarrow -i\gamma_i$.
As an example, let us choose the internal symmetry of the system correspond to $U(1)_{q_1}\times\cdots\times U(1)_{q_r}$, where the $q_i$ are the conserved charges. The character of $U(1)_{q_i}$ is $e^{iq_i\gamma_i}$, so the character of the direct product group is $\exp(i\sum_{i=1}^r q_i\gamma_i)$. The canonical partition function respecting the exact conservation of charges $q_i$ has now the form $$\begin{aligned}
Z_{q_1,\ldots,q_r}(T,V) &=&
\frac{1}{(2\pi)^r}\int_{0}^{2\pi}{\rm d}\gamma_1\cdots\int_{0}^{2\pi}
{\rm d}\gamma_r \nonumber \\
&&\times \exp\!\left[-i\sum_{i=1}^r q_i \gamma_i\right]
\hat{Z}(T,V,\gamma_1,\cdots,\gamma_r).\end{aligned}$$ The special cases, $Z_S$ and $Z_{B,Q,S}$ for a Boltzmannian hadron resonance gas are considered in previous sections.
Summary
=======
The particle abundances have been computed in the canonical formalism using the formulation for the exact conservation of baryon number, strangeness and charge in the thermal | _ { k=1}^{r}Q_k\gamma_k\right) \nonumber \\
& = & \prod_{j=1}^{\infty}\prod_{\rho=1}^{d(\nu)}\sum_n
\exp\!\left[n\left(-\beta E_j + i\sum_{k=1}^{r }
q_k^{(\rho)}\gamma_k\right) \right ].
\label{hatZ}\end{aligned}$$ In the last step, we have expressed the trace in the footing of $ n$ -particle Hamiltonian eigenstates. The $ q_k^{(\rho)}$ are the conserve charges, and the $ \gamma_k$ are the variables of the Cartan subgroup of the group $ G$ of membership $ r$. Eq. (\[hatZ\ ]) resemble the grand canonical partition affair, and is actually obtained from it by the Wick rotation: $ \beta\mu_i \rightarrow -i\gamma_i$.
As an example, permit us choose the internal symmetry of the organization correspond to $ U(1)_{q_1}\times\cdots\times U(1)_{q_r}$, where the $ q_i$ are the conserved charge. The fictional character of $ U(1)_{q_i}$ is $ e^{iq_i\gamma_i}$, so the character of the direct product group is $ \exp(i\sum_{i=1}^r q_i\gamma_i)$. The canonical partition affair respecting the exact conservation of charges $ q_i$ receive now the form $ $ \begin{aligned }
Z_{q_1,\ldots, q_r}(T, V) & = &
\frac{1}{(2\pi)^r}\int_{0}^{2\pi}{\rm d}\gamma_1\cdots\int_{0}^{2\pi }
{ \rm d}\gamma_r \nonumber \\
& & \times \exp\!\left[-i\sum_{i=1}^r q_i \gamma_i\right ]
\hat{Z}(T, V,\gamma_1,\cdots,\gamma_r).\end{aligned}$$ The special casing, $ Z_S$ and $ Z_{B, Q, S}$ for a Boltzmannian hadron resonance gas are considered in previous sections.
Summary
= = = = = = =
The atom abundances have been computed in the canonical formalism using the formulation for the exact conservation of baryon numeral, strangeness and charge in the thermal | _{k=1}^{r}Q_n\gamma_k\right) \nonumber \\
&=& \pvod_{j=1}^{\infty}\prod_{\rho=1}^{b(\bu)}\sum_n
\xxp\!\left[h\left(-\betx E_j + i\sum_{k=1}^{r}
q_k^{(\rho)}\gamma_k\rigit) \rught].
\lqbel{hatZ}\end{aligned}$$ In ghe last dtep, we yave wxpressed vge tracc in fme bavms of $n$ -particlg Hamiltoniat eigenstates. Dhd $e_k^{(\rho)}$ are the conserved charges, and the $\gakmw_k$ are the varyablts jf tgv Gartan subgroup of the group $G$ or rank $g$. Eq. (\[hatZ\]) resemblrs the grand canonical parhitiln function, and is actually ovtaigwd from it bh the Wick rotation: $\beja\mu_i \rightarrow -i\gamma_i$.
As an exxmple, let us chiowe hve internal symmvtry of the snxtem cmrrespomd to $U(1)_{q_1}\times\ccotv\tines U(1)_{q_r}$, where the $q_i$ ere the conserved chwrges. The ckaracter of $U(1)_{q_i}$ is $e^{iw_i\tamma_h}$, so the xhafacuer or the firxct product group is $\ezp(i\sum_{i=1}^r q_i\gamma_i)$. Tne bsnonical parfition ftnction respecting the exact conservatimn kf charges $q_i$ has now tye form $$\begin{aligned}
Z_{e_1,\ldots,q_r}(T,D) &=&
\frac{1}{(2\pi)^r}\int_{0}^{2\pi}{\rm d}\gamma_1\cdots\int_{0}^{2\pi}
{\rm d}\gamma_r \nonukber \\
&&\vioes \ewp\!\lewr[-i\dum_{i=1}^r q_i \gamma_i\right]
\hat{Z}(T,V,\gamma_1,\cdots,\gamma_r).\end{wmibnvd}$$ The special cafes, $Z_S$ and $A_{B,E,S}$ sor a Boltzmavnian kzddon resonance gas wre confiderwd in predioux sections.
Summary
=======
The particoe abundancef have been computeb in the canunicsl fotmalism using the formuuatikn for the fxact conadrvation of baryun kumter, strangeness and charge in the tiermak | _{k=1}^{r}Q_k\gamma_k\right) \nonumber \\ &=& \prod_{j=1}^{\infty}\prod_{\rho=1}^{d(\nu)}\sum_n \exp\!\left[n\left(-\beta E_j q_k^{(\rho)}\gamma_k\right) \label{hatZ}\end{aligned}$$ In last step, we the of $n$ -particle eigenstates. The $q_k^{(\rho)}$ the conserved charges, and the $\gamma_k$ the variables of the Cartan subgroup of the group $G$ of rank $r$. (\[hatZ\]) resembles the grand canonical partition function, and is actually obtained from it the rotation: \rightarrow As an example, let us choose the internal symmetry of the system correspond to $U(1)_{q_1}\times\cdots\times U(1)_{q_r}$, the $q_i$ are the conserved charges. The character $U(1)_{q_i}$ is $e^{iq_i\gamma_i}$, so character of the direct product is q_i\gamma_i)$. The partition respecting exact conservation of $q_i$ has now the form $$\begin{aligned} Z_{q_1,\ldots,q_r}(T,V) &=& \frac{1}{(2\pi)^r}\int_{0}^{2\pi}{\rm d}\gamma_1\cdots\int_{0}^{2\pi} {\rm d}\gamma_r \nonumber \\ &&\times \exp\!\left[-i\sum_{i=1}^r q_i \hat{Z}(T,V,\gamma_1,\cdots,\gamma_r).\end{aligned}$$ The $Z_S$ and for Boltzmannian resonance gas are previous sections. Summary ======= The particle computed in the canonical formalism using the formulation the exact of baryon number, strangeness and charge the thermal | _{k=1}^{r}Q_k\gamma_k\right) \nonumber \\
&=& \prOd_{j=1}^{\infty}\prOd_{\rho=1}^{D(\nu)}\Sum_N
\eXp\!\leFt[n\lEft(-\beta E_j + i\sum_{k=1}^{R}
Q_k^{(\rhO)}\gamma_k\right) \right].
\label{HatZ}\eNd{ALignED}$$ IN the lAst step, WE hAVE exPrEsSed ThE TrAce in The Basis of $N$ -particle HAmiLtOnian eigenstATeS. The $q_k^{(\rho)}$ aRe tHe conserved cHarGes, and ThE $\gaMMa_k$ arE thE variAbles oF The CarTan subgroUp OF the grOUp $G$ of raNK $R$. EQ. (\[hatz\]) resembles the granD CaNOnical partitioN functIoN, AnD IS acTuaLly obtaineD fRom it BY the WicK RoTATIon: $\BEta\mu_i \rightarRow -i\gamma_i$.
AS An eXample, LeT us CHoose tHe intErNAl sYmmetry of thE sysTem corresPond to $u(1)_{Q_1}\times\cDOts\timeS U(1)_{q_r}$, whEre The $Q_i$ arE ThE cOnsErVEd cHArGes. tHe cHaracter Of $u(1)_{q_I}$ is $e^{iQ_i\gaMMA_I}$, So thE chAracTer of The direct prodUct GrouP Is $\eXp(i\suM_{i=1}^r q_i\GammA_i)$. the caNonicaL partItIon function respEctiNg the exacT coNsErvAtIon of CHarges $Q_i$ hAs nOw the foRm $$\begin{ALigNeD}
z_{Q_1,\LdOts,q_r}(T,V) &=&
\frac{1}{(2\pi)^r}\int_{0}^{2\pI}{\rM D}\GaMma_1\cdots\Int_{0}^{2\pi}
{\rM D}\gAmMA_r \nonumbEr \\
&&\TimEs \exP\!\LEft[-i\sUm_{i=1}^r Q_I \gAmma_i\rigHt]
\hat{Z}(t,v,\gAmMa_1,\cdots,\GaMma_r).\enD{aLigNed}$$ the spECial Cases, $Z_s$ and $Z_{B,Q,S}$ For a BOLtzmannian hadrON resonance gas ARe CONsIDereD in Previous secTionS.
summAry
=======
THE pArtICle abUndanCeS HaVE been computed in the cAnOnical FormaLism using the fOrmulation FOR The exact ConsERvATion of baryon nuMber, sTrangeness ANd charge In the Thermal | _{k=1}^{r}Q_k\gamma_k\righ t) \nonumb er \\
&= & \ pr od_{ j=1} ^{\infty}\prod _ {\rh o=1}^{d(\nu)}\sum_n
\e xp\!\ le f t[n\ l ef t(-\b eta E_j + i\s um _{ k=1 }^ { r}
q_k^ {(\ rho)}\g amma_k\rig ht) \ right].
\lab e l{ hatZ}\end{ ali gned}$$ In t helast s te p,w e hav e e xpres sed th e trace in the b as i s of $ n $ -part i c le Ham iltonian eigensta t es . The $q_k^{(\r ho)}$ar e t h e co nse rved charg es , and the $\g a mm a _ k $ a r e the variabl es of the C a rta n subg ro upo f thegroup $ G $ o f rank $r$. Eq. (\[hatZ\ ]) res e mbles t h e grand canon ica l p arti t io nfun ct i on, an d i s ac tually o bt ai ned f romi t b y th e W ickrotat ion: $\beta\m u_i \ri g hta rrow-i\ga mma_ i$ .
As an ex ample ,let us choose t he i nternal s ymm et ryof thes ystemcor res pond to $U(1)_ { q_1 }\ t i m es \cdots\times U(1)_ {q _ r }$ , wherethe $q _ i$ a r e the co ns erv ed c h a rges. The ch aracterof $U( 1 )_ {q _i}$ is $ e^{iq_ i\ gam ma_ i}$,s o th e char acter of thed irect productg roup is $\exp ( i\ s u m_ { i=1} ^rq_i\gamma_i )$.T he c anon i ca l p a rtiti on fu nc t io n respecting the exa ct conse rvati on of charges $q_i$ has n o w the fo rm $ $ \b e gin{aligned}
Z _{q_1 ,\ldots,q_ r }(T,V) & =&
\ frac{1}{ (2\pi)^r} \ i nt_{0}^{ 2\p i}{ \rm d} \ g am ma_1\cdots\in t _ {0}^ {2 \pi}
{\ rmd}\gamm a_r \n onu mbe r\\
&&\tim es \exp\ !\ le ft [- i\s um_{i = 1}^r q_i \ gam ma _i\ right ]
\hat{ Z}(T, V,\g am ma _ 1,\ cdots,\ g am m a _r). \e nd {ali gne d} $$ Th e sp e cia l cases , $Z_S$ a nd$ Z_{B ,Q ,S }$ fora Boltzmannia nhadron res on anc e gasa r e consid ered in previous sectio n s.
Sum mar y
=== ====
The par tic le abu nda n ces ha ve bee n com pu ted i n the c an oni ca l formalis m usi ng th eform ulation for the exact con s erv ation of bary onnumb e r ,str a ng e nes sa ndc h arge in the the rmal | _{k=1}^{r}Q_k\gamma_k\right) \nonumber_\\
&=& \prod_{j=1}^{\infty}\prod_{\rho=1}^{d(\nu)}\sum_n
\exp\!\left[n\left(-\beta_E_j + _i\sum_{k=1}^{r}
q_k^{(\rho)}\gamma_k\right) \right].
\label{hatZ}\end{aligned}$$_In_the last_step,_we have expressed_the trace in_the basis of $n$_-particle Hamiltonian eigenstates._The_$q_k^{(\rho)}$ are the conserved charges, and the $\gamma_k$ are the variables of the Cartan_subgroup_of the_group_$G$_of rank $r$. Eq. (\[hatZ\])_resembles the grand canonical partition_function, and_is actually obtained from it by the Wick_rotation:_$\beta\mu_i \rightarrow -i\gamma_i$.
As_an example, let us choose the internal symmetry of_the system correspond to $U(1)_{q_1}\times\cdots\times U(1)_{q_r}$,_where the $q_i$_are_the_conserved charges. The character_of $U(1)_{q_i}$ is $e^{iq_i\gamma_i}$, so the_character of the direct product group_is $\exp(i\sum_{i=1}^r q_i\gamma_i)$. The canonical partition function_respecting the exact conservation of charges_$q_i$ has now the form_$$\begin{aligned}
Z_{q_1,\ldots,q_r}(T,V) &=&_
\frac{1}{(2\pi)^r}\int_{0}^{2\pi}{\rm d}\gamma_1\cdots\int_{0}^{2\pi}
{\rm d}\gamma_r \nonumber \\
&&\times_\exp\!\left[-i\sum_{i=1}^r q_i \gamma_i\right]
\hat{Z}(T,V,\gamma_1,\cdots,\gamma_r).\end{aligned}$$_The special_cases, $Z_S$ and_$Z_{B,Q,S}$ for a Boltzmannian hadron resonance_gas are considered_in previous sections.
Summary
=======
The particle abundances have_been_computed in the_canonical_formalism_using the_formulation for the_exact_conservation of_baryon_number, strangeness and charge in the_thermal |
, DFG and GSI.
[10]{}
J. C. LeGuillou and J. Zinn-Justin, [*Large-Order Behaviour of Perturbation Theory*]{} (North-Holland, Amsterdam, 1990).
J. Zinn-Justin, [*Quantum Field Theory and Critical Phenomena*]{}, 3rd ed. (Clarendon Press, Oxford, 1996).
J. Fischer, Int. J. Mod. Phys. A [**12**]{}, 3625 (1997).
U. D. Jentschura, E. Weniger, and G. Soff, Asymptotic Improvement of Resummation and Perturbative Predictions, Los Alamos preprint hep-ph/0005198, submitted.
D. Broadhurst and D. Kreimer, Phys. Lett. B [**475**]{}, 63 (2000).
I. Caprini and J. Fischer, Phys. Rev. D [**60**]{}, 054014 (1999).
I. Caprini and J. Fischer, Convergence of the expansion of the Laplace-Borel integral in perturbative QCD improved by conformal mapping, Los Alamos preprint hep-ph/0002016.
U. D. Jentschura, Resummation of Nonalternating Divergent Perturbative Expansions, Los Alamos preprint hep-ph/0001135, Phys. Rev. D (in press).
P. A. Raczka, Phys. Rev. D [**43**]{}, R9 (1991).
M. Pindor, Padé Approximants and Borel Summation for QCD Perturbation Series, Los Alamos preprint hep-th/9903151.
R. Seznec and J. Zinn-Justin, J. Math. Phys. [**20**]{}, 1398 (1979).
J. C. L. Guillou and J. Zinn-Justin, Ann. Phys. (N. Y.) [**147**]{}, 57 (1983).
R. Guida, K. Konishi, and H. Suzuki, Ann. Phys. (N. Y.) [**241**]{}, 152 (1995).
D. J. Broadhurst, P. A. Baikov, V. A. Ilyin, J. Fleischer, O. V. | , DFG and GSI.
[ 10 ] { }
J. C. LeGuillou and J. Zinn - Justin, [ * Large - Order Behaviour of Perturbation Theory * ] { } (North - Holland, Amsterdam, 1990).
J. Zinn - Justin, [ * Quantum Field Theory and Critical Phenomena * ] { }, 3rd ed. (Clarendon Press, Oxford, 1996).
J. Fischer, Int. J. Mod. Phys. A [ * * 12 * * ] { }, 3625 (1997).
U. D. Jentschura, E. Weniger, and G. Soff, Asymptotic Improvement of Resummation and Perturbative Predictions, Los Alamos preprint hep - ph/0005198, submit.
D. Broadhurst and D. Kreimer, Phys. Lett. boron [ * * 475 * * ] { }, 63 (2000).
I. Caprini and J. Fischer, Phys. Rev. D [ * * 60 * * ] { }, 054014 (1999).
I. Caprini and J. Fischer, Convergence of the expansion of the Laplace - Borel integral in perturbative QCD improved by conformal mapping, Los Alamos preprint hep - ph/0002016.
U. D. Jentschura, Resummation of Nonalternating Divergent Perturbative Expansions, Los Alamos preprint hep - ph/0001135, Phys. Rev. D (in crush).
P. A. Raczka, Phys. Rev. D [ * * 43 * * ] { }, R9 (1991).
M. Pindor, Padé Approximants and Borel Summation for QCD Perturbation Series, Los Alamos preprint hep - th/9903151.
R. Seznec and J. Zinn - Justin, J. Math. Phys. [ * * 20 * * ] { }, 1398 (1979).
J. C. L. Guillou and J. Zinn - Justin, Ann. Phys. (N. Y.) [ * * 147 * * ] { }, 57 (1983).
R. Guida, K. Konishi, and H. Suzuki, Ann. Phys. (N. Y.) [ * * 241 * * ] { }, 152 (1995).
D. J. Broadhurst, P. A. Baikov, V. A. Ilyin, J. Fleischer, O. V. | , DFH and GSI.
[10]{}
J. C. LeGuillou ana J. Zinn-Justin, [*Large-Orver Behzviour ow Perturbation Theory*]{} (North-Hlloand, Qmsterdam, 1990).
J. Zinn-Justin, [*Quantum Vield Thwory qnd Criticem Phenomena*]{}, 3rs ed. (Elerendon Press, Owford, 1996).
J. Fiswher, Int. J. Mod. Pfyd. A [**12**]{}, 3625 (1997).
U. D. Jentschura, E. Weniger, and G. Soff, Axylptotic Improvgment jf Rssummation and Perturbative Predicfions, Lms Alamos prelrint hep-ph/0005198, submitted.
D. Browdhugst and D. Kreimer, Ohys. Lett. B [**475**]{}, 63 (2000).
I. Xaprini and G. Fischer, Khvs. Rev. D [**60**]{}, 054014 (1999).
J. Caprini and J. Fischer, Convergdnce pf the expqnwioj of the Lapoace-Bjrel integral in perdurbatife QCD improvec bb cobformal mapping, Los Anamos preprint hep-kh/0002016.
U. D. Jentswhbra, Resummation of Nobaoterndtinc Dixwrgdnt Pxrthrbatige Xxpansions, Mos Alamos preprint hep-ph/0001135, Phys. Rqn. D (in press).
P. Z. Raczha, Phys. Rev. D [**43**]{}, R9 (1991).
M. Pindor, Padé Approximantv ahd Borel Summation for WCD Perturbation Serigs, Los Alaios preprint hep-th/9903151.
R. Seznec and J. Zinn-Justin, J. Matv. Phya. [**20**]{}, 1398 (1979).
L. G. J. Tulllou and J. Zinn-Justin, Ann. Phys. (N. Y.) [**147**]{}, 57 (1983).
R. Guida, I. Lokishi, and H. Suzuhi, Ann. Phys. (N. Y.) [**241**]{}, 152 (1995).
D. J. Broadhurrt, P. A. Yziiov, V. A. Ilyin, J. Flelscher, J. V. | , DFG and GSI. [10]{} J. C. J. [*Large-Order Behaviour Perturbation Theory*]{} (North-Holland, Field and Critical Phenomena*]{}, ed. (Clarendon Press, 1996). J. Fischer, Int. J. Mod. A [**12**]{}, 3625 (1997). U. D. Jentschura, E. Weniger, and G. Soff, Asymptotic of Resummation and Perturbative Predictions, Los Alamos preprint hep-ph/0005198, submitted. D. Broadhurst and Kreimer, Lett. [**475**]{}, (2000). I. Caprini and J. Fischer, Phys. Rev. D [**60**]{}, 054014 (1999). I. Caprini and J. Convergence of the expansion of the Laplace-Borel integral perturbative QCD improved by mapping, Los Alamos preprint hep-ph/0002016. D. Resummation of Divergent Expansions, Alamos preprint hep-ph/0001135, Rev. D (in press). P. A. Raczka, Phys. Rev. D [**43**]{}, R9 (1991). M. Pindor, Padé Approximants Borel Summation Perturbation Series, Alamos hep-th/9903151. Seznec and J. Math. Phys. [**20**]{}, 1398 (1979). J. and J. Zinn-Justin, Ann. Phys. (N. Y.) [**147**]{}, (1983). R. K. Konishi, and H. Suzuki, Ann. (N. Y.) [**241**]{}, 152 (1995). D. J. Broadhurst, A. Baikov, V. A. Ilyin, J. Fleischer, O. V. | , DFG and GSI.
[10]{}
J. C. LeGuillou and J. ZInn-Justin, [*LArge-ORdeR BeHaViouR of PErturbation TheORy*]{} (NoRth-Holland, Amsterdam, 1990).
J. ZiNn-JusTiN, [*quanTUm field theory aND CRITicAl phEnoMeNA*]{}, 3rD ed. (ClAreNdon PreSs, Oxford, 1996).
J. FIscHeR, Int. J. Mod. Phys. a [**12**]{}, 3625 (1997).
u. D. jentschura, e. WeNiger, and G. SofF, AsYmptotIc impROvemeNt oF ResuMmatioN And PerTurbative prEDictioNS, Los AlaMOS pReprInt hep-ph/0005198, submitted.
d. brOAdhurst and D. KreImer, PhYs. lEtT. b [**475**]{}, 63 (2000).
i. CaPriNi and J. FiscHeR, Phys. rEv. D [**60**]{}, 054014 (1999).
I. CapRInI AND J. FIScher, ConvergeNce of the expANsiOn of thE LAplACe-BoreL inteGrAL in PerturbativE QCD Improved bY confoRMal mappINg, Los AlAmos prEprInt Hep-pH/0002016.
u. D. jeNtsChURa, RESuMmaTIon Of NonaltErNaTing DIverGENT pertUrbAtivE ExpaNsions, Los AlamOs pReprINt hEp-ph/0001135, PHys. ReV. D (in PrEss).
P. A. raczka, phys. REv. d [**43**]{}, R9 (1991).
M. Pindor, Padé ApProxImants and borEl sumMaTion fOR QCD PeRtuRbaTion SerIes, Los ALAmoS pREPRiNt hep-th/9903151.
R. Seznec and J. ziNN-juStin, J. MatH. Phys. [**20**]{}, 1398 (1979).
J. c. l. GUiLLou and J. ZInN-JuStin, aNN. Phys. (n. Y.) [**147**]{}, 57 (1983).
R. GUIdA, K. KonishI, and H. SUZuKi, ann. Phys. (n. Y.) [**241**]{}, 152 (1995).
d. J. BroaDhUrsT, P. A. baikoV, v. A. IlYin, J. FlEischer, O. v. | , DFG and GSI.
[10]{}
J. C. LeGuil lou a ndJ.Zi nn-J usti n, [*Large-Ord e r Be haviour of Perturbatio n The or y *]{} (N orth- Holland , A m s ter da m, 19 90 ) .
J. Z inn -Justin , [*Quantu m F ie ld Theory an d C ritical Ph eno mena*]{}, 3r d e d. (Cl ar end o n Pre ss, Oxfo rd, 19 9 6).
J . Fischer ,I nt. J. Mod. Ph y s .A [* *12**]{}, 3625 (1 9 97 ) .
U. D. Jents chura, E . W e n ige r,and G. Sof f, Asym p totic I m pr o v e men t of Resummati on and Pert u rba tive P re dic t ions,Los A la m ospreprint he p-ph /0005198, submi t ted.
D . Broadh urst a ndD.Krei m er ,Phy s. Let t .B [ * *47 5**]{},63 ( 2000) .
I . C a prin i a nd J . Fis cher, Phys. R ev. D [ * *60 **]{} , 054 014(1 999).
I. C aprin iand J. Fischer, Con vergenceofth e e xp ansio n of th e L apl ace-Bor el inte g ral i n p er turbative QCD impr ov e d b y confor mal ma p pi ng , Los Ala mo s p repr i n t hep -ph/ 0 00 2016.
U . D. J e nt sc hura, R es ummati on of No nalte r nati ng Div ergent P ertur b ative Expansio n s, Los Alamos pr e p ri n t he p-p h/0001135,Phys . Rev . D( in pr e ss).
P. A .R ac z ka, Phys. Rev. D [* *4 3**]{} , R9(1991).
M. P indor, Pad é A pproxima ntsa nd Borel Summatio n for QCD Pertu r bation S eries , Los Al amos prep r i nt hep-t h/9 903 151 .
R . S eznec and J.Z i nn-J us tin, J. Ma th. Phy s.[** 20* *]{ }, 1398 (19 79).
J. C .L. G uil lou a n d J. Zin n- Jus ti n,Ann.P hys. ( N. Y. ) [* *1 47 * *]{ }, 57 ( 1 98 3 ) .
R .Gu ida, K. K onish i, a n d H . Suzuk i, Ann. P hys . (N. Y .) [**241 **]{}, 152 (1 99 5).
D. J. B roa dhurst , P. A. Ba ikov, V. A. Ilyin, J. F l eischer , O . V. | , DFG_and GSI.
[10]{}
J. C._LeGuillou and J. Zinn-Justin,_[*Large-Order Behaviour_of_Perturbation Theory*]{}_(North-Holland,_Amsterdam, 1990).
J. Zinn-Justin,_[*Quantum Field Theory_and Critical Phenomena*]{}, 3rd_ed. (Clarendon Press,_Oxford,_1996).
J. Fischer, Int. J. Mod. Phys. A [**12**]{}, 3625 (1997).
U. D. Jentschura, E. Weniger, and_G._Soff, Asymptotic_Improvement_of_Resummation and Perturbative Predictions, Los_Alamos preprint hep-ph/0005198, submitted.
D. Broadhurst_and D._Kreimer, Phys. Lett. B [**475**]{}, 63 (2000).
I. Caprini_and_J. Fischer, Phys._Rev. D [**60**]{}, 054014 (1999).
I. Caprini and J. Fischer,_Convergence of the expansion of the_Laplace-Borel integral in_perturbative_QCD_improved by conformal mapping,_Los Alamos preprint hep-ph/0002016.
U. D. Jentschura, Resummation_of Nonalternating Divergent Perturbative Expansions, Los_Alamos preprint hep-ph/0001135, Phys. Rev. D (in_press).
P. A. Raczka, Phys. Rev. D [**43**]{},_R9 (1991).
M. Pindor, Padé Approximants_and Borel_Summation for QCD Perturbation Series,_Los Alamos preprint_hep-th/9903151.
R. Seznec_and J. Zinn-Justin,_J. Math. Phys. [**20**]{}, 1398 (1979).
J. C. L._Guillou and J._Zinn-Justin, Ann. Phys. (N. Y.) [**147**]{},_57_(1983).
R. Guida, K._Konishi,_and_H. Suzuki,_Ann. Phys. (N._Y.)_[**241**]{}, 152_(1995).
D. J._Broadhurst, P. A. Baikov, V. A. Ilyin, J._Fleischer,_O. V. |
reflection coefficient $R=\gamma/\delta$. $R_\theta$ represents a wave plate which is used to rotate the horizontal polarization with an angle $\theta=arccos(\beta/\alpha)$. DL denotes a time-delay device which is used to make the two wavepackets of the two spatial modes arrive at PBS$_5$ (or PBS$_6$) in the same time. $D_i$ ($i=1,2,3$) represents a single-photon detector.[]{data-label="figure3.1"}](fig13_hyperECP1.eps){width="7.5cm"}
First, Alice splits the parameter of the spatial-mode state by performing a unitary operation on spatial mode $a_2$, resorting to an unbalanced beam splitter (i.e., UBS) with reflection coefficient $R=\gamma/\delta$ (shown in Fig. \[figure3.1\]a). The state of the photon pair $AB$ is changed from $|\phi_0\rangle_{AB}$ to $|\phi_1\rangle_{AB}$. Here $$\begin{aligned}
% Eq.3.2 % Eq. 24
|\phi_1\rangle_{AB}\!&=&\!(\alpha|HH\rangle+\beta|VV\rangle)_{AB}
\otimes\Big[\gamma(|a_1b_1\rangle
+|a_2b_2\rangle)\nonumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}|a_3b_2\rangle\Big].\end{aligned}$$ If photon $A$ is not detected in the spatial mode $a_3$, the spatial-mode state of the photon pair $AB$ is transformed into a maximally entangled Bell state.
Subsequently, Alice splits the parameter of the polarization state by performing the same polarization unitary operations on the spatial modes $a_1$ and $a_2$ as shown in Fig. \[figure3.1\]b. After two spatial modes $a_1$ and $a_2$ pass through PBSs (i.e., PBS$_1$ and PBS$_2$) and $R_\theta$, the state of the photon pair $AB$ is transformed from $|\phi_1\rangle_{AB}$ to $|\phi_2\rangle_{AB}$. Here $$\begin{aligned}
% Eq.(4) % Eq. 25
|\ | reflection coefficient $ R=\gamma/\delta$. $ R_\theta$ represents a wave plate which is used to revolve the horizontal polarization with an slant $ \theta = arccos(\beta/\alpha)$. DL denotes a time - delay device which is use to cause the two wavepackets of the two spatial modes arrive at PBS$_5 $ (or PBS$_6 $) in the same prison term. $ D_i$ ($ i=1,2,3 $) represent a single - photon detector.[]{data - label="figure3.1"}](fig13_hyperECP1.eps){width="7.5 centimeter " }
First, Alice splits the argument of the spatial - mode state by performing a unitary process on spatial mode $ a_2 $, resorting to an unbalanced beam splitter (i.e., UBS) with contemplation coefficient $ R=\gamma/\delta$ (shown in Fig. \[figure3.1\]a). The state of the photon pair $ AB$ is switch from $ |\phi_0\rangle_{AB}$ to $ |\phi_1\rangle_{AB}$. Here $ $ \begin{aligned }
% Eq.3.2 % Eq. 24
|\phi_1\rangle_{AB}\!&=&\!(\alpha|HH\rangle+\beta|VV\rangle)_{AB }
\otimes\Big[\gamma(|a_1b_1\rangle
+ |a_2b_2\rangle)\nonumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}|a_3b_2\rangle\Big].\end{aligned}$$ If photon $ A$ is not detected in the spatial mode $ a_3 $, the spatial - mode state of the photon couple $ AB$ is transformed into a maximally entangle Bell state.
Subsequently, Alice splits the parameter of the polarization department of state by performing the same polarization unitary operations on the spatial modes $ a_1 $ and $ a_2 $ as shown in Fig. \[figure3.1\]b. After two spatial modes $ a_1 $ and $ a_2 $ pass through PBSs (i.e., PBS$_1 $ and PBS$_2 $) and $ R_\theta$, the state of the photon pair $ AB$ is transformed from $ |\phi_1\rangle_{AB}$ to $ |\phi_2\rangle_{AB}$. Here $ $ \begin{aligned }
% Eq.(4) % Eq. 25
|\ | revlection coefficient $R=\gaoma/\delta$. $R_\theta$ represxnts a save plage which is used to rotate tie hirizobtal polarization with an angle $\theta=arxcos(\ueta/\alpha)$. DL denotes a tlie-demwy dzvmce which is usgd to make tve two wavepacnegs of the two spatial modes arrive at PBS$_5$ (or PHS$_6$) in the same timt. $D_y$ ($i=1,2,3$) dvpvesents a single-photon detector.[]{dzta-laben="figure3.1"}](fig13_hyprrECP1.eps){width="7.5cm"}
First, Alice spllts the parameter lf the spatual-mjee state by oerforming a unitary okeration on spatial mode $a_2$, resorging co an unbalqnxed team splittxr (i.e., LBS) with reflcbtion cmefficirnt $R=\gamma/\delts$ (siown in Fig. \[figure3.1\]a). The stete of the photon payr $AB$ is whcnged from $|\phi_0\rangle_{AV}$ ro $|\phh_1\rancle_{AC}$. Hefe $$\gejin{zlignef}
% Eq.3.2 % Eq. 24
|\pho_1\rwbgle_{AB}\!&=&\!(\alpha|HH\dangle+\feea|VV\rangle)_{AB}
\otimes\Big[\gamma(|a_1b_1\rangle
+|a_2b_2\rancle)\honumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}|a_3v_2\rangle\Big].\end{aligned}$$ Lf photon $A$ is not detected in the spatial mode $a_3$, the spathal-move stcbc of rhf photon pair $AB$ is transformed into a maximajmy ektangled Bell stcte.
Subsequently, Sllcr splits the pxrametzd kf the polarizatioj state by pwrforming the same polarization unitary iperations ou tye spatial modes $a_1$ and $a_2$ as skown im Fig. \[gigure3.1\]b. After two spaticl modss $a_1$ and $a_2$ oass throheh PBSs (i.e., PBS$_1$ avd IBS$_2$) dnd $R_\theux$, the state of thq photon 'air $CB$ is trxnsfprmed srom $|\phi_1\rajgle_{AN}$ to $|\phi_2\rangle_{AB}$. Hege $$\beyin{alhgned}
% Eq.(4) % Xa. 25
|\ | reflection coefficient $R=\gamma/\delta$. $R_\theta$ represents a wave is to rotate horizontal polarization with a device which is to make the wavepackets of the two spatial modes at PBS$_5$ (or PBS$_6$) in the same time. $D_i$ ($i=1,2,3$) represents a single-photon First, Alice splits the parameter of the spatial-mode state by performing a unitary on mode resorting an unbalanced beam splitter (i.e., UBS) with reflection coefficient $R=\gamma/\delta$ (shown in Fig. \[figure3.1\]a). The state the photon pair $AB$ is changed from $|\phi_0\rangle_{AB}$ $|\phi_1\rangle_{AB}$. Here $$\begin{aligned} % % Eq. 24 |\phi_1\rangle_{AB}\!&=&\!(\alpha|HH\rangle+\beta|VV\rangle)_{AB} \otimes\Big[\gamma(|a_1b_1\rangle \!&&+\sqrt{|\delta|^2-|\gamma|^2}|a_3b_2\rangle\Big].\end{aligned}$$ photon $A$ not in spatial mode $a_3$, spatial-mode state of the photon pair $AB$ is transformed into a maximally entangled Bell state. Subsequently, Alice the parameter polarization state performing same unitary operations on modes $a_1$ and $a_2$ as shown After two spatial modes $a_1$ and $a_2$ pass PBSs (i.e., and PBS$_2$) and $R_\theta$, the state the photon pair $AB$ is transformed from $|\phi_1\rangle_{AB}$ $|\phi_2\rangle_{AB}$. Here $$\begin{aligned} % Eq.(4) % Eq. 25 |\ | reflection coefficient $R=\gamMa/\delta$. $R_\thEta$ rePreSenTs A wavE plaTe which is used tO RotaTe the horizontal polarizAtion WiTH an aNGlE $\thetA=arccos(\BEtA/\ALphA)$. Dl dEnoTeS A tIme-deLay Device wHich is used To mAkE the two wavepACkEts of the twO spAtial modes arRivE at PBS$_5$ (Or pBS$_6$) IN the sAme Time. $D_I$ ($i=1,2,3$) reprESents a Single-phoToN DetectOR.[]{data-laBEL="fIgurE3.1"}](fig13_hyperECP1.eps){wiDTh="7.5CM"}
First, Alice splIts the PaRAmETEr oF thE spatial-moDe State BY perforMInG A UNitARy operation on Spatial mode $A_2$, ResOrting To An uNBalancEd beaM sPLitTer (i.e., UBS) witH refLection coEfficiENt $R=\gammA/\Delta$ (shOwn in FIg. \[fIguRe3.1\]a). THE sTaTe oF tHE phOToN paIR $AB$ Is changeD fRoM $|\phi_0\rAnglE_{ab}$ TO $|\phi_1\RanGle_{Ab}$. Here $$\Begin{aligned}
% EQ.3.2 % Eq. 24
|\Phi_1\rANglE_{AB}\!&=&\!(\alPha|HH\RangLe+\Beta|Vv\ranglE)_{AB}
\otImEs\Big[\gamma(|a_1b_1\ranGle
+|a_2B_2\rangle)\noNumBeR\\
\!&&+\sqRt{|\Delta|^2-|\GAmma|^2}|a_3b_2\RanGle\big].\end{aLigned}$$ IF PhoToN $a$ IS nOt detected in the spaTiAL MoDe $a_3$, the spAtial-mODe StATe of the pHoTon Pair $ab$ Is traNsfoRMeD into a maXimallY EnTaNgled BeLl State.
SUbSeqUenTly, AlICe spLits thE parametEr of tHE polarization sTAte by performiNG tHE SaME polAriZation unitaRy opERatiOns oN ThE spATial mOdes $a_1$ AnD $A_2$ aS Shown in Fig. \[figure3.1\]b. AfTeR two spAtial Modes $a_1$ and $a_2$ pasS through PBsS (I.E., PBS$_1$ and PbS$_2$) anD $r_\tHEta$, the state of tHe phoTon pair $AB$ iS TransforMed frOm $|\phi_1\ranGle_{AB}$ to $|\phI_2\RAngle_{AB}$. HEre $$\BegIn{aLigNED}
% EQ.(4) % Eq. 25
|\ | reflection coefficient $R =\gamma/\d elta$ . $ R_\ th eta$ rep resents a wave plat e which is used to rot ate t he hori z on tal p olariza t io n wit han an gl e $ \thet a=a rccos(\ beta/\alph a)$ .DL denotes a ti me-delay d evi ce which isuse d to m ak e t h e two wa vepac kets o f the t wo spatia lm odes a r rive at P BS $_5$ (or PBS$_6$) int he same time. $D_ i$ ($i =1 , 2, 3 $ ) r epr esents a s in gle-p h oton de t ec t o r .[] { data-label="f igure3.1"}] ( fig 13_hyp er ECP 1 .eps){ width =" 7 .5c m"}
First, Ali ce splits the p a rameter of thespatia l-m ode sta t eby pe rf o rmi n ga u n ita ry opera ti on on s pati a l m ode$a_ 2$,resor ting to an un bal ance d be am sp litte r (i .e ., UB S) wit h ref le ction coefficie nt $ R=\gamma/ \de lt a$(s howni n Fig. \[ fig ure3.1\ ]a). Th e st at e o fthe photon pair $A B$ i schangedfrom $ | \p hi _ 0\rangle _{ AB} $ to $ |\phi _1\r a ng le_{AB}$ . Here $$ \b egin{al ig ned}
% Eq .3.2 % Eq. 24
|\p h i_1\rangle_{A B }\ ! & =& \ !(\a lph a|HH\rangle +\be t a|VV \ran g le )_{ A B}
\o times \B i g[ \ gamma(|a_1b_1\rangl e+|a_2b _2\ra ngle)\nonumbe r\\
\!&&+\ s q r t{|\delt a|^2 - |\ g amma|^2}|a_3b_ 2\ran gle\Big].\ e nd{align ed}$$ If phot on $A$ is n ot detec ted in th e s p a ti al mode $a_3$ , thesp atial-m ode stateofthe ph oto npair $AB$ is tran sf or me dint o a m a ximallyen tan gl edBells tate.
Subs eque nt ly , Al ice spl i ts t he p ar am eter of t he po lari z ati on stat e by perf orm i ng t he s ame pol arization uni ta ry operati on s o n thes p atial mo des $a_1$ and $a_2$ ass hown in Fi g. \[ figu re3.1\]b. Af ter tw o s p atialmodes$a_1$ a nd$ a _2$ p a s sthr ou gh PBSs (i . e .,PBS$_ 1$ and PBS$_2 $) and $R_\theta$, the state of the ph oton p ai r $ A B$ istr a nsf o r med from $|\phi _1\rangle_ {A B }$ to $|\phi _ 2\r an gle_{AB }$. Her e $$\ b egin{al igned}
% Eq .(4) % Eq. 25| \ | reflection_coefficient $R=\gamma/\delta$._$R_\theta$ represents a wave_plate which_is_used to_rotate_the horizontal polarization_with an angle_$\theta=arccos(\beta/\alpha)$. DL denotes a_time-delay device which_is_used to make the two wavepackets of the two spatial modes arrive at PBS$_5$_(or_PBS$_6$) in_the_same_time. $D_i$ ($i=1,2,3$) represents a_single-photon detector.[]{data-label="figure3.1"}](fig13_hyperECP1.eps){width="7.5cm"}
First, Alice splits the_parameter of_the spatial-mode state by performing a unitary operation_on_spatial mode $a_2$,_resorting to an unbalanced beam splitter (i.e., UBS) with_reflection coefficient $R=\gamma/\delta$ (shown in Fig. \[figure3.1\]a)._The state of_the_photon_pair $AB$ is changed_from $|\phi_0\rangle_{AB}$ to $|\phi_1\rangle_{AB}$. Here $$\begin{aligned}
_ _ _ _ _ _ % Eq.3.2 _ % _Eq. 24
|\phi_1\rangle_{AB}\!&=&\!(\alpha|HH\rangle+\beta|VV\rangle)_{AB}
\otimes\Big[\gamma(|a_1b_1\rangle
+|a_2b_2\rangle)\nonumber\\
\!&&+\sqrt{|\delta|^2-|\gamma|^2}|a_3b_2\rangle\Big].\end{aligned}$$_If photon $A$_is not detected in the spatial_mode $a_3$, the_spatial-mode state of the photon pair_$AB$_is transformed into_a_maximally_entangled Bell_state.
Subsequently, Alice splits_the_parameter of_the_polarization state by performing the same_polarization_unitary operations on the spatial modes $a_1$_and $a_2$ as shown_in_Fig. \[figure3.1\]b. After two spatial_modes $a_1$ and $a_2$ pass_through PBSs (i.e., PBS$_1$ and PBS$_2$)_and $R_\theta$,_the state_of the photon pair $AB$ is transformed from $|\phi_1\rangle_{AB}$ to $|\phi_2\rangle_{AB}$._Here $$\begin{aligned}
_ _ __ __ _ _ _ % Eq.(4) %_ Eq. 25
|\ |
_N.ps "fig:"){width="34.50000%"}![\[fig:D10\]Examples of the $D$-dependence of the ground state chiral condensate for $m/g=0.125$, $x=10$ and five system sizes (left). The right plot shows a zoom into the region $D\in[80,160]$ for $N=84$. See comments in the text about the irregular approach to the $1/D=0$ limit. The red band represents the uncertainty related to the bond dimension, taken as explained in the text. ](cond_mg0.125_x10_N84.ps "fig:"){width="34.50000%"}
A typical example of such extrapolation is shown in Fig. \[fig:D200\] for $x=200$ and in Fig. \[fig:D10\] for $x=10$, at $m/g=0.125$. In both cases, we observe very good convergence towards the $1/D=0$ limit, with the above defined error from this step being of $\mathcal{O}(10^{-9})$ for the former and $\mathcal{O}(10^{-12})$ for the latter. This error is represented by a red band. Note that despite going to $D=160$, the convergence in bond dimension is so good that actually even with $D=40$ we would already obtain the result with an outstanding precision, of $\mathcal{O}(10^{-8})$ for $x=200$ (i.e. only an order of magnitude worse than with $D=160$) or even of $\mathcal{O}(10^{-12})$ for $x=10$ (i.e. the same as with $D=160$). The $x=10$ case illustrates that in some cases the convergence in $D$ is so good that our uncertainty comes from issues with the numerical precision. The MPS optimization procedure is considered to be converged when the relative change in the ground state energy in subsequent sweeps falls below a certain tolerance parameter, taken to be $\varepsilon_{\mathrm{tol}}=10^{-12}$ in our case. Notice, however, that this precision refers to the ground state energy, which typically converges better than other observables, so it will correspond to a somewhat worse precision in the chiral condensate, which we estimate to be in the $10^{-11}-10^{-10}$ region. In the $x=10$ case, the | _ N.ps " fig:"){width="34.50000%"}![\[fig: D10\]Examples of the $ D$-dependence of the ground state chiral condensate for $ m / g=0.125 $, $ x=10 $ and five system size (leave). The right plot shows a zoom into the area $ D\in[80,160]$ for $ N=84$. See comments in the text about the irregular access to the $ 1 / D=0 $ limit. The red set represents the uncertainty relate to the bond dimension, lead as explained in the text. ] (cond_mg0.125_x10_N84.ps " fig:"){width="34.50000% " }
A typical model of such extrapolation is shown in Fig. \[fig: D200\ ] for $ x=200 $ and in Fig. \[fig: D10\ ] for $ x=10 $, at $ m / g=0.125$. In both cases, we observe very adept convergence towards the $ 1 / D=0 $ limit, with the above defined error from this measure being of $ \mathcal{O}(10^{-9})$ for the former and $ \mathcal{O}(10^{-12})$ for the latter. This error is represented by a red band. notice that despite going to $ D=160 $, the convergence in bond dimension is so good that actually even with $ D=40 $ we would already receive the result with an outstanding precision, of $ \mathcal{O}(10^{-8})$ for $ x=200 $ (i.e. only an order of magnitude worse than with $ D=160 $) or even of $ \mathcal{O}(10^{-12})$ for $ x=10 $ (i.e. the same as with $ D=160 $). The $ x=10 $ case illustrates that in some subject the convergence in $ D$ is so good that our doubt hail from issues with the numerical preciseness. The MPS optimization procedure is regard to be converged when the relative variety in the ground state energy in subsequent sweeps falls below a certain tolerance parameter, taken to be $ \varepsilon_{\mathrm{tol}}=10^{-12}$ in our case. Notice, however, that this precision refers to the flat coat state department of energy, which typically converge better than other observables, so it will correspond to a somewhat bad precision in the chiral condensate, which we estimate to be in the $ 10^{-11}-10^{-10}$ area. In the $ x=10 $ case, the | _N.ps "fig:"){width="34.50000%"}![\[fig:D10\]Examples of the $D$-dependencg if the grouns state zhiral condensate for $m/g=0.125$, $x=10$ aid fuve ststem sizes (left). The rkght plot shows a zoon into the csgion $D\lu[80,160]$ for $K=84$. See romments in the text aboud the irregulas xp'roach to the $1/D=0$ limit. The red band rqpresenys the uncertainjy rekwted no the bond dimension, taken as sxplaintd in the text. ](conc_mg0.125_x10_N84.ps "fig:"){width="34.50000%"}
A typical fxamole of such extrapllation is whowg in Fig. \[fig:D200\] wor $x=200$ and pu Fig. \[fig:D10\] fkr $x=10$, at $m/g=0.125$. In both cases, we obsdrve rery good cincerhgnce towards the $1/Q=0$ limit, with the aboee defimed error from thms srep being of $\mathcal{O}(10^{-9})$ for the former anq $\mathcal{M}(10^{-12})$ yor the latter. This eerir is repsesevred by a rsd banf. Nkte that dsspite goint to $D=160$, the convergemcq in bond dimehsion ys so good that actually even with $D=40$ we wmuls already obtain the rewult with an outstandlng precifion, of $\mathcal{O}(10^{-8})$ for $x=200$ (i.e. only an order of magnitgde wkfse tman dutj $D=160$) or even of $\mathcal{O}(10^{-12})$ for $x=10$ (i.e. the same as rjtn $C=160$). The $x=10$ case ijlustrates yhwt yn some cases the convsrgence in $D$ is so good trat oyr uncertwintu comes from issues with thw numerical irecusion. The MPS optilization prucedore is considered to be convexged wgen the relwtive chahee in the ground ststa energy in subsequent sweqps falls belpw a ceftaim tolewance paraleter, taken to be $\varepsllon_{\mcthrm{dol}}=10^{-12}$ in our case. Notice, however, that this 'cecision refets do nhe grounb statc energy, which eypically convgrges betcer thxn other ogservabnes, so it wyll correspong to a somewhet worse [recusiob in thd chiral condenxate, whick we wstimate to be in bhe $10^{-11}-10^{-10}$ tefion. In the $x=10$ ccrt, rhe | _N.ps "fig:"){width="34.50000%"}![\[fig:D10\]Examples of the $D$-dependence of the chiral for $m/g=0.125$, and five system shows zoom into the $D\in[80,160]$ for $N=84$. comments in the text about the approach to the $1/D=0$ limit. The red band represents the uncertainty related to bond dimension, taken as explained in the text. ](cond_mg0.125_x10_N84.ps "fig:"){width="34.50000%"} A typical example such is in \[fig:D200\] for $x=200$ and in Fig. \[fig:D10\] for $x=10$, at $m/g=0.125$. In both cases, we observe good convergence towards the $1/D=0$ limit, with the defined error from this being of $\mathcal{O}(10^{-9})$ for the and for the This is by a red Note that despite going to $D=160$, the convergence in bond dimension is so good that actually even $D=40$ we obtain the with outstanding of $\mathcal{O}(10^{-8})$ for only an order of magnitude worse or even of $\mathcal{O}(10^{-12})$ for $x=10$ (i.e. the as with The $x=10$ case illustrates that in cases the convergence in $D$ is so good our uncertainty comes from issues with the numerical precision. The MPS optimization procedure is considered converged when the relative in the ground energy subsequent falls a certain parameter, taken to be $\varepsilon_{\mathrm{tol}}=10^{-12}$ in our case. Notice, however, that precision refers to the ground state energy, which typically converges other so it will to a somewhat worse in chiral condensate, which we be the the case, | _N.ps "fig:"){width="34.50000%"}![\[fig:D10\]Examples of The $D$-dependEnce oF thE grOuNd stAte cHiral condensatE For $m/G=0.125$, $x=10$ and five system sizes (leFt). The RiGHt plOT sHows a Zoom intO ThE REgiOn $d\iN[80,160]$ foR $N=84$. sEe CommeNts In the teXt about the IrrEgUlar approach TO tHe $1/D=0$ limit. ThE reD band represeNts The uncErTaiNTy relAteD to thE bond dIMensioN, taken as eXpLAined iN The text. ](CONd_Mg0.125_x10_N84.Ps "fig:"){width="34.50000%"}
A typicaL ExAMple of such extrApolatIoN Is SHOwn In FIg. \[fig:D200\] for $x=200$ AnD in FiG. \[Fig:D10\] for $X=10$, At $M/G=0.125$. iN boTH cases, we obserVe very good cONveRgence ToWarDS the $1/D=0$ lImit, wItH The Above defineD errOr from thiS step bEIng of $\maTHcal{O}(10^{-9})$ foR the foRmeR anD $\matHCaL{O}(10^{-12})$ For ThE LatTEr. thiS ErrOr is reprEsEnTed by A red BAND. note ThaT desPite gOing to $D=160$, the conVerGencE In bOnd diMensiOn is So Good tHat actUally EvEn with $D=40$ we would aLreaDy obtain tHe rEsUlt WiTh an oUTstandIng PreCision, oF $\mathcaL{o}(10^{-8})$ foR $x=200$ (I.E. ONlY an order of magnitudE wORSe Than with $d=160$) or eveN Of $\MaTHcal{O}(10^{-12})$ for $X=10$ (i.E. thE samE AS with $d=160$). The $X=10$ CaSe illustRates tHAt In Some casEs The conVeRgeNce In $D$ is SO gooD that oUr uncertAinty COmes from issues WIth the numericAL pRECiSIon. THe MpS optimizatIon pROcedUre iS CoNsiDEred tO be coNvERgED when the relative chaNgE in the GrounD state energy iN subsequenT SWEeps fallS belOW a CErtain tolerancE paraMeter, taken TO be $\varepSilon_{\Mathrm{toL}}=10^{-12}$ in our casE. nOtice, howEveR, thAt tHis PREcIsion refers to THE groUnD state eNerGy, which TypIcaLly ConVeRges betteR than othEr ObSeRvAblEs, so iT Will corrEsPonD tO a sOmewhAT worse PreciSion In ThE ChiRal condENsATE, whiCh We EstiMatE tO be in The $10^{-11}-10^{-10}$ rEGioN. In the $x=10$ Case, the | _N.ps "fig:"){width="34.50 000%"}![\[ fig:D 10\ ]Ex am ples ofthe $D$-depend e nceof the ground state ch iralco n dens a te for$m/g=0. 1 25 $ , $x =1 0$ an df iv e sys tem sizes(left). Th e r ig ht plot show s a zoom into th e region $D\ in[ 80,160 ]$ fo r $N=8 4$. Seecommen t s in t he text a bo u t thei rregula r ap proa ch to the $1/D=0$ li m it. The red ba nd rep re s en t s th e u ncertainty r elate d to the bo n d dim e nsion, takenas explaine d in the t ex t.] (cond_ mg0.1 25 _ x10 _N84.ps "fi g:") {width="3 4.5000 0 %"}
At ypicalexampl e o f s uche xt ra pol at i oni ssho w n i n Fig. \ [f ig :D200 \] f o r $ x=20 0$andin Fi g. \[fig:D10\ ] f or $ x =10 $, at $m/g =0.1 25 $. In bothcases ,we observe very goo d converg enc etow ar ds th e $1/D= 0$lim it, wit h the a b ove d e f i ne d error from thisst e p b eing of$\math c al {O } (10^{-9} )$ fo r th e forme r an d $ \mathcal {O}(10 ^ {- 12 })$ for t he lat te r.Thi s err o r is repre sented b y a r e d band. Note t h at despite go i ng t o$ D=16 0$, the conver genc e inbond di men s ion i s sogo o dt hat actually even w it h $D=4 0$ we would alread y obtain t h e result w itha no utstanding pre cisio n, of $\ma t hcal{O}( 10^{- 8})$ for $x=200$( i .e. only an or der of m ag nitude worset h an w it h $D=16 0$) or eve n o f $ \ma thc al {O}(10^{- 12})$ fo r$x =1 0$ (i .e. t h e same a swit h$D= 160$) . The $ x=10$ cas eil l ust rates t h at i n so me c ases th econve rgen c e i n $D$ i s so good th a t ou run certain ty comes from i ssues with t henumeri c a l precis ion. The MPS optimizati o n proce dur e iscons idered to be conve rge d whenthe re lativ echa n g e int h egro un d state en e r gyin su bs eque nt swee ps falls below a c e rta in tolerancepar amet e r ,tak e nt o b e$ \va r e psilon_{\mathrm {tol}}=10^ {- 1 2} $ in our c a se. N otice,however , tha t this p recisionrefers to t he g r o und state ene rgy, whi ch typica l ly co n ve rgesbet ter th an ot her o bserva b les , soit wil lcorres pondto a somew hat worse precision inthe ch iralcon densate,whi c h w e estimat e to be in the $1 0^{ -11}- 10^ { -10}$ reg i on . I n the$x=1 0 $ case, t h e | _N.ps "fig:"){width="34.50000%"}![\[fig:D10\]Examples_of the_$D$-dependence of the ground_state chiral_condensate_for $m/g=0.125$,_$x=10$_and five system_sizes (left). The_right plot shows a_zoom into the_region_$D\in[80,160]$ for $N=84$. See comments in the text about the irregular approach to the_$1/D=0$_limit. The_red_band_represents the uncertainty related to_the bond dimension, taken as_explained in_the text. ](cond_mg0.125_x10_N84.ps "fig:"){width="34.50000%"}
A typical example of such_extrapolation_is shown in_Fig. \[fig:D200\] for $x=200$ and in Fig. \[fig:D10\] for $x=10$, at_$m/g=0.125$. In both cases, we observe_very good convergence_towards_the_$1/D=0$ limit, with the_above defined error from this step_being of $\mathcal{O}(10^{-9})$ for the former_and $\mathcal{O}(10^{-12})$ for the latter. This error_is represented by a red band._Note that despite going to_$D=160$, the_convergence in bond dimension is_so good that_actually even_with $D=40$ we_would already obtain the result with_an outstanding precision,_of $\mathcal{O}(10^{-8})$ for $x=200$ (i.e. only_an_order of magnitude_worse_than_with $D=160$)_or even of_$\mathcal{O}(10^{-12})$_for $x=10$_(i.e._the same as with $D=160$). The_$x=10$_case illustrates that in some cases the_convergence in $D$ is_so_good that our uncertainty_comes from issues with the_numerical precision. The MPS optimization procedure_is considered_to be_converged when the relative change in the ground state energy in_subsequent sweeps falls below a certain_tolerance parameter, taken to_be $\varepsilon_{\mathrm{tol}}=10^{-12}$_in_our case. Notice,_however,_that this_precision refers to the ground state energy,_which typically_converges better than other observables, so_it will correspond to_a_somewhat worse precision in the chiral_condensate, which we estimate to be_in the $10^{-11}-10^{-10}$ region. In_the_$x=10$_case, the |
)}$, (d) $\delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $\delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$. The experimental parameters are: $\alpha{\left({0}\right)}=6\cdot2\pi KHz$, $\gamma= 50\cdot2\pi M(Hz)^2$ with $\mu{\left({0}\right)}=-1$. The varying values of $|\delta|=|d\mu/dt|$ are related to the quality of the inertial approximation, for slow change in $\mu$ the inertial approximation is satisfied, Cf. panels (c) and (d). Varying $\mu$ rapidly leads to the breakdown of the inertial theorem, see panels (a),(b),(e) and (f).[]{data-label="fig:1"}](graph_b.pdf "fig:"){width="23.50000%"}![The normalized energy as a function of time for the experimental result (blue), inertial solution (red) and numerical solution (dashed-black) for different values of $\delta$; (a) $\delta=- \alpha{\left({0}\right)}$, (b) $\delta=-0.05\cdot\alpha{\left({0}\right)}$, (c) $\delta=-0.01\cdot\alpha{\left({0}\right)}$, (d) $\delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $\delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$. The experimental parameters are: $\alpha{\left({0}\right)}=6\cdot2\pi KHz$, $\gamma= 50\cdot2\pi M(Hz)^2$ with $\mu{\left({0}\right)}=-1$. The varying values of $|\delta|=|d\mu/dt|$ are related to the quality of the inertial approximation, for slow change in $\mu$ the inertial approximation is satisfied, Cf. panels (c) and (d). Varying $\mu$ rapidly leads to the breakdown of the inertial theorem, see panels (a),(b),(e) and (f).[]{data-label="fig:1"}](graph_c.pdf "fig:"){width="23.5 | ) } $, (d) $ \delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $ \delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $ \delta=0.1\cdot\alpha{\left({0}\right)}$. The experimental parameters are: $ \alpha{\left({0}\right)}=6\cdot2\pi KHz$, $ \gamma= 50\cdot2\pi M(Hz)^2 $ with $ \mu{\left({0}\right)}=-1$. The varying values of $ |\delta|=|d\mu / dt|$ are relate to the timbre of the inertial approximation, for slow variety in $ \mu$ the inertial approximation is quenched, Cf. panels (c) and (d). vary $ \mu$ quickly leads to the breakdown of the inertial theorem, see panels (a),(b),(e) and (f).[]{data - label="fig:1"}](graph_b.pdf " fig:"){width="23.50000%"}![The normalized department of energy as a function of time for the experimental result (blue), inertial solution (crimson) and numerical solution (dashed - black) for different values of $ \delta$; (a) $ \delta=- \alpha{\left({0}\right)}$, (b) $ \delta=-0.05\cdot\alpha{\left({0}\right)}$, (coke) $ \delta=-0.01\cdot\alpha{\left({0}\right)}$, (d) $ \delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $ \delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $ \delta=0.1\cdot\alpha{\left({0}\right)}$. The experimental parameters are: $ \alpha{\left({0}\right)}=6\cdot2\pi KHz$, $ \gamma= 50\cdot2\pi M(Hz)^2 $ with $ \mu{\left({0}\right)}=-1$. The varying value of $ |\delta|=|d\mu / dt|$ are related to the quality of the inertial approximation, for slow variety in $ \mu$ the inertial approximation is quenched, Cf. panels (c) and (d). Varying $ \mu$ rapidly leave to the breakdown of the inertial theorem, see panels (a),(b),(e) and (f).[]{data - label="fig:1"}](graph_c.pdf " fig:"){width="23.5 | )}$, (d) $\felta=0.01\cdot\alpha{\left({0}\right)}$, (t) $\delta=0.05\cdot\alpha{\lgfr({0}\right)}$, (f) $\delfa=0.1\cdot\aloha{\left({0}\right)}$. The experimentap paramtners are: $\alpha{\left({0}\rigft)}=6\cdot2\pi NHz$, $\gammq= 50\cdir2\pi M(Hz)^2$ wivg $\mu{\lefb({0}\xight)}=-1$. Bhe vcrbing values of $|\celta|=|d\mu/dt|$ are related tm ghz quality of the inertial approximatyon, for spow change in $\iu$ tnq insgtlal approximation is satisfied, Cr. panelv (c) and (d). Varuing $\mu$ rapidly leads to tje bgeakdown of the infrtial theotsm, fwe panels (a),(b),(d) and (f).[]{data-label="fig:1"}](grakh_b.pdf "fig:"){width="23.50000%"}![The normalized endrgy cs a functiin of jime for the expegimental result (blue), hnertiak solution (red) anv nunerical solution (dashxd-black) for differenj values ox $\belta$; (a) $\delta=- \alpha{\ledt({0}\eight)}$, (b) $\dalta=-0.05\zeot\xlpga{\kert({0}\righh)}$, (c) $\delta=-0.01\cdot\zlpha{\left({0}\ritht)}$, (d) $\delta=0.01\cdot\alphs{\lqdt({0}\right)}$, (e) $\delfa=0.05\cdot\wl[ha{\left({0}\right)}$, (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$. Uhe esperimental parameters qre: $\alpha{\left({0}\right)}=6\cdoj2\pi KHz$, $\gaima= 50\cdot2\pi M(Hz)^2$ with $\mu{\left({0}\right)}=-1$. The varying valuev of $|\veuta|=|b\nu/dt|$ xee related to the quality of the inertial approvjmstpon, for slow channe in $\mu$ the inertoap skproximation ir satisfisd, Cf. panels (c) and (d). Varyyng $\my$ rapidly leacs to the breakdown of the unertial thejeem, see panels (a),(b),(e) and (f).[]{data-lcbel="fib:1"}](grapn_c.pdf "fig:"){width="23.5 | )}$, (d) $\delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $\delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$. parameters $\alpha{\left({0}\right)}=6\cdot2\pi KHz$, 50\cdot2\pi M(Hz)^2$ with $|\delta|=|d\mu/dt|$ related to the of the inertial for slow change in $\mu$ the approximation is satisfied, Cf. panels (c) and (d). Varying $\mu$ rapidly leads to breakdown of the inertial theorem, see panels (a),(b),(e) and (f).[]{data-label="fig:1"}](graph_b.pdf "fig:"){width="23.50000%"}![The normalized energy a of for experimental result (blue), inertial solution (red) and numerical solution (dashed-black) for different values of $\delta$; (a) \alpha{\left({0}\right)}$, (b) $\delta=-0.05\cdot\alpha{\left({0}\right)}$, (c) $\delta=-0.01\cdot\alpha{\left({0}\right)}$, (d) $\delta=0.01\cdot\alpha{\left({0}\right)}$, (e) (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$. The experimental are: $\alpha{\left({0}\right)}=6\cdot2\pi KHz$, $\gamma= 50\cdot2\pi with The varying of are to the quality the inertial approximation, for slow change in $\mu$ the inertial approximation is satisfied, Cf. panels (c) and Varying $\mu$ to the of inertial see panels (a),(b),(e) "fig:"){width="23.5 | )}$, (d) $\delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $\Delta=0.05\cdot\aLpha{\lEft({0}\RigHt)}$, (F) $\delTa=0.1\cdOt\alpha{\left({0}\rigHT)}$. The Experimental parameters Are: $\alPhA{\Left({0}\RIgHt)}=6\cdoT2\pi KHz$, $\gAMmA= 50\CDot2\Pi m(HZ)^2$ wiTh $\MU{\lEft({0}\riGht)}=-1$. the varyIng values oF $|\deLtA|=|d\mu/dt|$ are relATeD to the qualIty Of the inertiaL apProximAtIon, FOr sloW chAnge iN $\mu$ the INertiaL approximAtIOn is saTIsfied, CF. PAnEls (c) And (d). Varying $\mu$ rapiDLy LEads to the breakDown of ThE InERTiaL thEorem, see paNeLs (a),(b),(e) ANd (f).[]{data-LAbEL="FIg:1"}](gRAph_b.pdf "fig:"){widTh="23.50000%"}![The normalIZed Energy As A fuNCtion oF time FoR The ExperimentaL resUlt (blue), inErtial SOlution (REd) and nuMericaL soLutIon (dAShEd-BlaCk) FOr dIFfEreNT vaLues of $\deLtA$; (a) $\Delta=- \AlphA{\LEFT({0}\rigHt)}$, (b) $\DeltA=-0.05\cdot\Alpha{\left({0}\righT)}$, (c) $\dElta=-0.01\CDot\Alpha{\Left({0}\rIght)}$, (D) $\dElta=0.01\cDot\alpHa{\lefT({0}\rIght)}$, (e) $\delta=0.05\cdot\aLpha{\Left({0}\right)}$, (F) $\deLtA=0.1\cdOt\Alpha{\LEft({0}\rigHt)}$. THe eXperimeNtal parAMetErS ARE: $\aLpha{\left({0}\right)}=6\cdot2\pI KhZ$, $\GaMma= 50\cdot2\pI M(Hz)^2$ wiTH $\mU{\lEFt({0}\right)}=-1$. THe VarYing VALues oF $|\delTA|=|d\Mu/dt|$ are rElated TO tHe Quality Of The ineRtIal AppRoximATion, For sloW change iN $\mu$ thE Inertial approxIMation is satisFIeD, cF. pANels (C) anD (d). Varying $\mu$ RapiDLy leAds tO ThE brEAkdowN of thE iNErTIal theorem, see panels (A),(b),(E) and (f).[]{dAta-laBel="fig:1"}](graph_c.pDf "fig:"){width="23.5 | )}$, (d) $\delta=0.01\cdot \alpha{\le ft({0 }\r igh t) }$,(e)$\delta=0.05\c d ot\a lpha{\left({0}\right)} $, (f )$ \del t a= 0.1\c dot\alp h a{ \ l eft ({ 0} \ri gh t )} $. Th e e xperime ntal param ete rs are: $\alph a {\ left({0}\r igh t)}=6\cdot2\ piKHz$,$\ gam m a= 50 \cd ot2\p i M(Hz ) ^2$ wi th $\mu{\ le f t({0}\ r ight)}= - 1 $. The varying values o f $ | \delta|=|d\mu/ dt|$ a re re l a ted to the quali ty of t h e inert i al a p pro x imation, forslow change in$\mu$th e i n ertial appr ox i mat ion is sati sfie d, Cf. pa nels ( c ) and ( d ). Vary ing $\ mu$ ra pidl y l ea dsto the br eak d own of thein er tialtheo r e m , see pa nels (a), (b),(e) and ( f). []{d a ta- label ="fig :1"} ]( graph _b.pdf "fig :" ){width="23.500 00%" }![The no rma li zed e nergy as a f unc tio n of ti me fort heex p e r im ental result (blue ), i ne rtial so lution (r ed ) and num er ica l so l u tion(das h ed -black)for di f fe re nt valu es of $\ de lta $;(a) $ \ delt a=- \a lpha{\le ft({0 } \right)}$, (b) $\delta=-0.05 \ cd o t \a l pha{ \le ft({0}\righ t)}$ , (c) $\d e lt a=- 0 .01\c dot\a lp h a{ \ left({0}\right)}$,(d ) $\de lta=0 .01\cdot\alph a{\left({0 } \ r ight)}$, (e) $\ d elta=0.05\cdot \alph a{\left({0 } \right)} $, (f ) $\delt a=0.1\cdo t \ alpha{\l eft ({0 }\r igh t ) }$ . The experim e n talpa rameter s a re: $\a lph a{\ lef t({ 0} \right)}= 6\cdot2\ pi K Hz $, $\ gamma = 50\cdot 2\ piM( Hz) ^2$ w i th $\m u{\le ft({ 0} \r i ght )}=-1$. Th e vary in gvalu esof $|\d elta | =|d \mu/dt| $ are rel ate d toth equality of the inert ia l approxim at ion , fors l ow chang e in $\mu$ the inertial approxi mat ion i s sa tisfied,Cf. panel s ( c ) and(d). V aryin g$\m u $ rapi d l ylea ds to the br e a kdo wn of t he i nertial theorem, see pane l s ( a),(b),(e) an d ( f).[ ] { da ta- l ab e l=" fi g :1" } ] (graph_c.pdf "f ig:"){widt h= " 23 .5 | )}$, (d)_$\delta=0.01\cdot\alpha{\left({0}\right)}$, (e)_$\delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$. The_experimental parameters_are:_$\alpha{\left({0}\right)}=6\cdot2\pi KHz$,_$\gamma=_50\cdot2\pi M(Hz)^2$ with_$\mu{\left({0}\right)}=-1$. The varying_values of $|\delta|=|d\mu/dt|$ are_related to the_quality_of the inertial approximation, for slow change in $\mu$ the inertial approximation is satisfied,_Cf._panels (c)_and_(d)._Varying $\mu$ rapidly leads to_the breakdown of the inertial_theorem, see_panels (a),(b),(e) and (f).[]{data-label="fig:1"}](graph_b.pdf "fig:"){width="23.50000%"}![The normalized energy as_a_function of time_for the experimental result (blue), inertial solution (red) and_numerical solution (dashed-black) for different values_of $\delta$; (a)_$\delta=-_\alpha{\left({0}\right)}$,_(b) $\delta=-0.05\cdot\alpha{\left({0}\right)}$, (c) $\delta=-0.01\cdot\alpha{\left({0}\right)}$,_(d) $\delta=0.01\cdot\alpha{\left({0}\right)}$, (e) $\delta=0.05\cdot\alpha{\left({0}\right)}$, (f) $\delta=0.1\cdot\alpha{\left({0}\right)}$._The experimental parameters are: $\alpha{\left({0}\right)}=6\cdot2\pi KHz$,_$\gamma= 50\cdot2\pi M(Hz)^2$ with $\mu{\left({0}\right)}=-1$. The varying_values of $|\delta|=|d\mu/dt|$ are related to_the quality of the inertial_approximation, for_slow change in $\mu$ the_inertial approximation is_satisfied, Cf._panels (c) and_(d). Varying $\mu$ rapidly leads to_the breakdown of_the inertial theorem, see panels (a),(b),(e)_and_(f).[]{data-label="fig:1"}](graph_c.pdf "fig:"){width="23.5 |
ant {#ind}
==============================================
Let $f,g\in\mathbb{C}[x,y]$ be two polynomials which are algebraically independant. In this section, we will prove that the complex $f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is essentially concentrated in degree $0$. Finally we obtain a formula for the irregularity number at $c\in\mathbb{P}^1$ in terms of some geometric data associated with $f$ and $g$.
The complex $f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is essentially concentrated in degree zero
-------------------------------------------------------------------------------------------
The complex $f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is concentrated in degree zero except at a finite number of points.
- First of all, we recall the result of F. Maaref [@Ma] about the generic fibre of the sheaf of horizontal analytic sections of $\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))$.
\[theoma\] There exists a finite subset $\Sigma$ of $\mathbb{C}$ such that for all $c\in\mathbb{C}\setminus\Sigma$ and all $\rho\in\mathbb{C}$, such that $Re(-\rho)$ is big enough, $$i_c^+\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))\simeq H^k(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C}),$$ where $i_c$ is the inclusion of $\{c\}$ in $\mathbb{C}$.
- For all $c,\rho\in\mathbb{C}$, we have the long exact sequence of relative cohomology:
$$\xymatrix{0\ar[r]& H^0(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]&
H^0(f^{-1}(c),\mathbb{C})\ar[r]^-{\alpha}& H^0((f,g)^{-1}(c,\rho),\mathbb{C})\ar[dll] \\
&H^1(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]&
| ant { # ind }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Let $ f, g\in\mathbb{C}[x, y]$ be two polynomials which are algebraically independant. In this section, we will rise that the complex $ f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is basically concentrated in degree $ 0$. ultimately we prevail a formula for the irregularity issue at $ c\in\mathbb{P}^1 $ in term of some geometric datum consociate with $ f$ and $ g$.
The complex $ f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is essentially digest in degree zero
-------------------------------------------------------------------------------------------
The complex $ f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is concentrated in academic degree zero except at a finite number of points.
- First of all, we recall the result of F. Maaref [ @Ma ] about the generic fibre of the sheaf of horizontal analytic sections of $ \mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))$.
\[theoma\ ] There exist a finite subset $ \Sigma$ of $ \mathbb{C}$ such that for all $ c\in\mathbb{C}\setminus\Sigma$ and all $ \rho\in\mathbb{C}$, such that $ Re(-\rho)$ is big enough, $ $ i_c^+\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))\simeq H^k(f^{-1}(c),(f, g)^{-1}(c,\rho),\mathbb{C}),$$ where $ i_c$ is the inclusion of $ \{c\}$ in $ \mathbb{C}$.
- For all $ c,\rho\in\mathbb{C}$, we have the farseeing exact sequence of relative cohomology:
$ $ \xymatrix{0\ar[r ] & H^0(f^{-1}(c),(f, g)^{-1}(c,\rho),\mathbb{C})\ar[r ] &
H^0(f^{-1}(c),\mathbb{C})\ar[r]^-{\alpha } & H^0((f, g)^{-1}(c,\rho),\mathbb{C})\ar[dll ] \\
& H^1(f^{-1}(c),(f, g)^{-1}(c,\rho),\mathbb{C})\ar[r ] & | ant {#ind}
==============================================
Let $f,g\in\mathbb{C}[x,y]$ be two polynomials which are amgebraicxlly independant. In this secvion, we wull prove that the comolex $f_+(\matjcal{O}_{\matybb{C}^2}t^g)$ is essentially concentvcted jk degxex $0$. Finally we ontain a forkula for the isrdgblarity number at $c\in\mathbb{P}^1$ in termf of soke geometric datw asxjciafvd with $f$ and $g$.
The complex $f_+(\matgcal{O}_{\mauhbb{C}^2}e^g)$ is essentislly concentrated in degref zego
-------------------------------------------------------------------------------------------
The complex $f_+(\matjcal{O}_{\mathbb{X}^2}e^g)$ yw concentratdd in degrte zero excekt at a finite number of points.
- Firxt of all, qe refdll the resnlt of F. Maaref [@Ma] about tve genetic fibre of tme shxaf if horizontal analytir sections of $\mathcaj{H}^{k-1}(f_+(\mathcdl{K}_{\mathbb{C}^2}e^g))$.
\[theimq\] Thete exhsts q fkniue xugset $\Slgme$ of $\mathbb{D}$ such that for all $c\in\mathbb{C}\xeeninus\Sigma$ ans all $\whj\in\mathbb{C}$, such that $Re(-\rho)$ is big enougv, $$i_d^+\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathvb{C}^2}e^g))\simeq H^k(f^{-1}(c),(f,g)^{-1}(c,\rho),\lathbb{C}),$$ wrere $i_c$ is the inclusion of $\{c\}$ in $\mathbb{C}$.
- For aln $c,\rhk\kn\mcbmbb{C}$, qe have the long exact sequence of relative cohjjokony:
$$\xymatrix{0\ar[r]& H^0(f^{-1}(g),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]&
H^0(g^{-1}(c),\layrbb{C})\ar[r]^-{\alpha}& F^0((f,g)^{-1}(c,\rho),\mafhbb{C})\ar[dll] \\
&H^1(f^{-1}(c),(f,g)^{-1}(c,\rjo),\mathbf{C})\ar[r]&
| ant {#ind} ============================================== Let $f,g\in\mathbb{C}[x,y]$ be two are independant. In section, we will is concentrated in degree Finally we obtain formula for the irregularity number at in terms of some geometric data associated with $f$ and $g$. The complex is essentially concentrated in degree zero ------------------------------------------------------------------------------------------- The complex $f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is concentrated in zero at finite of points. - First of all, we recall the result of F. Maaref [@Ma] about the fibre of the sheaf of horizontal analytic sections $\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))$. \[theoma\] There exists finite subset $\Sigma$ of $\mathbb{C}$ that all $c\in\mathbb{C}\setminus\Sigma$ all such $Re(-\rho)$ is big $$i_c^+\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))\simeq H^k(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C}),$$ where $i_c$ is the inclusion of $\{c\}$ in $\mathbb{C}$. - For all $c,\rho\in\mathbb{C}$, we have long exact relative cohomology: H^0(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]& H^0((f,g)^{-1}(c,\rho),\mathbb{C})\ar[dll] &H^1(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]& | ant {#ind}
==============================================
Let $f,g\in\mathbb{C}[x,y]$ be tWo polynomiAls whIch Are AlGebrAicaLly independant. iN thiS section, we will prove thaT the cOmPLex $f_+(\MAtHcal{O}_{\Mathbb{C}^2}E^G)$ iS ESseNtIaLly CoNCeNtratEd iN degree $0$. finally we oBtaIn A formula for tHE iRregularitY nuMber at $c\in\matHbb{p}^1$ in terMs Of sOMe geoMetRic daTa assoCIated wIth $f$ and $g$.
THe COmplex $F_+(\Mathcal{o}_{\MAtHbb{C}^2}E^g)$ is essentially coNCeNTrated in degree Zero
-------------------------------------------------------------------------------------------
ThE cOMpLEX $f_+(\mAthCal{O}_{\mathbb{c}^2}e^G)$ is coNCentratED iN DEGreE Zero except at a Finite numbeR Of pOints.
- FIrSt oF All, we rEcall ThE ResUlt of F. MaareF [@Ma] aBout the geNeric fIBre of thE Sheaf of HorizoNtaL anAlytIC sEcTioNs OF $\maTHcAl{H}^{K-1}(F_+(\maThcal{O}_{\maThBb{c}^2}e^g))$.
\[thEoma\] tHERE exiSts A finIte suBset $\Sigma$ of $\maThbB{C}$ suCH thAt for All $c\iN\matHbB{C}\setMinus\SIgma$ aNd All $\rho\in\mathbb{C}$, Such That $Re(-\rho)$ Is bIg EnoUgH, $$i_c^+\maTHcal{H}^{k-1}(F_+(\maThcAl{O}_{\mathBb{C}^2}e^g))\siMEq H^K(f^{-1}(C),(F,G)^{-1}(C,\rHo),\mathbb{C}),$$ where $i_c$ is ThE INcLusion of $\{C\}$ in $\matHBb{c}$.
- FOR all $c,\rho\In\MatHbb{C}$, WE Have tHe loNG eXact sequEnce of RElAtIve cohoMoLogy:
$$\xyMaTriX{0\ar[R]& H^0(f^{-1}(c),(f,G)^{-1}(C,\rho),\Mathbb{c})\ar[r]&
H^0(f^{-1}(c),\mAthbb{c})\Ar[r]^-{\alpha}& H^0((f,g)^{-1}(c,\rhO),\Mathbb{C})\ar[dll] \\
&H^1(F^{-1}(C),(f,G)^{-1}(C,\RhO),\MathBb{C})\Ar[r]&
| ant {#ind}
=============== ========== ===== === === == ==== ====
Let $f,g\in\ m athb b{C}[x,y]$ be two poly nomia ls whic h a re al gebraic a ll y ind ep en dan t. In this se ction,we will pr ove t hat the comp l ex $f_+(\mat hca l{O}_{\mathb b{C }^2}e^ g) $ i s esse nti allyconcen t ratedin degree $ 0 $. Fin a lly weo b ta in a formula for thei rr e gularity numbe r at $ c\ i n\ m a thb b{P }^1$ in te rm s ofs ome geo m et r i c da t a associatedwith $f$ an d $g $.
Th ecom p lex $f _+(\m at h cal {O}_{\mathb b{C} ^2}e^g)$is ess e ntially concent ratedindeg reez er o--- -- - --- - -- --- - --- -------- -- -- ----- ---- - - - - ---- --- ---- ----- ------------- --- ---- - --- ----- --
T he c om plex$f_+(\ mathc al {O}_{\mathbb{C} ^2}e ^g)$ is c onc en tra te d ind egreezer o e xcept a t a fin i tenu m b e rof points.
- Fi rs t of all, we recal l t he result o fF.Maar e f [@Ma ] ab o ut the gen eric f i br eof thesh eaf of h ori zon tal a n alyt ic sec tions of $\ma t hcal{H}^{k-1}( f _+(\mathcal{O } _{ \ m at h bb{C }^2 }e^g))$.
\ [ theo ma\] Th ere exist s a f in i te subset $\Sigma$ of$\ mathbb {C}$such that for all $c\in \ m a thbb{C}\ setm i nu s \Sigma$ and al l $\r ho\in\math b b{C}$, s uch t hat $Re( -\rho)$ i s big enou gh, $$ i_c ^+\ m a th cal{H}^{k-1}( f _ +(\m at hcal{O} _{\ mathbb{ C}^ 2}e ^g) )\s im eq H^k(f^ {-1}(c), (f ,g )^ {- 1}( c,\rh o ),\mathb b{ C}) ,$ $ w here$ i_c$ i s the inc lu si o n o f $\{c\ } $i n $\m at hb b{C} $.
- Fo r al l $c ,\rho\i n\mathbb{ C}$ , weha ve the lo ng exact sequ en ce of rela ti vecohomo l o gy:
$$\ xymatrix{0\ar[r]& H^0(f ^ {-1}(c) ,(f ,g)^{ -1}( c,\rho),\ mat hbb{C} )\a r [r]&
H ^0(f^{ -1}(c ), \ma t h bb{C} ) \ ar [r] ^- {\alpha}&H ^ 0(( f,g)^ {- 1}(c ,\rho), \mathbb{C})\ar[dll ] \\
&H^1(f^{-1}( c), (f,g ) ^ {- 1}( c ,\ r ho) ,\ m ath b b {C})\ar[r]&
| ant {#ind}
==============================================
Let_$f,g\in\mathbb{C}[x,y]$ be_two polynomials which are_algebraically independant._In_this section,_we_will prove that_the complex $f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$_is essentially concentrated in_degree $0$. Finally_we_obtain a formula for the irregularity number at $c\in\mathbb{P}^1$ in terms of some geometric_data_associated with_$f$_and_$g$.
The complex $f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is essentially_concentrated in degree zero
-------------------------------------------------------------------------------------------
The complex_$f_+(\mathcal{O}_{\mathbb{C}^2}e^g)$ is_concentrated in degree zero except at a finite_number_of points.
- _ First of all, we recall the result of_F. Maaref [@Ma] about the generic_fibre of the_sheaf_of_horizontal analytic sections of_$\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))$.
\[theoma\] There_exists a finite subset $\Sigma$ of_$\mathbb{C}$ such that for all $c\in\mathbb{C}\setminus\Sigma$ and_all $\rho\in\mathbb{C}$, such that $Re(-\rho)$ is_big enough, $$i_c^+\mathcal{H}^{k-1}(f_+(\mathcal{O}_{\mathbb{C}^2}e^g))\simeq H^k(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C}),$$ where_$i_c$ is_the inclusion of $\{c\}$ in_$\mathbb{C}$.
- _For all_$c,\rho\in\mathbb{C}$, we have_the long exact sequence of relative_cohomology:
$$\xymatrix{0\ar[r]& H^0(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]&
H^0(f^{-1}(c),\mathbb{C})\ar[r]^-{\alpha}& H^0((f,g)^{-1}(c,\rho),\mathbb{C})\ar[dll]_\\
&H^1(f^{-1}(c),(f,g)^{-1}(c,\rho),\mathbb{C})\ar[r]&
|
).
A. H. Safavi-Naeini, T. P.Mayer Alegre, J.Chan, M. Eichenfield, M. Winger, Q. Lin, J. T. Hill, D. E. Chang, and O. Painter, Electromagnetically induced transparency and slow light with optomechanics, Nature (London) **472**, 69 (2011).
L. Tian, Robust photon entanglement via quantum interference in optomechanical interfaces, Phys. Rev. Lett. **110**, 233602 (2013).
Y. D. Wang and A. A. Clerk, Reservoir-engineered entanglement in optomechanical systems, Phys. Rev. Lett. **110**, 253601 (2013).
I. Marinković, A. Wallucks, R. Riedinger, S. Hong, M. Aspelmeyer, and S. Gröblacher, Optomechanical Bell test, Phys. Rev. Lett. **121**, 220404 (2018).
J. Margueritat, A. V. Carlotta, S. Monnier, H. D. Ayari, H. C. Mertani, A. Berthelot, Q. Martinet, X. Dagany, C. Rivière, J. P. Rieu, and T. Dehoux, High-frequency mechanical properties of tumors measured by brillouin light scattering, Phys. Rev. Lett. **122**, 018101 (2019).
L. J. Aplet and J. W. Carson, A Faraday effect optical isolator, Appl. Opt. **3**, 544 (1964).
H. Lira, Z. Yu, S. Fan, and M. Lipson, Electrically driven nonreciprocity induced by interband photonic transition on a silicon chip, Phys. Rev. Lett. **109**, 033901 (2012).
K. Fang, Z. Yu, and S. Fan, Photonic aharonov-bohm effect based on dynamic modulation, Phys. Rev. Lett. **108**, 153901 (2012).
R. Fleury, D. L. Sounas, C. F. Sieck, M. R. Haberman, and A. Alù, Sound isolation and giant linear nonreciprocity in a compact acoustic circulator, Science **343**, 516 (2014). | ).
A. H. Safavi - Naeini, T. P.Mayer Alegre, J.Chan, M. Eichenfield, M. Winger, Q. Lin, J. T. Hill, D. E. Chang, and O. Painter, Electromagnetically induced transparency and slow light with optomechanics, Nature (London) * * 472 * *, 69 (2011).
L. Tian, full-bodied photon web via quantum hindrance in optomechanical interfaces, Phys. Rev. Lett. * * 110 * *, 233602 (2013).
Y. D. Wang and A. A. Clerk, Reservoir - engineer web in optomechanical systems, Phys. Rev. Lett. * * 110 * *, 253601 (2013).
I. Marinković, A. Wallucks, R. Riedinger, S. Hong, M. Aspelmeyer, and S. Gröblacher, Optomechanical Bell test, Phys. Rev. Lett. * * 121 * *, 220404 (2018).
J. Margueritat, A. V. Carlotta, S. Monnier, H. D. Ayari, H. C. Mertani, A. Berthelot, Q. Martinet, X. Dagany, C. Rivière, J. P. Rieu, and T. Dehoux, High - frequency mechanical property of tumors measured by brillouin luminosity scattering, Phys. Rev. Lett. * * 122 * *, 018101 (2019).
L. J. Aplet and J. W. Carson, A Faraday effect ocular isolator, Appl. Opt. * * 3 * *, 544 (1964).
H. Lira, Z. Yu, S. Fan, and M. Lipson, Electrically driven nonreciprocity induced by interband photonic conversion on a silicon bit, Phys. Rev. Lett. * * 109 * *, 033901 (2012).
K. Fang, Z. Yu, and S. Fan, Photonic aharonov - bohm effect based on dynamic modulation, Phys. Rev. Lett. * * 108 * *, 153901 (2012).
R. Fleury, D. L. Sounas, C. F. Sieck, M. R. Haberman, and A. Alù, Sound isolation and elephantine linear nonreciprocity in a compact acoustic circulator, Science * * 343 * *, 516 (2014). | ).
A. H. Safavi-Naeini, T. P.Mayer Auegre, J.Chan, M. Enxhenfixld, M. Wjnger, Q. Uin, J. T. Hill, D. E. Chang, and O. Pqintee, Electromagnetically knduced tgansparenxy aid slow light wivg optomcehanidd, Nacuce (London) **472**, 69 (2011).
L. Tlan, Robust [hoton entanglamdnc via quantum interference in optomeshanicak lnterfaces, Phyf. Ren. Jett. **110**, 233602 (2013).
Y. D. Wang and A. A. Clerk, Reservoid-enginetred entanglement on optomechanical systems, Ohys. Rev. Lett. **110**, 253601 (2013).
I. Marijković, A. Walouckf, R. Riedinger, S. Hong, M. Aspelmeyer, ahd S. Gröblacher, Optomechanical Bdll tzst, Phys. Rec. Oeth. **121**, 220404 (2018).
J. Marguecitat, W. V. Carlotta, S. Monniar, H. D. Syari, H. C. Mertsni, A. Verthelot, Q. Martinet, E. Dagany, C. Rivière, J. K. Rieu, and T. Dehoux, High-frequwnxy mewhanhcal prooeruiex kf tumlrs measured gy brillouib light scattering, Khyf. Rev. Lett. **122**, 018101 (2019).
L. J. Aplqt and J. W. Carson, A Faraday effect optican iaolator, Appl. Opt. **3**, 544 (1964).
H. Liea, Z. Yu, S. Fan, and M. Llpson, Elestrically driven nonreciprocity induced by interbdnd piogonng geajsition on a silicon chip, Phys. Rev. Lett. **109**, 033901 (2012).
K. Fwhg, Z. Yu, and S. Fan, Ihotonic aharonov-bphl rsfect based ov dynamic modulation, Phys. Rfv. Lett. **108**, 153901 (2012).
R. Dleury, D. J. Soinas, C. F. Sieck, M. R. Haberman, and A. Alù, Solnd usolation and gianc linear nonxeciprpcity in a compact acoustic eirculztor, Sciencf **343**, 516 (2014). | ). A. H. Safavi-Naeini, T. P.Mayer Alegre, Eichenfield, Winger, Q. J. T. Hill, Painter, induced transparency and light with optomechanics, (London) **472**, 69 (2011). L. Tian, photon entanglement via quantum interference in optomechanical interfaces, Phys. Rev. Lett. **110**, 233602 Y. D. Wang and A. A. Clerk, Reservoir-engineered entanglement in optomechanical systems, Phys. Lett. 253601 I. A. Wallucks, R. Riedinger, S. Hong, M. Aspelmeyer, and S. Gröblacher, Optomechanical Bell test, Phys. Rev. **121**, 220404 (2018). J. Margueritat, A. V. Carlotta, Monnier, H. D. Ayari, C. Mertani, A. Berthelot, Q. X. C. Rivière, P. and Dehoux, High-frequency mechanical of tumors measured by brillouin light scattering, Phys. Rev. Lett. **122**, 018101 (2019). L. J. Aplet and W. Carson, effect optical Appl. **3**, (1964). H. Lira, S. Fan, and M. Lipson, Electrically by interband photonic transition on a silicon chip, Rev. Lett. 033901 (2012). K. Fang, Z. Yu, S. Fan, Photonic aharonov-bohm effect based on dynamic Phys. Rev. Lett. **108**, 153901 (2012). R. Fleury, D. L. Sounas, C. F. Sieck, M. and A. Alù, Sound and giant linear in compact circulator, **343**, 516 | ).
A. H. Safavi-Naeini, T. P.Mayer AlegRe, J.Chan, M. EiChenfIelD, M. WInGer, Q. lin, J. t. Hill, D. E. Chang, anD o. PaiNter, ElectromagneticallY induCeD TranSPaRency And slow LIgHT WitH oPtOmeChANiCs, NatUre (london) **472**, 69 (2011).
L. tian, Robust PhoToN entanglemenT ViA quantum inTerFerence in optOmeChanicAl IntERfaceS, PhYs. Rev. lett. **110**, 233602 (2013).
Y. D. wAng and a. A. Clerk, ReSeRVoir-enGIneered ENTaNgleMent in optomechaniCAl SYstems, Phys. Rev. LEtt. **110**, 253601 (2013).
I. MaRiNKoVIĆ, A. WAllUcks, R. RiediNgEr, S. HoNG, M. AspelMEyER, ANd S. gRöblacher, OptoMechanical BELl tEst, PhyS. REv. LETt. **121**, 220404 (2018).
J. MarGueriTaT, a. V. CArlotta, S. MonNier, h. D. Ayari, H. C. mertanI, a. BertheLOt, Q. MartInet, X. DAgaNy, C. riviÈRe, j. P. rieU, aND T. DEHoUx, HIGh-fRequency MeChAnicaL proPERTIes oF tuMors MeasuRed by brillouiN liGht sCAttEring, phys. REv. LeTt. **122**, 018101 (2019).
l. J. AplEt and J. w. CarsOn, a Faraday effect oPticAl isolatoR, ApPl. opt. **3**, 544 (1964).
h. LIra, Z. YU, s. Fan, anD M. LIpsOn, ElectRically DRivEn NONReCiprocity induced by InTERbAnd photoNic traNSiTiON on a siliCoN chIp, PhYS. rev. LeTt. **109**, 033901 (2012).
K. FANg, z. Yu, and S. FAn, PhotONiC aHaronov-BoHm effeCt BasEd oN dynaMIc moDulatiOn, Phys. ReV. Lett. **108**, 153901 (2012).
r. fleury, D. L. Sounas, c. f. Sieck, M. R. HaberMAn, AND A. aLù, SoUnd Isolation anD giaNT linEar nONrEciPRocitY in a cOmPAcT Acoustic circulator, SCiEnce **343**, 516 (2014). | ).
A. H. Safavi-Naeini, T . P.MayerAlegr e,J.C ha n, M . Ei chenfield, M.W inge r, Q. Lin, J. T. Hill, D. E .C hang , a nd O. Painte r ,E l ect ro ma gne ti c al ly in duc ed tran sparency a ndsl ow light wit h o ptomechani cs, Nature (Lon don ) **47 2* *,6 9 (20 11) .
L. Tian, Robust photon e nt a ngleme n t via q u a nt um i nterference in op t om e chanical inter faces, P h ys . Rev . L ett. **110 ** , 233 6 02 (201 3 ). Y . D . Wang and A.A. Clerk, R e ser voir-e ng ine e red en tangl em e ntin optomech anic al system s, Phy s . Rev.L ett. ** 110**, 25 360 1 (2 0 13 ).
I .M ari n ko vić , A. Walluck s, R . Rie ding e r , S. H ong , M. Aspe lmeyer, and S . G röbl a che r, Op tomec hani ca l Bel l test , Phy s. Rev. Lett. **1 21** , 220404(20 18 ).
J . Mar g uerita t,A.V. Carl otta, S . Mo nn i e r ,H. D. Ayari, H. C. M e r ta ni, A. B erthel o t, Q . Martine t, X. Dag a n y, C. Riv i èr e, J. P. Rieu, an dT. Deho ux , High -f req uen cy me c hani cal pr operties of t u mors measuredb y brillouin l i gh t sc a tter ing , Phys. Rev . Le t t. * *122 * *, 01 8 101 ( 2019) .L. J. Aplet and J. W.Ca rson,A Far aday effect o ptical iso l a t or, Appl . Op t .* *3**, 544 (196 4).
H. Lira, Z . Yu, S.Fan,and M. L ipson, El e c trically dr ive n n onr e c ip rocity induce d by i nt erbandpho tonic t ran sit ion on a siliconchip, Ph ys .Re v. Le tt. * * 109**, 0 33 901 ( 201 2).
K . Fang , Z.Yu,an dS . F an, Pho t on i c aha ro no v-bo hmef fectbase d on dynami c modulat ion , Phy s. R ev. Let t. **108**, 1 53 901 (2012) .
R. Fleur y , D. L. S ounas, C. F. Sieck, M.R . Haber man , and A.Alù, Soun d i solati ona nd gia nt lin ear n on rec i p rocit y in aco mpact acou s t iccircu la tor, Scienc e **343**, 516 (20 1 4). | ).
A. H._Safavi-Naeini, T._P.Mayer Alegre, J.Chan, M._Eichenfield, M._Winger,_Q. Lin,_J._T. Hill, D._E. Chang, and_O. Painter, Electromagnetically induced_transparency and slow_light_with optomechanics, Nature (London) **472**, 69 (2011).
L. Tian, Robust photon entanglement via quantum interference_in_optomechanical interfaces,_Phys._Rev._Lett. **110**, 233602 (2013).
Y. D._Wang and A. A. Clerk,_Reservoir-engineered entanglement_in optomechanical systems, Phys. Rev. Lett. **110**, 253601_(2013).
I._Marinković, A. Wallucks,_R. Riedinger, S. Hong, M. Aspelmeyer, and S. Gröblacher,_Optomechanical Bell test, Phys. Rev. Lett._**121**, 220404 (2018).
J._Margueritat,_A._V. Carlotta, S. Monnier,_H. D. Ayari, H. C. Mertani,_A. Berthelot, Q. Martinet, X. Dagany,_C. Rivière, J. P. Rieu, and T._Dehoux, High-frequency mechanical properties of tumors_measured by brillouin light scattering,_Phys. Rev._Lett. **122**, 018101 (2019).
L. J._Aplet and J._W. Carson,_A Faraday effect_optical isolator, Appl. Opt. **3**, 544_(1964).
H. Lira, Z._Yu, S. Fan, and M. Lipson,_Electrically_driven nonreciprocity induced_by_interband_photonic transition_on a silicon_chip,_Phys. Rev._Lett._**109**, 033901 (2012).
K. Fang, Z. Yu,_and_S. Fan, Photonic aharonov-bohm effect based on_dynamic modulation, Phys. Rev._Lett._**108**, 153901 (2012).
R. Fleury,_D. L. Sounas, C. F._Sieck, M. R. Haberman, and A._Alù, Sound_isolation and_giant linear nonreciprocity in a compact acoustic circulator, Science **343**, 516_(2014). |
=0} =\left. \varphi'\frac{d}{dr}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0} = 0,$$ and so $$\begin{aligned}
\sum_{i=1}^{n-1}\frac{d^2}{dr^2}\left(\varphi\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)\right)
\right|_{r=0} &= \sum_{i=1}^{n-1}\left.\left(\varphi'\frac{d^2}{dr^2}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0} +\varphi''\left(\frac{d}{dr}\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0}\right)^2\right)\\
&\leq -2(n-1)\left.\varphi'{\mathbf{T}_\kappa}\right|_{d/2}.\end{aligned}$$ Now using the inequalities and, we have from that $$\label{trace inequality}
\mathcal{L}[\nabla^2\psi,\nabla\psi]\leq{\mathrm{trace}}\left(A\nabla^2\psi\right)\leq 2\left[\alpha(\varphi')\varphi''-(n-1)\beta(\varphi')\varphi'{\mathbf{T}_\kappa}\right]\big|_{d/2},$$ as required.
In the case $d=\frac{\pi}{\sqrt{K}}$ then we choose instead $\gamma_i(r,s) = \exp_{\gamma(s)}
\left(\frac{r \mathbf{C_{\kappa'}} (s)E_i}{\mathbf{C_{\kappa'}}(d/2)}\right)$, for arbitrary $\kappa'<\kappa$. Then the computation above gives $$\sum_{i=1}^{n-1}\nabla^2\psi((E_i,E_i),(E_i,E_i))\leq -2(n-1)\varphi'{\mathbf{T}_\kappa}.$$ Since the right hand side approaches $-\infty$ as $\kappa'$ increases to $\kappa$, we have a contradiction to the assumption that $\psi$ is smooth. Hence no such $\psi$ exists and there is nothing to prove | = 0 } = \left. \varphi'\frac{d}{dr}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0 } = 0,$$ and so $ $ \begin{aligned }
\sum_{i=1}^{n-1}\frac{d^2}{dr^2}\left(\varphi\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)\right)
\right|_{r=0 } & = \sum_{i=1}^{n-1}\left.\left(\varphi'\frac{d^2}{dr^2}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0 } + \varphi''\left(\frac{d}{dr}\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0}\right)^2\right)\\
& \leq -2(n-1)\left.\varphi'{\mathbf{T}_\kappa}\right|_{d/2}.\end{aligned}$$ Now using the inequalities and, we get from that $ $ \label{trace inequality }
\mathcal{L}[\nabla^2\psi,\nabla\psi]\leq{\mathrm{trace}}\left(A\nabla^2\psi\right)\leq 2\left[\alpha(\varphi')\varphi''-(n-1)\beta(\varphi')\varphi'{\mathbf{T}_\kappa}\right]\big|_{d/2},$$ as command.
In the case $ d=\frac{\pi}{\sqrt{K}}$ then we choose instead $ \gamma_i(r, s) = \exp_{\gamma(s) }
\left(\frac{r \mathbf{C_{\kappa' } } (s)E_i}{\mathbf{C_{\kappa'}}(d/2)}\right)$, for arbitrary $ \kappa'<\kappa$. Then the computation above yield $ $ \sum_{i=1}^{n-1}\nabla^2\psi((E_i, E_i),(E_i, E_i))\leq -2(n-1)\varphi'{\mathbf{T}_\kappa}.$$ Since the right hand english approach $ -\infty$ as $ \kappa'$ increases to $ \kappa$, we suffer a contradiction to the assumption that $ \psi$ is smooth. therefore no such $ \psi$ exists and there is nothing to prove | =0} =\levt. \varphi'\frac{d}{dr}\left(\frac{U[\gamma_{i}(r,.)]}{2}\right)
\riyyt|_{r=0} = 0,$$ end so $$\gegin{aliened}
\sum_{i=1}^{n-1}\frac{d^2}{dr^2}\left(\varphi\leht.\ledt(\frax{L[\gamma_{i}(r,.)]}{2}\right)\right)
\rigft|_{r=0} &= \sum_{i=1}^{j-1}\left.\lefr(\varkhi'\frac{d^2}{dr^2}\left(\frar{M[\gamma_{i}(v,.)]}{2}\xight)
\dlght|_{r=0} +\tarphi''\left(\frac{d}{cr}\left.\left(\xrac{L[\gamma_{i}(r,.)]}{2}\richg)
\rnght|_{r=0}\right)^2\right)\\
&\leq -2(n-1)\left.\varphi'{\mathbf{E}_\kappa}\rogjt|_{d/2}.\end{aligned}$$ Gow lsyng fhe inequalities and, we have from fhat $$\lauel{trace inequakity}
\mathcal{L}[\nabla^2\psi,\nabla\pdi]\lee{\mathrm{trace}}\left(A\nwbla^2\psi\righj)\meq 2\oeft[\alpha(\varohi')\varphi''-(n-1)\beta(\varphi')\vatphi'{\mathbf{T}_\kappa}\right]\big|_{d/2},$$ as reqjired.
Nn the case $d=\draf{\ki}{\sqrt{K}}$ then we croose instead $\gamma_i(s,s) = \exp_{\bamma(s)}
\left(\frac{v \matibf{C_{\jappa'}} (s)E_i}{\mathbf{C_{\kappa'}}(v/2)}\right)$, for arbitrary $\kappa'<\kap[a$. Then the computariin abmve civer $$\suo_{i=1}^{n-1}\haula^2\lsi((E_i,E_l),(E_i,X_i))\leq -2(n-1)\varpgi'{\mathbf{T}_\kappa}.$$ Since the right hwbd side approzches $-\ynsty$ as $\kappa'$ increases to $\kappa$, we have a dontradiction to the aswumption that $\psi$ is dmooth. Hegce no such $\psi$ exists and there is nothing to prmve | =0} =\left. \varphi'\frac{d}{dr}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right) \right|_{r=0} = 0,$$ and \sum_{i=1}^{n-1}\frac{d^2}{dr^2}\left(\varphi\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)\right) &= \sum_{i=1}^{n-1}\left.\left(\varphi'\frac{d^2}{dr^2}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right) +\varphi''\left(\frac{d}{dr}\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right) \right|_{r=0}\right)^2\right)\\ &\leq and, have from that inequality} \mathcal{L}[\nabla^2\psi,\nabla\psi]\leq{\mathrm{trace}}\left(A\nabla^2\psi\right)\leq 2\left[\alpha(\varphi')\varphi''-(n-1)\beta(\varphi')\varphi'{\mathbf{T}_\kappa}\right]\big|_{d/2},$$ required. In the case $d=\frac{\pi}{\sqrt{K}}$ then choose instead $\gamma_i(r,s) = \exp_{\gamma(s)} \left(\frac{r \mathbf{C_{\kappa'}} (s)E_i}{\mathbf{C_{\kappa'}}(d/2)}\right)$, for arbitrary $\kappa'<\kappa$. Then the above gives $$\sum_{i=1}^{n-1}\nabla^2\psi((E_i,E_i),(E_i,E_i))\leq -2(n-1)\varphi'{\mathbf{T}_\kappa}.$$ Since the right hand side approaches $-\infty$ as $\kappa'$ to we a to the assumption that $\psi$ is smooth. Hence no such $\psi$ exists and there is nothing prove | =0} =\left. \varphi'\frac{d}{dr}\left(\frac{l[\gamma_{i}(r,.)]}{2}\riGht)
\riGht|_{R=0} = 0,$$ anD sO $$\begIn{alIgned}
\sum_{i=1}^{n-1}\frac{D^2}{Dr^2}\leFt(\varphi\left.\left(\frac{L[\gAmma_{i}(R,.)]}{2}\rIGht)\rIGhT)
\righT|_{r=0} &= \sum_{i=1}^{n-1}\LEfT.\LEft(\VaRpHi'\fRaC{D^2}{dR^2}\left(\FraC{L[\gamma_{I}(r,.)]}{2}\right)
\rigHt|_{r=0} +\VaRphi''\left(\frac{D}{Dr}\Left.\left(\frAc{L[\Gamma_{i}(r,.)]}{2}\right)
\RigHt|_{r=0}\rigHt)^2\RigHT)\\
&\leq -2(n-1)\LefT.\varpHi'{\mathBF{T}_\kappA}\right|_{d/2}.\enD{aLIgned}$$ NOW using tHE InEquaLities and, we have frOM tHAt $$\label{trace inEqualiTy}
\MAtHCAl{L}[\NabLa^2\psi,\nabla\PsI]\leq{\mAThrm{traCE}}\lEFT(a\naBLa^2\psi\right)\leq 2\Left[\alpha(\vaRPhi')\Varphi''-(N-1)\bEta(\VArphi')\vArphi'{\MaTHbf{t}_\kappa}\right]\Big|_{d/2},$$ As requireD.
In the CAse $d=\fraC{\Pi}{\sqrt{K}}$ Then we ChoOse InstEAd $\GaMma_I(r,S) = \Exp_{\GAmMa(s)}
\LEft(\Frac{r \matHbF{C_{\Kappa'}} (S)E_i}{\mATHBF{C_{\kaPpa'}}(D/2)}\rigHt)$, for Arbitrary $\kappA'<\kaPpa$. THEn tHe comPutatIon aBoVe givEs $$\sum_{i=1}^{N-1}\nablA^2\pSi((E_i,E_i),(E_i,E_i))\leq -2(n-1)\vArphI'{\mathbf{T}_\kAppA}.$$ SIncE tHe rigHT hand sIde AppRoaches $-\Infty$ as $\KAppA'$ iNCREaSes to $\kappa$, we have a cOnTRAdIction to The assUMpTiON that $\psi$ Is SmoOth. HENCe no sUch $\pSI$ eXists and There iS NoThIng to prOvE | =0} =\left. \varphi'\frac{ d}{dr}\lef t(\fr ac{ L[\ ga mma_ {i}( r,.)]}{2}\righ t )
\r ight|_{r=0} = 0,$$ and so $ $\ b egin { al igned }
\sum_ { i= 1 } ^{n -1 }\ fra c{ d ^2 }{dr^ 2}\ left(\v arphi\left .\l ef t(\frac{L[\g a mm a_{i}(r,.) ]}{ 2}\right)\ri ght )
\rig ht |_{ r =0} & = \ sum_{ i=1}^{ n -1}\le ft.\left( \v a rphi'\ f rac{d^2 } { dr ^2}\ left(\frac{L[\gam m a_ { i}(r,.)]}{2}\r ight)\r i gh t | _{r =0} +\varphi' '\ left( \ frac{d} { dr } \ l eft . \left(\frac{L [\gamma_{i} ( r,. )]}{2} \r igh t )
\rig ht|_{ r= 0 }\r ight)^2\rig ht)\ \
&\leq - 2(n-1) \ left.\v a rphi'{\ mathbf {T} _\k appa } \r ig ht| _{ d /2} . \e nd{ a lig ned}$$ N ow u singthei n e q uali tie s an d, we have from th at$$\l a bel {trac e ine qual it y}
\m athcal {L}[\ na bla^2\psi,\nabl a\ps i]\leq{\m ath rm {tr ac e}}\l e ft(A\n abl a^2 \psi\ri ght)\le q 2\ le f t [ \a lpha(\varphi')\var ph i ' '- (n-1)\be ta(\va r ph i' ) \varphi' {\ mat hbf{ T } _\kap pa}\ r ig ht]\big| _{d/2} , $$ a s requi re d.
In t hecas e $d= \ frac {\pi}{ \sqrt{K} }$ th e n we choose in s tead $\gamma_ i (r , s )= \ex p_{ \gamma(s)}\lef t (\fr ac{r \m ath b f{C_{ \kapp a' } }( s)E_i}{\mathbf{C_{\ ka ppa'}} (d/2) }\right)$, fo r arbitrar y $ \kappa'< \kap p a$ . Then the comp utati on above g i ves $$\s um_{i =1}^{n-1 }\nabla^2 \ p si((E_i, E_i ),( E_i ,E_ i ) )\ leq -2(n-1)\v a r phi' {\ mathbf{ T}_ \kappa} .$$ Si nce th eright han d side a pp ro ac he s $ -\inf t y$ as $\ ka ppa '$ in creas e s to $ \kapp a$,we h a vea contr a di c t ionto t he a ssu mp tionthat $\p si$ issmooth. H enc e nosu ch $\psi$ exists and t he re is noth in g t o prov e | =0} =\left._\varphi'\frac{d}{dr}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0} =_0,$$ and so $$\begin{aligned}
\sum_{i=1}^{n-1}\frac{d^2}{dr^2}\left(\varphi\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)\right)
\right|_{r=0}_&= \sum_{i=1}^{n-1}\left.\left(\varphi'\frac{d^2}{dr^2}\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0}_+\varphi''\left(\frac{d}{dr}\left.\left(\frac{L[\gamma_{i}(r,.)]}{2}\right)
\right|_{r=0}\right)^2\right)\\
&\leq_-2(n-1)\left.\varphi'{\mathbf{T}_\kappa}\right|_{d/2}.\end{aligned}$$ Now_using_the inequalities and,_we have from_that $$\label{trace inequality}
\mathcal{L}[\nabla^2\psi,\nabla\psi]\leq{\mathrm{trace}}\left(A\nabla^2\psi\right)\leq 2\left[\alpha(\varphi')\varphi''-(n-1)\beta(\varphi')\varphi'{\mathbf{T}_\kappa}\right]\big|_{d/2},$$_as required.
In the_case_$d=\frac{\pi}{\sqrt{K}}$ then we choose instead $\gamma_i(r,s) = \exp_{\gamma(s)}
\left(\frac{r \mathbf{C_{\kappa'}} (s)E_i}{\mathbf{C_{\kappa'}}(d/2)}\right)$, for arbitrary $\kappa'<\kappa$. Then_the_computation above_gives_$$\sum_{i=1}^{n-1}\nabla^2\psi((E_i,E_i),(E_i,E_i))\leq_-2(n-1)\varphi'{\mathbf{T}_\kappa}.$$ Since the right hand_side approaches $-\infty$ as $\kappa'$_increases to_$\kappa$, we have a contradiction to the assumption_that_$\psi$ is smooth._Hence no such $\psi$ exists and there is nothing_to prove |
\in \partial {\|\beta^\star\|}_\infty$ where $$\begin{aligned}
\partial {\|\beta^\star\|}_\infty
&= \bigl\{ g:~ \langle g, \beta^\star\rangle = {\|\beta^\star\|}_\infty,~ {\|g\|}_1\leq 1 \bigr\} \nonumber\\
&= \bigl\{ g:~ g_i = 0 \text{ if } {|\beta^\star_i|}<{\|\beta^\star\|}_\infty,~
{\|g\|}_1=1,~ g \circ \beta \geq 0 \bigr\}.\label{eq:subdiff-linf}\end{aligned}$$ This time, note that a vector with all equal values minimizes both the $\ell_2$ and the $\ell_\infty$ norm subject to $\ell_1$ constraints. Therefore, for $t = {|\{i\in[p]:~ {|\beta^\star_i|} = {\|\beta^\star\|}_\infty\}|}$, the optimal $g$ has $t$ nonzero entries with absolute values equal to $1/t$, which yields $$\begin{aligned}
\varphi^2(\beta^\star)
\leq \frac{1}{t} + 1 + 2 \cdot\frac{1}{t}
= 1 + \frac{3}{t}
\leq 4\end{aligned}$$ and finishes the proof.
Consider and observe that ${{\operatorname{ext}}}({\mathcal{B}}^\star) = \{\pm e_i:~ i\in[p]\}$ where $e_i$ is the $i$-th standard basis vector. Define $S = \{i\in[p]:~ \beta^\star_i = {\|\beta^\star\|}_\infty \} $ and $t={|S|}$.
- Case 1: For $i\not\in S$ and $z=\pm e_i$ we have ${{\operatorname{dist}}}^2(z, \partial {\|\beta^\star\|}_\infty) = \min_{g\in \partial {\|\beta^\star\|}_\infty} 1+{\|g\|}_2^2=1 + \frac{1}{t}$. If $S=[p]$, we ignore this case in the maximum over $z\in {{\operatorname{ext}}}({\mathcal{B}}^\star)$ in.
- | \in \partial { \|\beta^\star\|}_\infty$ where $ $ \begin{aligned }
\partial { \|\beta^\star\|}_\infty
& = \bigl\ { g:~ \langle g, \beta^\star\rangle = { \|\beta^\star\|}_\infty,~ { \|g\|}_1\leq 1 \bigr\ } \nonumber\\
& = \bigl\ { g:~ g_i = 0 \text { if } { |\beta^\star_i|}<{\|\beta^\star\|}_\infty,~
{ \|g\|}_1=1,~ g \circ \beta \geq 0 \bigr\}.\label{eq: subdiff - linf}\end{aligned}$$ This time, note that a vector with all equal values minimize both the $ \ell_2 $ and the $ \ell_\infty$ average subject to $ \ell_1 $ constraints. consequently, for $ t = { |\{i\in[p]:~ { |\beta^\star_i| } = { \|\beta^\star\|}_\infty\}|}$, the optimum $ g$ has $ t$ nonzero entries with absolute value adequate to $ 1 / t$, which yields $ $ \begin{aligned }
\varphi^2(\beta^\star)
\leq \frac{1}{t } + 1 + 2 \cdot\frac{1}{t }
= 1 + \frac{3}{t }
\leq 4\end{aligned}$$ and finishes the proof.
Consider and observe that $ { { \operatorname{ext}}}({\mathcal{B}}^\star) = \{\pm e_i:~ i\in[p]\}$ where $ e_i$ is the $ i$-th standard footing vector. Define $ S = \{i\in[p]:~ \beta^\star_i = { \|\beta^\star\|}_\infty \ } $ and $ t={|S|}$.
- Case 1: For $ i\not\in S$ and $ z=\pm e_i$ we have $ { { \operatorname{dist}}}^2(z, \partial { \|\beta^\star\|}_\infty) = \min_{g\in \partial { \|\beta^\star\|}_\infty } 1+{\|g\|}_2 ^ 2=1 + \frac{1}{t}$. If $ S=[p]$, we ignore this shell in the maximum over $ z\in { { \operatorname{ext}}}({\mathcal{B}}^\star)$ in.
- | \in \oartial {\|\beta^\star\|}_\infty$ whtre $$\begin{aligned}
\pcetial {\|\ueta^\stad\|}_\infty
&= \bkgl\{ g:~ \langle g, \beta^\star\ranglx = {\|\bwta^\stqr\|}_\infty,~ {\|g\|}_1\leq 1 \bigr\} \nonjmber\\
&= \bigp\{ g:~ g_i = 0 \texu{ if } {|\beta^\star_i|}<{\|\bevz^\star\|}_\inncy,~
{\|g\|}_1=1,~ f \cire \ueta \geq 0 \bigr\}.\lsbel{eq:subdhff-linf}\end{aligtea}$$ Chis time, note that a vector with alj equal vwlues minimizef bouh ehe $\sll_2$ and the $\ell_\infty$ norm subject fo $\ell_1$ bonstraints. Theregore, for $t = {|\{i\in[p]:~ {|\beta^\star_i|} = {\|\beha^\star\|}_\infty\}|}$, the ophimal $g$ has $j$ nogzero entries with absolute values gqual to $1/t$, which yields $$\begin{aliened}
\vcrphi^2(\beta^\stqr)
\lee \frac{1}{t} + 1 + 2 \rdot\frwc{1}{t}
= 1 + \frac{3}{t}
\leq 4\end{dligned}$$ and finishes bhe pcoof.
Xonsider and observe vhat ${{\operatorname{ext}}}({\iathcal{B}}^\sdax) = \{\pm e_i:~ i\in[p]\}$ where $e_u$ us thg $i$-th staveara bzsms bector. Dehine $S = \{i\in[l]:~ \beta^\star_i = {\|\beta^\star\|}_\infty \} $ amd $n={|X|}$.
- Case 1: Fod $i\not\yn S$ and $z=\pm e_i$ we have ${{\operatorname{dist}}}^2(z, \padtial {\|\beta^\star\|}_\infty) = \mib_{g\in \partial {\|\beta^\star\|}_\lnfty} 1+{\|g\|}_2^2=1 + \srac{1}{t}$. If $S=[p]$, we ignore this case in the maximum ovar $z\ii {{\upexqtornxne{fxt}}}({\mathcal{B}}^\star)$ in.
- | \in \partial {\|\beta^\star\|}_\infty$ where $$\begin{aligned} \partial {\|\beta^\star\|}_\infty g:~ g, \beta^\star\rangle {\|\beta^\star\|}_\infty,~ {\|g\|}_1\leq 1 g_i 0 \text{ if {|\beta^\star_i|}<{\|\beta^\star\|}_\infty,~ {\|g\|}_1=1,~ g \beta \geq 0 \bigr\}.\label{eq:subdiff-linf}\end{aligned}$$ This time, that a vector with all equal values minimizes both the $\ell_2$ and the norm subject to $\ell_1$ constraints. Therefore, for $t = {|\{i\in[p]:~ {|\beta^\star_i|} = {\|\beta^\star\|}_\infty\}|}$, optimal has nonzero with absolute values equal to $1/t$, which yields $$\begin{aligned} \varphi^2(\beta^\star) \leq \frac{1}{t} + 1 + 2 = 1 + \frac{3}{t} \leq 4\end{aligned}$$ and finishes proof. Consider and observe ${{\operatorname{ext}}}({\mathcal{B}}^\star) = \{\pm e_i:~ i\in[p]\}$ $e_i$ the $i$-th basis Define = \{i\in[p]:~ \beta^\star_i {\|\beta^\star\|}_\infty \} $ and $t={|S|}$. - Case 1: For $i\not\in S$ and $z=\pm e_i$ we have ${{\operatorname{dist}}}^2(z, {\|\beta^\star\|}_\infty) = {\|\beta^\star\|}_\infty} 1+{\|g\|}_2^2=1 \frac{1}{t}$. $S=[p]$, ignore this case maximum over $z\in {{\operatorname{ext}}}({\mathcal{B}}^\star)$ in. - | \in \partial {\|\beta^\star\|}_\infty$ wheRe $$\begin{aliGned}
\pArtIal {\|\BeTa^\stAr\|}_\inFty
&= \bigl\{ g:~ \langle G, \Beta^\Star\rangle = {\|\beta^\star\|}_\inftY,~ {\|g\|}_1\leq 1 \BiGR\} \nonUMbEr\\
&= \bigL\{ g:~ g_i = 0 \texT{ If } {|\BETa^\sTaR_i|}<{\|\BetA^\sTAr\|}_\Infty,~
{\|G\|}_1=1,~ g \cIrc \beta \Geq 0 \bigr\}.\labEl{eQ:sUbdiff-linf}\enD{AlIgned}$$ This tIme, Note that a vecTor With alL eQuaL ValueS miNimizEs both THe $\ell_2$ aNd the $\ell_\iNfTY$ norm sUBject to $\ELL_1$ cOnstRaints. Therefore, foR $T = {|\{i\IN[p]:~ {|\beta^\star_i|} = {\|\betA^\star\|}_\iNfTY\}|}$, tHE OptImaL $g$ has $t$ nonzErO entrIEs with aBSoLUTE vaLUes equal to $1/t$, whIch yields $$\beGIn{aLigned}
\VaRphI^2(\Beta^\stAr)
\leq \FrAC{1}{t} + 1 + 2 \cDot\frac{1}{t}
= 1 + \fraC{3}{t}
\leQ 4\end{alignEd}$$ and fINishes tHE proof.
COnsideR anD obServE ThAt ${{\OpeRaTOrnAMe{Ext}}}({\MAthCal{B}}^\star) = \{\Pm E_i:~ I\in[p]\}$ wHere $E_I$ IS The $i$-Th sTandArd baSis vector. DefiNe $S = \{I\in[p]:~ \BEta^\Star_i = {\|\Beta^\sTar\|}_\iNfTy \} $ and $T={|S|}$.
- Case 1: for $i\nOt\In S$ and $z=\pm e_i$ we haVe ${{\opEratornamE{diSt}}}^2(Z, \paRtIal {\|\beTA^\star\|}_\iNftY) = \miN_{g\in \parTial {\|\betA^\StaR\|}_\iNFTY} 1+{\|g\|}_2^2=1 + \Frac{1}{t}$. If $S=[p]$, we ignore tHiS CAsE in the maXimum oVEr $Z\iN {{\OperatorNaMe{eXt}}}({\maTHCal{B}}^\sTar)$ iN.
- | \in \partial {\|\beta^\sta r\|}_\inft y$ wh ere $$ \b egin {ali gned}
\partial {\|\ beta^\star\|}_\infty
& = \bi gl \ { g: ~ \ langl e g, \b e ta ^ \ sta r\ ra ngl e= { \|\be ta^ \star\| }_\infty,~ {\ |g \|}_1\leq 1\ bi gr\} \nonu mbe r\\
&= \bigl \{g:~ g_ i= 0 \text { i f } { |\beta ^ \star_ i|}<{\|\b et a ^\star \ |}_\inf t y ,~
{\ |g\|}_1=1,~ g \ci r c\ beta \geq 0 \b igr\}. \l a be l { eq: sub diff-linf} \e nd{al i gned}$$ Th i s tim e , note that a vector wit h al l equa lval u es min imize sb oth the $\ell_ 2$ a nd the $\ ell_\i n fty$ no r m subje ct to$\e ll_ 1$ c o ns tr ain ts . Th e re for e , f or $t ={| \{ i\in[ p]:~ { | \ beta ^\s tar_ i|} = {\|\beta^\st ar\ |}_\ i nft y\}|} $, th e op ti mal $ g$ has $t$no nzero entries w ithabsoluteval ue s e qu al to $1/t$, wh ich yields $$\beg i n{a li g n e d}
\varphi^2(\beta^\ st a r )
\leq \f rac{1} { t} + 1 + 2 \c do t\f rac{ 1 } {t} = 1+ \ frac{3}{ t}
\l e q4\ end{ali gn ed}$$an d f ini shest he p roof.
Conside r and observe that $ { {\operatornam e {e x t }} } ({\m ath cal{B}}^\st ar)= \{\ pm e _ i: ~ i \ in[p] \}$ w he r e$ e_i$ is the $i$-thst andard basi s vector. Def ine $S = \ { i \ in[p]:~\bet a ^\ s tar_i = {\|\be ta^\s tar\|}_\in f ty \} $and $ t={|S|}$ .
- Ca s e 1: For$i\ not \in S$ a nd $z=\pm e_i$w e hav e${{\ope rat orname{ dis t}} }^2 (z, \ partial { \|\beta^ \s ta r\ |} _\i nfty) = \min_{ g\ in\p art ial { \ |\beta ^\sta r\|} _\ in f ty} 1+{\|g \ |} _ 2 ^2=1 + \ frac {1} {t }$. I f $S = [p] $, we i gnore thi s c a se i nth e maxim um over $z\in { {\operator na me{ ext}}} ( { \mathcal {B}}^\star)$ in.
- | \in \partial_{\|\beta^\star\|}_\infty$ where_$$\begin{aligned}
\partial {\|\beta^\star\|}_\infty
&= \bigl\{ g:~_\langle g,_\beta^\star\rangle_= {\|\beta^\star\|}_\infty,~_{\|g\|}_1\leq_1 \bigr\} \nonumber\\
&=_\bigl\{ g:~ g_i_= 0 \text{ if_} {|\beta^\star_i|}<{\|\beta^\star\|}_\infty,~
{\|g\|}_1=1,~_g_\circ \beta \geq 0 \bigr\}.\label{eq:subdiff-linf}\end{aligned}$$ This time, note that a vector with all equal_values_minimizes both_the_$\ell_2$_and the $\ell_\infty$ norm subject_to $\ell_1$ constraints. Therefore, for_$t =_{|\{i\in[p]:~ {|\beta^\star_i|} = {\|\beta^\star\|}_\infty\}|}$, the optimal $g$ has $t$_nonzero_entries with absolute_values equal to $1/t$, which yields $$\begin{aligned}
\varphi^2(\beta^\star)
\leq \frac{1}{t}_+ 1 + 2 \cdot\frac{1}{t}
=_1 + \frac{3}{t}_
\leq_4\end{aligned}$$_and finishes the proof.
Consider_and observe that ${{\operatorname{ext}}}({\mathcal{B}}^\star) = \{\pm_e_i:~ i\in[p]\}$ where $e_i$ is the_$i$-th standard basis vector. Define $S =_\{i\in[p]:~ \beta^\star_i = {\|\beta^\star\|}_\infty \} $_and $t={|S|}$.
- Case_1: For_$i\not\in S$ and $z=\pm e_i$_we have ${{\operatorname{dist}}}^2(z,_\partial {\|\beta^\star\|}_\infty)_= \min_{g\in \partial_{\|\beta^\star\|}_\infty} 1+{\|g\|}_2^2=1 + \frac{1}{t}$. If $S=[p]$,_we ignore this_case in the maximum over $z\in_{{\operatorname{ext}}}({\mathcal{B}}^\star)$_in.
- _ |
mathcal{R}|_g \geq (\omega_n -\epsilon_i )(\alpha \lambda_i^{-1})^4$$ for some sequence $\epsilon_i \to 0$. The main idea is to use the fact that, for $i\in \mathbb{N}$ sufficiently large, $(B^X(y_i,\alpha \lambda_i^{-1}),\lambda_i d,y_i)$ is arbitrarily close to a Euclidean ball in the pointed Gromov-Hausdroff sense, and to then appeal to a volume convergence theorem for Riemannian manifolds with integral Ricci lower bounds.
Observe that, by Lemma 6.1 of [@BZ1], we have $$|Rc|_{\widetilde{g}^i}(\cdot,0)\leq C(A)(r_{Rm}^{\widetilde{g}^i})^{-1}(\cdot,0),$$ so combining this with the integral estimate for the curvature scale (Theorem 1.7 of [@bam2]) gives $$\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}|Rc|^{3}(\cdot,0) d\widetilde{g}_{0}^{i}\leq\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}(r_{Rm}^{g^i}(\cdot,0))^{-3}d\widetilde{g}_{0}^{i}\leq C(A).$$ Note that we actually have a local $L^p$ bound for $Rc$ for any $p<4$, and the following arguments will work for any $p\in (2,4)$, but we choose $p=3$ for convenience.
Let $\mathcal{H}_d^4=\mathcal{H}^4$ be the $4$-dimensional Hausdorff measure on the metric space $(X,d)$. Because $\mathcal{H}^{4}(X\setminus\mathcal{R})=0$, and because $\mathcal{H}^{4}$ agrees with the Riemannian volume measure on any 4-dimensional Riemannian manifold (in particular, on $\mathcal{R})$, we have $\mathcal{H}^{4}(S)=|S\cap\mathcal{R}|$ for any subset $S\subseteq X$. Thus $$\mathcal{H}_{\lambda_i d}^4(B^X(y_i,\alpha \lambda_i^{-1}))=\lambda_i^4 |B^X(y_i,\alpha \lambda_i | mathcal{R}|_g \geq (\omega_n -\epsilon_i) (\alpha \lambda_i^{-1})^4$$ for some sequence $ \epsilon_i \to 0$. The main idea is to use the fact that, for $ i\in \mathbb{N}$ sufficiently big, $ (B^X(y_i,\alpha \lambda_i^{-1}),\lambda_i d, y_i)$ is randomly close to a Euclidean musket ball in the pointed Gromov - Hausdroff common sense, and to then appeal to a volume convergence theorem for Riemannian manifold with integral Ricci lower boundary.
Observe that, by Lemma 6.1 of [ @BZ1 ], we have $ $ |Rc|_{\widetilde{g}^i}(\cdot,0)\leq C(A)(r_{Rm}^{\widetilde{g}^i})^{-1}(\cdot,0),$$ indeed combining this with the integral appraisal for the curvature scale (Theorem 1.7 of [ @bam2 ]) gives $ $ \int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}|Rc|^{3}(\cdot,0) d\widetilde{g}_{0}^{i}\leq\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}(r_{Rm}^{g^i}(\cdot,0))^{-3}d\widetilde{g}_{0}^{i}\leq C(A).$$ Note that we actually have a local $ L^p$ bound for $ Rc$ for any $ p<4 $, and the following arguments will work for any $ p\in (2,4)$, but we choose $ p=3 $ for public toilet.
Let $ \mathcal{H}_d^4=\mathcal{H}^4 $ be the $ 4$-dimensional Hausdorff measure on the metric quad $ (X, d)$. Because $ \mathcal{H}^{4}(X\setminus\mathcal{R})=0 $, and because $ \mathcal{H}^{4}$ agrees with the Riemannian volume bill on any 4 - dimensional Riemannian manifold (in particular, on $ \mathcal{R})$, we have $ \mathcal{H}^{4}(S)=|S\cap\mathcal{R}|$ for any subset $ S\subseteq X$. Thus $ $ \mathcal{H}_{\lambda_i d}^4(B^X(y_i,\alpha \lambda_i^{-1}))=\lambda_i^4 |B^X(y_i,\alpha \lambda_i | matjcal{R}|_g \geq (\omega_n -\epsilok_i )(\alpha \lambda_i^{-1})^4$$ for soke seqhence $\eprilon_i \to 0$. The main idea is vo uwe tht fact that, for $i\in \mathbb{N}$ dufficiebtly oarge, $(B^X(y_i,\empha \lambda_i^{-1}),\lzlbda_n v,y_i)$ is arbitrarlly close tm a Euclidean taul in the pointed Gromov-Hausdroff senfe, and yo then appeal tj a nojume bokvergence theorem for Riemannian manifonds with intebral Ricci lower bounds.
Obsfrve that, by Lemma 6.1 of [@BZ1], we have $$|Rc|_{\wyeetilde{g}^i}(\cdog,0)\leq C(A)(r_{Rm}^{\widetilde{g}^i})^{-1}(\csot,0),$$ so combining this with the kntegxal estimatg rog the curvatnre scwle (Theorem 1.7 of [@bam2]) cives $$\imt_{B_{\widetilde{g}^{i}}(w_{i},0,1)}|Rc|^{3}(\cvot,0) e\widetilde{g}_{0}^{i}\leq\int_{B_{\wivetilde{g}^{i}}(x_{i},0,1)}(r_{Rm}^{g^i}(\cdot,0))^{-3}q\widetilda{g}_{0}^{n}\leq C(A).$$ Note that we qcrualli hava a uicau $L^k$ bpuhd for $Rc$ for any $p<4$, and the foolowing arguments wolj work for any $p\in (2,4)$, fue we choose $p=3$ for convenience.
Let $\mathcan{H}_d^4=\jathcal{H}^4$ be the $4$-dimensiinal Hausdorff measurg on the mqtric space $(X,d)$. Because $\mathcal{H}^{4}(X\setminus\mathcal{R})=0$, dnd bxcxust $\nathcxo{H}^{4}$ agrees with the Riemannian volume measure on znu 4-cimensional Ricmannian manifold (on pstticular, on $\majhcal{R})$, cs gave $\mathcal{H}^{4}(S)=|S\cap\lathcal{T}|$ for qny subseu $S\sunseteq X$. Thus $$\mathcal{H}_{\lambdq_i d}^4(B^X(y_i,\alphc \lqmbda_i^{-1}))=\lambda_i^4 |B^X(y_i,\clpha \lambda_n | mathcal{R}|_g \geq (\omega_n -\epsilon_i )(\alpha \lambda_i^{-1})^4$$ for $\epsilon_i 0$. The idea is to $i\in sufficiently large, $(B^X(y_i,\alpha d,y_i)$ is arbitrarily to a Euclidean ball in the Gromov-Hausdroff sense, and to then appeal to a volume convergence theorem for Riemannian with integral Ricci lower bounds. Observe that, by Lemma 6.1 of [@BZ1], we $$|Rc|_{\widetilde{g}^i}(\cdot,0)\leq so this the integral estimate for the curvature scale (Theorem 1.7 of [@bam2]) gives $$\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}|Rc|^{3}(\cdot,0) d\widetilde{g}_{0}^{i}\leq\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}(r_{Rm}^{g^i}(\cdot,0))^{-3}d\widetilde{g}_{0}^{i}\leq C(A).$$ Note we actually have a local $L^p$ bound for for any $p<4$, and following arguments will work for $p\in but we $p=3$ convenience. $\mathcal{H}_d^4=\mathcal{H}^4$ be the Hausdorff measure on the metric space $(X,d)$. Because $\mathcal{H}^{4}(X\setminus\mathcal{R})=0$, and because $\mathcal{H}^{4}$ agrees with the Riemannian volume on any manifold (in on we $\mathcal{H}^{4}(S)=|S\cap\mathcal{R}|$ for any X$. Thus $$\mathcal{H}_{\lambda_i d}^4(B^X(y_i,\alpha \lambda_i^{-1}))=\lambda_i^4 |B^X(y_i,\alpha | mathcal{R}|_g \geq (\omega_n -\epsilon_I )(\alpha \lambDa_i^{-1})^4$$ foR soMe sEqUencE $\epsIlon_i \to 0$. The main IDea iS to use the fact that, for $i\iN \mathBb{n}$ SuffICiEntly Large, $(B^X(Y_I,\aLPHa \lAmBdA_i^{-1}),\lAmBDa_I d,y_i)$ iS arBitrariLy close to a eucLiDean ball in thE PoInted GromoV-HaUsdroff sense, And To then ApPeaL To a voLumE convErgencE TheoreM for RiemaNnIAn maniFOlds witH INtEgraL Ricci lower bounds.
oBsERve that, by Lemma 6.1 Of [@BZ1], we HaVE $$|RC|_{\WIdeTilDe{g}^i}(\cdot,0)\leQ C(a)(r_{Rm}^{\wIDetilde{G}^I})^{-1}(\cDOT,0),$$ So cOMbining this wiTh the integrAL esTimate FoR thE CurvatUre scAlE (theOrem 1.7 of [@bam2]) giVes $$\iNt_{B_{\widetiLde{g}^{i}}(x_{I},0,1)}|rc|^{3}(\cdot,0) d\WIdetildE{g}_{0}^{i}\leq\Int_{b_{\wiDetiLDe{G}^{i}}(X_{i},0,1)}(r_{rm}^{G^I}(\cdOT,0))^{-3}d\WidETilDe{g}_{0}^{i}\leq C(a).$$ NOtE that We acTUALLy haVe a LocaL $L^p$ boUnd for $Rc$ for anY $p<4$, aNd thE FolLowinG arguMentS wIll woRk for aNy $p\in (2,4)$, BuT we choose $p=3$ for coNvenIence.
Let $\mAthCaL{H}_d^4=\MaThcal{h}^4$ Be the $4$-dImeNsiOnal HauSdorff mEAsuRe ON THe Metric space $(X,d)$. BecauSe $\MAThCal{H}^{4}(X\setMinus\mAThCaL{r})=0$, and becaUsE $\maThcaL{h}^{4}$ AgreeS witH ThE RiemannIan volUMe MeAsure on AnY 4-dimenSiOnaL RiEmannIAn maNifold (In particUlar, oN $\Mathcal{R})$, we have $\MAthcal{H}^{4}(S)=|S\cap\mAThCAL{R}|$ FOr anY suBset $S\subsetEq X$. THUs $$\maThcaL{h}_{\lAmbDA_i d}^4(B^X(Y_i,\alpHa \LAmBDa_i^{-1}))=\lambda_i^4 |B^X(y_i,\alpha \LaMbda_i | mathcal{R}|_g \geq (\omega _n -\epsil on_i)(\ alp ha \la mbda _i^{-1})^4$$ f o r so me sequence $\epsilon_ i \to 0 $ . Th e m ain i dea ist ou s e t he f act t h at , for $i \in \ma thbb{N}$ s uff ic iently large , $ (B^X(y_i,\ alp ha \lambda_i ^{- 1}),\l am bda _ i d,y _i) $ isarbitr a rily c lose to a E u clidea n ball i n th e po inted Gromov-Haus d ro f f sense, and t o then a p pe a l to avolume con ve rgenc e theore m f o r Rie m annian manifo lds with in t egr al Ric ci lo w er bou nds.
O b ser ve that, by Lem ma 6.1 of [@BZ1 ] , we ha v e $$|Rc |_{\wi det ild e{g} ^ i} (\ cdo t, 0 )\l e qC(A ) (r_ {Rm}^{\w id et ilde{ g}^i } ) ^ { -1}( \cd ot,0 ),$$so combiningthi s wi t h t he in tegra l es ti matefor th e cur va ture scale (The orem 1.7 of [ @ba m2 ])gi ves $ $ \int_{ B_{ \wi detilde {g}^{i} } (x_ {i } , 0 ,1 )}|Rc|^{3}(\cdot,0 )d \ wi detilde{ g}_{0} ^ {i }\ l eq\int_{ B_ {\w idet i l de{g} ^{i} } (x _{i},0,1 )}(r_{ R m} ^{ g^i}(\c do t,0))^ {- 3}d \wi detil d e{g} _{0}^{ i}\leq C (A).$ $ Note that wea ctually havea l o c al $L^p $ b ound for $R c$ f o r an y $p < 4$ , a n d the foll ow i ng arguments will work f or any $p\i n (2,4)$, but we choose $ p =3$ forconv e ni e nce.
Let $\ma thcal {H}_d^4=\m a thcal{H} ^4$ b e the $4 $-dimensi o n al Hausd orf f m eas ure o nthe metric sp a c e $( X, d)$. Be cau se $\ma thc al{ H}^ {4} (X \setminus \mathcal {R }) =0 $, an d bec a use $\ma th cal {H }^{ 4}$ a g rees w ith t he R ie ma n nia n volum e m e a sure o nany4-d im ensio nalR iem annianmanifold(in part ic ul ar, on$\mathcal{R}) $, we have $ \m ath cal{H} ^ { 4}(S)=|S \cap\mathcal{R}|$ for a n y subse t $ S\sub sete q X$. Thu s $ $\math cal { H}_{\l ambda_ i d}^ 4( B^X ( y _i,\a l p ha \l am bda_i^{-1} ) ) =\l ambda _i ^4 | B^X(y_i ,\alpha \lambda_i | mathcal{R}|_g \geq_(\omega_n -\epsilon_i_)(\alpha \lambda_i^{-1})^4$$ for some_sequence $\epsilon_i_\to_0$. The_main_idea is to_use the fact_that, for $i\in \mathbb{N}$_sufficiently large, $(B^X(y_i,\alpha_\lambda_i^{-1}),\lambda_i_d,y_i)$ is arbitrarily close to a Euclidean ball in the pointed Gromov-Hausdroff sense, and_to_then appeal_to_a_volume convergence theorem for Riemannian_manifolds with integral Ricci lower_bounds.
Observe that,_by Lemma 6.1 of [@BZ1], we have $$|Rc|_{\widetilde{g}^i}(\cdot,0)\leq_C(A)(r_{Rm}^{\widetilde{g}^i})^{-1}(\cdot,0),$$_so combining this_with the integral estimate for the curvature scale (Theorem_1.7 of [@bam2]) gives $$\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}|Rc|^{3}(\cdot,0) d\widetilde{g}_{0}^{i}\leq\int_{B_{\widetilde{g}^{i}}(x_{i},0,1)}(r_{Rm}^{g^i}(\cdot,0))^{-3}d\widetilde{g}_{0}^{i}\leq_C(A).$$ Note that_we_actually_have a local $L^p$_bound for $Rc$ for any $p<4$,_and the following arguments will work_for any $p\in (2,4)$, but we choose_$p=3$ for convenience.
Let $\mathcal{H}_d^4=\mathcal{H}^4$ be the_$4$-dimensional Hausdorff measure on the_metric space_$(X,d)$. Because $\mathcal{H}^{4}(X\setminus\mathcal{R})=0$, and because_$\mathcal{H}^{4}$ agrees with_the Riemannian_volume measure on_any 4-dimensional Riemannian manifold (in particular,_on $\mathcal{R})$, we_have $\mathcal{H}^{4}(S)=|S\cap\mathcal{R}|$ for any subset $S\subseteq_X$._Thus $$\mathcal{H}_{\lambda_i d}^4(B^X(y_i,\alpha_\lambda_i^{-1}))=\lambda_i^4_|B^X(y_i,\alpha_\lambda_i |
2007**]{} (2007) 237, \[[[arXiv:0711.0023]{}](http://arxiv.org/abs/0711.0023)\].
B. B. Brandt, G. Endrödi, and S. Schmalzbauer, [*QCD phase diagram for nonzero isospin-asymmetry*]{}, [*Phys. Rev. D*]{} [**97**]{} (2018) 054514, \[[[arXiv:1712.08190]{}](http://arxiv.org/abs/1712.08190)\].
J. B. Kogut, M. A. Stephanov, D. Toublan, J. J. M. Verbaarschot, and A. Zhitnitsky, [*QCD-like theories at finite baryon density*]{}, [*Nucl. Phys. B*]{} [**582**]{} (2000) 477–513, \[[[arXiv:hep-ph/0001171]{}](http://arxiv.org/abs/hep-ph/0001171)\].
J. B. Kogut, D. K. Sinclair, S. J. Hands, and S. E. Morrison, [*Two-color QCD at nonzero quark-number density*]{}, [*Phys. Rev. D*]{} [**64**]{} (2001) 094505, \[[[arXiv:hep-lat/0105026]{}](http://arxiv.org/abs/hep-lat/0105026)\].
G. V. Dunne and S. M. Nishigaki, [*Two-color QCD in 3D at finite baryon density*]{}, [*Nucl. Phys. B*]{} [**654**]{} (2003) 445–465, \[[[arXiv:hep-ph/0210219]{}](http://arxiv.org/abs/hep-ph/0210219)\].
E. Bilgici, F. Bruckmann, C. Gattringer, and C. Hagen, [*[Dual quark condensate and dressed Polyakov loops]{}*]{}, [*Phys. Rev. D*]{} [**77**]{} (2008) 094007, \[[[arXiv:0801.4051]{}](http://arxiv.org/abs/0801.4051)\].
T. DeGrand and R. Hoffmann, [*[QCD with | 2007 * * ] { } (2007) 237, \[[[arXiv:0711.0023]{}](http://arxiv.org / abs/0711.0023)\ ].
B. B. Brandt, G. Endrödi, and S. Schmalzbauer, [ * QCD phase diagram for nonzero isospin - asymmetry * ] { }, [ * Phys. Rev. D * ] { } [ * * 97 * * ] { } (2018) 054514, \[[[arXiv:1712.08190]{}](http://arxiv.org / abs/1712.08190)\ ].
J. B. Kogut, M. A. Stephanov, D. Toublan, J. J. M. Verbaarschot, and A. Zhitnitsky, [ * QCD - like theories at finite baryon density * ] { }, [ * Nucl. Phys. B * ] { } [ * * 582 * * ] { } (2000) 477–513, \[[[arXiv: hep - ph/0001171]{}](http://arxiv.org / abs / hep - ph/0001171)\ ].
J. B. Kogut, D. K. Sinclair, S. J. Hands, and S. E. Morrison, [ * Two - color QCD at nonzero quark - number concentration * ] { }, [ * Phys. Rev. D * ] { } [ * * 64 * * ] { } (2001) 094505, \[[[arXiv: hep - lat/0105026]{}](http://arxiv.org / abs / hep - lat/0105026)\ ].
G. V. Dunne and S. M. Nishigaki, [ * Two - semblance QCD in 3D at finite baryon density * ] { }, [ * Nucl. Phys. B * ] { } [ * * 654 * * ] { } (2003) 445–465, \[[[arXiv: hep - ph/0210219]{}](http://arxiv.org / abs / hep - ph/0210219)\ ].
E. Bilgici, F. Bruckmann, C. Gattringer, and C. Hagen, [ * [ Dual quark cheese condensate and appareled Polyakov loops ] { } * ] { }, [ * Phys. Rev. D * ] { } [ * * 77 * * ] { } (2008) 094007, \[[[arXiv:0801.4051]{}](http://arxiv.org / abs/0801.4051)\ ].
T. DeGrand and R. Hoffmann, [ * [ QCD with | 2007**]{} (2007) 237, \[[[wrXiv:0711.0023]{}](http://arxiv.org/abs/0711.0023)\].
B. B. Brxndt, G. Endrödi, anb S. Schmelzbaued, [*QCD phxse diagram for nonzero isos'in-awymmeugy*]{}, [*Phys. Rev. D*]{} [**97**]{} (2018) 054514, \[[[arXix:1712.08190]{}](http://arxin.org/abs/1712.08190)\].
J. B. Joguu, M. A. Stephanov, D. Tonglan, J. J. M. Verbazvschoc, end A. Zhitnitsky, [*QCD-like tveories at finhtd yaryon density*]{}, [*Nucl. Phys. B*]{} [**582**]{} (2000) 477–513, \[[[arXiv:hqp-ph/0001171]{}](httl://agxiv.org/abs/hep-pr/0001171)\].
J. B. Kpdut, S. K. Sinclair, S. J. Hands, and S. E. Morrison, [*Fwo-colog QCD at nonzero auark-number density*]{}, [*Phys. Rfv. D*]{} [**64**]{} (2001) 094505, \[[[arXiv:hep-lat/0105026]{}](htto://arxiv.org/abw/hep-jqt/0105026)\].
G. V. Dunne ana S. M. Nishigaki, [*Two-color QCD in 3D at finite baryon denskty*]{}, [*Nbcl. Phys. B*]{} [**654**]{} (2003) 445–465, \[[[arDhv:hep-ph/0210219]{}](http://erxiv.ogg/abs/hep-ph/0210219)\].
E. Bilgici, F. Bsuckmanm, C. Gattringer, snd C. Hqgen, [*[Dual quark condeisate and dressed Pojyakov lompa]{}*]{}, [*Phys. Rev. D*]{} [**77**]{} (2008) 094007, \[[[aeXuv:0801.4051]{}](httk://arxie.org/xvs/0801.4051)\].
T. AeGdaid znd R. Hlffjann, [*[QCD wjth | 2007**]{} (2007) 237, \[[[arXiv:0711.0023]{}](http://arxiv.org/abs/0711.0023)\]. B. B. Brandt, and Schmalzbauer, [*QCD diagram for nonzero (2018) \[[[arXiv:1712.08190]{}](http://arxiv.org/abs/1712.08190)\]. J. B. M. A. Stephanov, Toublan, J. J. M. Verbaarschot, and Zhitnitsky, [*QCD-like theories at finite baryon density*]{}, [*Nucl. Phys. B*]{} [**582**]{} (2000) 477–513, J. B. Kogut, D. K. Sinclair, S. J. Hands, and S. E. Morrison, QCD nonzero density*]{}, Rev. D*]{} [**64**]{} (2001) 094505, \[[[arXiv:hep-lat/0105026]{}](http://arxiv.org/abs/hep-lat/0105026)\]. G. V. Dunne and S. M. Nishigaki, [*Two-color QCD in at finite baryon density*]{}, [*Nucl. Phys. B*]{} [**654**]{} 445–465, \[[[arXiv:hep-ph/0210219]{}](http://arxiv.org/abs/hep-ph/0210219)\]. E. Bilgici, Bruckmann, C. Gattringer, and C. [*[Dual condensate and Polyakov [*Phys. D*]{} [**77**]{} (2008) \[[[arXiv:0801.4051]{}](http://arxiv.org/abs/0801.4051)\]. T. DeGrand and R. Hoffmann, [*[QCD with | 2007**]{} (2007) 237, \[[[arXiv:0711.0023]{}](http://arxiv.org/abs/0711.0023)\].
B. B. BraNdt, G. EndrödI, and S. schMalZbAuer, [*qCD pHase diagram for NOnzeRo isospin-asymmetry*]{}, [*Phys. rev. D*]{} [**97**]{} (2018) 054514, \[[[aRXIV:1712.08190]{}](httP://ArXiv.orG/abs/1712.08190)\].
J. B. KOGuT, m. a. StEpHaNov, d. TOUbLan, J. J. m. VeRbaarscHot, and A. ZhiTniTsKy, [*QCD-like theORiEs at finite BarYon density*]{}, [*NuCl. PHys. B*]{} [**582**]{} (2000) 477–513, \[[[arxiV:heP-Ph/0001171]{}](httP://arXiv.orG/abs/heP-Ph/0001171)\].
J. B. KoGut, D. K. SincLaIR, S. J. HanDS, and S. E. MORRiSon, [*TWo-color QCD at nonzeRO qUArk-number densiTy*]{}, [*Phys. reV. d*]{} [**64**]{} (2001) 094505, \[[[aRxIv:hEp-lAt/0105026]{}](http://arxiV.oRg/abs/HEp-lat/0105026)\].
G. V. dUnNE ANd S. m. nishigaki, [*Two-cOlor QCD in 3D aT FinIte barYoN deNSity*]{}, [*NuCl. PhyS. B*]{} [**654**]{} (2003) 445–465, \[[[ARXiV:hep-ph/0210219]{}](http://aRxiv.Org/abs/hep-Ph/0210219)\].
E. BilGIci, F. BruCKmann, C. GAttrinGer, And c. HagEN, [*[DUaL quArK ConDEnSatE And Dressed POlYaKov loOps]{}*]{}, [*PHYS. rEv. D*]{} [**77**]{} (2008) 094007, \[[[aRXiV:0801.4051]{}](httP://arxiV.org/abs/0801.4051)\].
T. DeGraNd aNd R. HOFfmAnn, [*[QCd with | 2007**]{} (2007) 237, \[[[ arXiv:0711 .0023 ]{} ](h tt p:// arxi v.org/abs/0711 . 0023 )\].
B. B. Brandt, G. Endr öd i , an d S . Sch malzbau e r, [ *QC Dph ase d i ag ram f ornonzero isospin-a sym me try*]{}, [*P h ys . Rev. D*] {}[**97**]{} ( 201 8) 054 51 4,\ [[[ar Xiv :1712 .08190 ] {}](ht tp://arxi v. o rg/abs / 1712.08 1 9 0) \].
J. B. Kogut, M.A .S tephanov, D. T oublan ,J .J . M. Ve rbaarschot ,and A . Zhitni t sk y , [*Q C D-like theori es at finit e ba ryon d en sit y *]{},[*Nuc l. Phy s. B*]{} [* *582 **]{} (20 00) 47 7 –513, \ [ [[arXiv :hep-p h/0 001 171] { }] (h ttp :/ / arx i v. org / abs /hep-ph/ 00 01 171)\ ].
J . B . Ko gut , D. K. S inclair, S. J . H ands , an d S.E. Mo rris on , [*T wo-col or QC Dat nonzero quar k-nu mber dens ity *] {}, [ *Phys . Rev.D*] {}[**64** ]{} (20 0 1)09 4 5 0 5, \[[[arXiv:hep-lat /0 1 0 50 26]{}](h ttp:// a rx iv . org/abs/ he p-l at/0 1 0 5026) \].G. V. Dunn e andS .M. Nishig ak i, [*T wo -co lor QCDi n 3D at fi nite bar yon d e nsity*]{}, [*N u cl. Phys. B*] { }[ * *6 5 4**] {}(2003) 445– 465, \[[[ arXi v :h ep- p h/021 0219] {} ] (h t tp://arxiv.org/abs/ he p-ph/0 21021 9)\].
E. Bil gici, F. B r u c kmann, C . Ga t tr i nger, and C. H agen, [*[Dual q u ark cond ensat e and dr essed Pol y a kov loop s]{ }*] {}, [* P h ys . Rev. D*]{}[ * *77* *] {} (200 8)094007, \[ [[a rXi v:0 80 1.4051]{} ](http:/ /a rx iv .o rg/ abs/0 8 01.4051) \] .
T. De Grand and R. Hoff mann ,[* [ QCD with | 2007**]{} (2007)_237, \[[[arXiv:0711.0023]{}](http://arxiv.org/abs/0711.0023)\].
B. B. Brandt,_G. Endrödi, and S. Schmalzbauer, [*QCD_phase diagram_for_nonzero isospin-asymmetry*]{},_[*Phys._Rev. D*]{} [**97**]{}_(2018) 054514, \[[[arXiv:1712.08190]{}](http://arxiv.org/abs/1712.08190)\].
J. B. Kogut,_M. A. Stephanov, D. Toublan, J. J. M. Verbaarschot, and_A. Zhitnitsky, [*QCD-like theories_at_finite baryon density*]{}, [*Nucl. Phys. B*]{} [**582**]{} (2000) 477–513, \[[[arXiv:hep-ph/0001171]{}](http://arxiv.org/abs/hep-ph/0001171)\].
J. B. Kogut, D. K. Sinclair, S. J. Hands, and S. E. Morrison,_[*Two-color_QCD at_nonzero_quark-number_density*]{}, [*Phys. Rev. D*]{} [**64**]{}_(2001) 094505, \[[[arXiv:hep-lat/0105026]{}](http://arxiv.org/abs/hep-lat/0105026)\].
G. V. Dunne and S. M. Nishigaki,_[*Two-color QCD_in 3D at finite baryon density*]{}, [*Nucl. Phys._B*]{}_[**654**]{} (2003) 445–465,_\[[[arXiv:hep-ph/0210219]{}](http://arxiv.org/abs/hep-ph/0210219)\].
E. Bilgici, F. Bruckmann, C. Gattringer, and C. Hagen, [*[Dual quark condensate and_dressed Polyakov loops]{}*]{}, [*Phys. Rev. D*]{}_[**77**]{} (2008) 094007,_\[[[arXiv:0801.4051]{}](http://arxiv.org/abs/0801.4051)\].
T. DeGrand_and_R. Hoffmann, [*[QCD with |
- A Macsyma package for Painlevé analysis of ordinary differential equations, [*Computer Physics Communications*]{} [**42**]{} (1986), 359-383.
Ince E.L., Ordinary differential equations, Dover, New York, 1956.
Ramani A., Grammaticos B., Tremblay S., Integrable system without the painlevé property, [J. Phys. A: Math. Gen]{} [**33**]{} (2000), 3045-3052.
Tamizhmani K.M., Gammaticos B., Ramani A., Do all integrable evolution equations have the Painlevé property?, [*SIGMA*]{} [**3**]{} (2007) 073, 6pp.
Lie S., Klassification und Integration von gewöhnlichen Differentialgleichugen zwischen $x, y,$ die eine Gruppe von Transformationen gestatten, III. [*Archiv for Mathematik og Naturvidenskrab*]{} [**8**]{} (1883), 371-458.
Ibragimov N. H., Magri F., Geometric Proof of Lie Linearization Theorem, Nonlinear Dynamics [**36**]{} (2004), 41-46.
---
abstract: 'We present a Very Large Array[^1] (VLA) 21cm map and optical observations of the region around one of the nearest damped Lyman-$\alpha$ absorber beyond the local group, SBS 1543+593. Two previously uncataloged galaxies have been discovered and a redshift has been determined for a third. All three of these galaxies are at the redshift of SBS 1543+593 and are [< $\buildrel <\over {_\sim}$]{}185 kpc from the damped Lyman-$\alpha$ absorber. We discuss the and optical properties of SBS 1543+593 and its newly identified neighbors. Both SBS 1543+593 and Dwarf 1 have baryonic components that are dominated by neutral gas – unusual for damped Lyman-$\alpha$ absorbers for which only $\sim$5% of the cross-section originates in such strongly gas-dominated systems. What remains unknown is whether low mass gas-rich groups are common surrounding gas-rich galaxies in the local universe and whether the low star-formation rate in these systems is indicative of a young system or a stable, slowly evolving system. We discuss these evolutionary scenarios and future prospects for answering these questions | - A Macsyma package for Painlevé analysis of ordinary differential equations, [ * Computer Physics Communications * ] { } [ * * 42 * * ] { } (1986), 359 - 383.
Ince E.L., average differential equation, Dover, New York, 1956.
Ramani A., Grammaticos B., Tremblay S., Integrable system without the painlevé property, [ J. Phys. deoxyadenosine monophosphate: Math. Gen ] { } [ * * 33 * * ] { } (2000), 3045 - 3052.
Tamizhmani K.M., Gammaticos B., Ramani A., Do all integrable development equations have the Painlevé property? , [ * SIGMA * ] { } [ * * 3 * * ] { } (2007) 073, 6pp.
Lie S., Klassification und Integration von gewöhnlichen Differentialgleichugen zwischen $ x, y,$ die eine Gruppe von Transformationen gestatten, III. [ * Archiv for Mathematik og Naturvidenskrab * ] { } [ * * 8 * * ] { } (1883), 371 - 458.
Ibragimov N. H., Magri F., Geometric Proof of Lie Linearization Theorem, Nonlinear Dynamics [ * * 36 * * ] { } (2004), 41 - 46.
---
abstract:' We deliver a Very Large Array[^1 ] (VLA) 21 cm function and ocular observations of the area around one of the nearest damped Lyman-$\alpha$ absorber beyond the local group, SBS 1543 + 593. Two previously uncataloged galaxies have been discovered and a red shift has been determined for a third. All three of these galaxy are at the redshift of SBS 1543 + 593 and are [ & lt; $ \buildrel < \over { _ \sim}$]{}185 kpc from the damped Lyman-$\alpha$ absorber. We discuss the and ocular properties of SBS 1543 + 593 and its newly identified neighbors. Both SBS 1543 + 593 and Dwarf 1 consume baryonic components that are dominated by neutral gas – strange for damped Lyman-$\alpha$ absorber for which only $ \sim$5% of the cross - section originates in such strongly gas - dominate systems. What remains unknown is whether low mass accelerator - rich groups are common surrounding gas - rich galaxies in the local universe and whether the low star - constitution rate in these systems is indicative of a unseasoned system or a stable, slowly evolving arrangement. We discuss these evolutionary scenarios and future prospects for answering these question | - A Macsyma package for Paiklevé analysis of ordinacy diffsrential equations, [*Computer Physics Rommynicaupons*]{} [**42**]{} (1986), 359-383.
Ince E.L., Ordinafy differvntial eqyatiibs, Dover, Nxs York, 1956.
Vcmani W., Grcmnaticos B., Tremnlay S., Intecrable system fighlut the painlevé property, [J. Phys. A: Mwth. Gen]{} [**33**]{} (2000), 3045-3052.
Tamizhmani K.M., Gamkwticks B., Ramani A., Do all integrable evklution equations hafe the Painlevé property?, [*SIHMA*]{} [**3**]{} (2007) 073, 6pp.
Lie S., Klassiflcation und Intqtration von eewöhnlichen Differentizlgleichugen zwischen $x, y,$ die ekne Gxuppe von Ttcbsfltmationen gewtattvn, III. [*Archiv for Matvematik og Naturvidenxkreb*]{} [**8**]{} (1883), 371-458.
Ibragimov N. H., Magri H., Geometric Proof of Lie Linedrnzation Theorem, Nonlibeqr Dytamiws [**36**]{} (2004), 41-46.
---
abrtrzcv: 'Ws presfnt a Very Ladge Array[^1] (VOA) 21cm map and opticsl ibservations kf the rqgion around one of the nearest damped Nymzn-$\alpha$ absorber beyond the local group, SBS 1543+593. Two prevyously uncataloged galaxies have been discovered dnd a fedwhlft fqs been determined for a third. All three of thefs bakaxies are at bhe redshift of SBX 1543+593 amq are [< $\builarel <\orsr {_\sim}$]{}185 kpc from the famped Jyman-$\qlpha$ absjrbet. We discuss the and opticao properties if SBS 1543+593 and its necly identifizd neibhborx. Both SBS 1543+593 and Dwarf 1 kave bzryonic comoonents tgxt are dominated by nautral gas – unusual for daiped Lymai-$\alphc$ absorbdrs gor whych only $\slm$5% of the cross-section ogiginctes hn such stgongly gas-dominated systems. Whav remains unkmofn ps whethex low kass gas-rich droups are comkon surxoundivg gas-rich galaxixs in the losal universe djd whether tie low stwr-foematuon ratd in these systrms is inbncative od a young system ov a sjagle, slowly evolrnnt system. We disvusr trede edmlutionary swenafior and wuture kxosiecgs fpr answering these qgestjons | - A Macsyma package for Painlevé analysis differential [*Computer Physics [**42**]{} (1986), 359-383. Dover, York, 1956. Ramani Grammaticos B., Tremblay Integrable system without the painlevé property, Phys. A: Math. Gen]{} [**33**]{} (2000), 3045-3052. Tamizhmani K.M., Gammaticos B., Ramani A., all integrable evolution equations have the Painlevé property?, [*SIGMA*]{} [**3**]{} (2007) 073, 6pp. S., und von Differentialgleichugen zwischen $x, y,$ die eine Gruppe von Transformationen gestatten, III. [*Archiv for Mathematik og Naturvidenskrab*]{} (1883), 371-458. Ibragimov N. H., Magri F., Geometric of Lie Linearization Theorem, Dynamics [**36**]{} (2004), 41-46. --- 'We a Very Array[^1] 21cm and optical observations the region around one of the nearest damped Lyman-$\alpha$ absorber beyond the local group, SBS 1543+593. Two uncataloged galaxies discovered and redshift been for a third. of these galaxies are at the 1543+593 and are [< $\buildrel <\over {_\sim}$]{}185 kpc the damped absorber. We discuss the and optical of SBS 1543+593 and its newly identified neighbors. SBS 1543+593 and Dwarf 1 have baryonic components that are dominated by neutral gas – damped Lyman-$\alpha$ absorbers for only $\sim$5% of cross-section in strongly systems. What unknown is whether low mass gas-rich groups are common surrounding gas-rich in the local universe and whether the low star-formation rate systems indicative of a system or a stable, evolving We discuss these evolutionary future for | - A Macsyma package for PainlevÉ analysis oF ordiNarY diFfErenTial Equations, [*CompuTEr PhYsics Communications*]{} [**42**]{} (1986), 359-383.
IncE E.L., OrDiNAry dIFfErentIal equaTIoNS, dovEr, neW YoRk, 1956.
rAmAni A., GRamMaticos b., Tremblay S., intEgRable system wIThOut the painLevÉ property, [J. PhYs. A: math. GeN]{} [**33**]{} (2000), 3045-3052.
TAmiZHmani k.M., GAmmatIcos B., RAMani A., DO all integRaBLe evolUTion equATIoNs haVe the Painlevé propERtY?, [*sIGMA*]{} [**3**]{} (2007) 073, 6pp.
Lie S., KlaSsificAtIOn UND InTegRation von gEwÖhnliCHen DiffEReNTIAlgLEichugen zwiscHen $x, y,$ die einE gruPpe von trAnsFOrmatiOnen gEsTAttEn, III. [*Archiv For MAthematik Og NatuRVidenskRAb*]{} [**8**]{} (1883), 371-458.
IbragImov N. H., magRi F., geomETrIc proOf OF LiE liNeaRIzaTion TheoReM, NOnlinEar DYNAMIcs [**36**]{} (2004), 41-46.
---
aBstRact: 'we preSent a Very LargE ArRay[^1] (Vla) 21cm Map anD optiCal oBsErvatIons of The reGiOn around one of thE neaRest dampeD LyMaN-$\alPhA$ absoRBer beyOnd The Local grOup, SBS 1543+593. TWO prEvIOUSlY uncataloged galaxiEs HAVe Been discOvered ANd A rEDshift haS bEen DeteRMIned fOr a tHIrD. All threE of theSE gAlAxies arE aT the reDsHifT of sBS 1543+593 anD Are [&lT; $\buildRel <\over {_\sIm}$]{}185 kpc FRom the damped LyMAn-$\alpha$ absorbER. WE DIsCUss tHe aNd optical prOperTIes oF SBS 1543+593 ANd Its NEwly iDentiFiED nEIghbors. Both SBS 1543+593 and DwArF 1 have bAryonIc components tHat are domiNATEd by neutRal gAS – uNUsual for damped lyman-$\Alpha$ absorBErs for whIch onLy $\sim$5% of tHe cross-seCTIon origiNatEs iN suCh sTROnGly gas-dominatED SystEmS. What reMaiNs unknoWn iS whEthEr lOw Mass gas-riCh groups ArE cOmMoN suRrounDIng gas-riCh GalAxIes In the LOcal unIversE and WhEtHEr tHe low stAR-fORMatiOn RaTe in TheSe SysteMs is INdiCative oF a young sySteM Or a sTaBlE, slowly Evolving systeM. WE discuss thEsE evOlutioNARy scenarIos and future prospects foR AnsweriNg tHese qUestIons | - A Macsyma package for P ainlevé an alysi s o f o rd inar y di fferential equ a tion s, [*Computer PhysicsCommu ni c atio n s* ]{} [ **42**] { }( 1 986 ), 3 59- 38 3 .
Ince E. L., Ord inary diff ere nt ial equation s ,Dover, New Yo rk, 1956.
R ama ni A., G ram m atico s B ., Tr emblay S., In tegrablesy s tem wi t hout th e pa inle vé property, [J.P hy s . A: Math. Gen ]{} [* *3 3 ** ] { } ( 200 0), 3045-3 05 2.
T a mizhman i K . M . , G a mmaticos B.,Ramani A.,D o a ll int eg rab l e evol ution e q uat ions have t he P ainlevé p ropert y ?, [*SI G MA*]{}[**3** ]{} (2 007) 07 3, 6p p.
Li e S .,K las sificati on u nd In tegr a t i o n vo n g ewöh nlich en Differenti alg leic h uge n zwi schen $x, y ,$ di e eine Grup pe von Transforma tion en gestat ten ,III .[*Arc h iv for Ma the matik o g Natur v ide ns k r a b* ]{} [**8**]{} (188 3) , 37 1-458.
Ibragi m ov N . H., Mag ri F. , Ge o m etric Pro o fof Lie L ineari z at io n Theor em , Nonl in ear Dy namic s [** 36**]{ } (2004) , 41- 4 6.
---
abstra c t: 'We presen t a V er y Lar geArray[^1] ( VLA) 21c m ma p a ndo ptica l obs er v at i ons of the region a ro und on e ofthe nearest d amped Lyma n - $ \alpha$abso r be r beyond the lo cal g roup, SBS1 543+593. Twoprevious ly uncata l o ged gala xie s h ave be e n d iscovered and a red sh ift has be en dete rmi ned fo r a t hird. All three o fth es egal axies are at t he re ds hif t ofS BS 154 3+593 and a re [&l t; $\bu i ld r e l <\ ov er {_\ sim }$ ]{}18 5 kp c fr om thedamped Ly man - $\al ph a$ absorb er. We discus sthe and o pt ica l prop e r ties ofSBS 1543+593 and its ne w ly iden tif ied n eigh bors. Bot h S BS 154 3+5 9 3 andDwarf1 hav ebar y o nic c o m po nen ts that ared o min atedby neu tral ga s – unusual for da m ped Lyman-$\alph a$abso r b er s f o rw hic ho nly $ \sim$5% of the cross-sec ti o noriginates insu ch stro ngly ga s-dom i nated s ystems. W hat remai ns unk n o wnis whether low mas s gas-ric h grou p sare c omm on sur ro und ing g as-ric h ga laxie s in t he local univ er se and w hether the low star-for mation rate in these sy ste m s i s indicat iveof a young sy ste m ora s t able, slo w ly ev o lving sys t em. We di s cu sst h es e evolution a r y sc enari osa nd fut ureprospects for ans w ering these qu esti o n s | -_A Macsyma_package for Painlevé analysis_of ordinary_differential_equations, [*Computer_Physics_Communications*]{} [**42**]{} (1986),_359-383.
Ince E.L., Ordinary_differential equations, Dover, New_York, 1956.
Ramani A.,_Grammaticos_B., Tremblay S., Integrable system without the painlevé property, [J. Phys. A: Math. Gen]{}_[**33**]{}_(2000), 3045-3052.
Tamizhmani_K.M.,_Gammaticos_B., Ramani A., Do all_integrable evolution equations have the_Painlevé property?,_[*SIGMA*]{} [**3**]{} (2007) 073, 6pp.
Lie S., Klassification und_Integration_von gewöhnlichen Differentialgleichugen_zwischen $x, y,$ die eine Gruppe von Transformationen gestatten,_III. [*Archiv for Mathematik og Naturvidenskrab*]{}_[**8**]{} (1883), 371-458.
Ibragimov_N._H.,_Magri F., Geometric Proof_of Lie Linearization Theorem, Nonlinear Dynamics_[**36**]{} (2004), 41-46.
---
abstract: 'We present_a Very Large Array[^1] (VLA) 21cm map_and optical observations of the region_around one of the nearest_damped Lyman-$\alpha$_absorber beyond the local group,_SBS 1543+593. Two_previously uncataloged_galaxies have been_discovered and a redshift has been_determined for a_third. All three of these galaxies_are_at the redshift_of_SBS_1543+593 and_are [< $\buildrel_<\over_{_\sim}$]{}185 kpc_from_the damped Lyman-$\alpha$ absorber. We discuss_the_ and optical properties of SBS 1543+593 and_its newly identified neighbors._Both_SBS 1543+593 and Dwarf_1 have baryonic components that_are dominated by neutral gas –_unusual for_damped Lyman-$\alpha$_absorbers for which only $\sim$5% of the cross-section originates in such_strongly gas-dominated systems. What remains unknown_is whether low mass_gas-rich groups_are_common surrounding gas-rich_galaxies_in the_local universe and whether the low star-formation_rate in_these systems is indicative of a_young system or a_stable,_slowly evolving system. We discuss these_evolutionary scenarios and future prospects for_answering these questions |
=0.3-0) = 0.75$. These are the predicted stellar mass growth factors of the ICL between $z=0.5-0$ and $z=0.3-0$ from [@Contini2014]. Substituting these values into the relevant equations above, we find $(\Delta M_{\star})_{\mathrm{ICL}} \sim 2 \times 10^{11} \, \mathrm{M_{\odot}}$.
We measure the amount of stellar mass contained in the close companions of BCGs to be $\sim 4 \times 10^{11} \, \mathrm{M_{\odot}}$, of which half is transferred to the ICL. Compared to the ICL stellar mass estimate given above, we find there is sufficient stellar material in the galaxies that are likely to merge with the BCGs as major mergers to account for the stellar mass growth of the ICL between $0.0 \leq z \leq 0.3$. This may also imply that our assumption of 50 per cent mass transfer to the BCG during mergers is enough to grow the ICL, arguing against a higher mass transfer.
Conclusions {#sec:concl}
===========
In this paper, we examine the role that mergers play in the stellar mass build-up of BCGs between $0.08 \leq z \leq 0.50$. For this purpose we identify close galaxy pairs within a 50 kpc physical search radius and assume that these close pairs are diagnostics of mergers. The close pair fraction is then used to determine how much stellar mass growth the BCGs have experience over the redshift range of interest due to major mergers.
- We observe a weak trend that the spectroscopically corrected BCG pair fraction increases with decreasing redshift, suggesting that major mergers may become more important towards the present day. The evolution of the pair fractions (Eq. \[eq:fm\_fit\]) within 30 and 50 kpc are respectively given by $m = -6.42 \pm 2.58$ and $m = -2.30 \pm 1.27$.
- Since $z=0.32$ we find the fractional contribution made by major mergers to the stellar mass of a present day BCG to be $24 \pm 14 $ per cent, on average.
- Using data from SALT we extend our study to $z=0.45$. From this redshift, we find the fractional contribution made by major mergers to the stellar mass of a present day BCG to be $29 \pm | = 0.3 - 0) = 0.75$. These are the predicted stellar mass growth divisor of the ICL between $ z=0.5 - 0 $ and $ z=0.3 - 0 $ from [ @Contini2014 ]. substitute these values into the relevant equations above, we find $ (\Delta M_{\star})_{\mathrm{ICL } } \sim 2 \times 10^{11 } \, \mathrm{M_{\odot}}$.
We quantify the amount of stellar mass contain in the close companions of BCGs to be $ \sim 4 \times 10^{11 } \, \mathrm{M_{\odot}}$, of which one-half is transplant to the ICL. Compared to the ICL stellar mass estimate given above, we find there embody sufficient stellar material in the galaxies that are probable to merge with the BCGs as major mergers to report for the stellar aggregate growth of the ICL between $ 0.0 \leq z \leq 0.3$. This may also imply that our assumption of 50 per cent bulk transfer to the BCG during mergers is enough to grow the ICL, arguing against a higher mass transfer.
Conclusions { # sec: concl }
= = = = = = = = = = =
In this paper, we examine the role that mergers play in the stellar mass build - up of BCGs between $ 0.08 \leq z \leq 0.50$. For this function we identify close galax pair within a 50 kpc forcible search radius and assume that these close pairs are diagnostics of mergers. The close pair fraction is then use to determine how much stellar mass growth the BCGs get experience over the redshift range of interest due to major mergers.
- We observe a weak tendency that the spectroscopically corrected BCG pair fraction increases with decreasing redshift, suggest that major mergers may become more important towards the present day. The evolution of the pair fractions (Eq. \[eq: fm\_fit\ ]) within 30 and 50 kpc are respectively given by $ m = -6.42 \pm 2.58 $ and $ megabyte = -2.30 \pm 1.27$.
- Since $ z=0.32 $ we find the fractional contribution make by major mergers to the stellar batch of a present day BCG to be $ 24 \pm 14 $ per penny, on average.
- Using data from SALT we extend our discipline to $ z=0.45$. From this redshift, we find the fractional contribution make by major mergers to the leading mass of a present day BCG to be $ 29 \pm | =0.3-0) = 0.75$. Hhese are the predicted rtellar mass growth fartors or the ICU between $z=0.5-0$ and $z=0.3-0$ from [@Contiii2014]. Sybstiulting these values ingo the repevant ewuatmons above, we fiis $(\Delta M_{\star})_{\jwthrk{MCL}} \sim 2 \times 10^{11} \, \mathrm{M_{\ogot}}$.
We measure dhd cmount of stellar mass contained in ehe cloxe companions of BCGx to gv $\wim 4 \times 10^{11} \, \mathrm{M_{\odot}}$, of wgich hanf is transfetred to the ICL. Compared tl thf ICL stellar mass estimate guven qbove, we fina there is sufficient atellar material in the galaxier thac are likeli fo kerge with vhe BCDs as major mergers do accoint for the stcllar maws growth of the ICL uetween $0.0 \leq z \leq 0.3$. Jhis may ansk imply that our qswumpthon mf 50 owr zenu mess transver to the BCF during meegers is enough to brjq the ICL, arghing adaynst a higher mass transfer.
Conclusions {#vec:doncl}
===========
In this paper, we ezamine the role that lergers pjay in the stellar mass build-up of BCGs between $0.08 \neq z \ueq 0.50$. Nor gyid purpose we identify close galaxy pairs withyh s 50 kpc physical fearch radiis amq assume that these clkse pairs are diagjostics of mwrgers. Tht cloxe pair fraction is then uswd to determpne yow much stellar mcss growth tke BCGx havr experience over the rzdshiff range of lnterest sje to major mergdrs.
- We observe a weak trend trat the s'ectrpscopicxlly corrested BCG pwir fvdction increases wlth dgcreashng redshivt, suggesting that major mergers may become mpra ikportant towavds the present day. The evolujion of tke paif fractiona (Eq. \[eq:hm\_fit\]) within 30 and 50 kpc ara respectivelb given br $m = -6.42 \pm 2.58$ and $m = -2.30 \pm 1.27$.
- Since $a=0.32$ we find the fractuonal contribution maas by major mergzxs to the stellar mars jf a pwasent day BCC to be $24 \pm 14 $ per cent, ok axerabe.
- Using data from SALF we extend our stidn to $z=0.45$. Frim this wedshift, we fond the fractional contcibutiun mace fy major mergers to the stellad mass of a iresent day BSG ti be $29 \pm | =0.3-0) = 0.75$. These are the predicted growth of the between $z=0.5-0$ and values the relevant equations we find $(\Delta \sim 2 \times 10^{11} \, \mathrm{M_{\odot}}$. measure the amount of stellar mass contained in the close companions of BCGs be $\sim 4 \times 10^{11} \, \mathrm{M_{\odot}}$, of which half is transferred to ICL. to ICL mass estimate given above, we find there is sufficient stellar material in the galaxies that are to merge with the BCGs as major mergers account for the stellar growth of the ICL between \leq \leq 0.3$. may imply our assumption of per cent mass transfer to the BCG during mergers is enough to grow the ICL, arguing against higher mass {#sec:concl} =========== this we the role that in the stellar mass build-up of \leq z \leq 0.50$. For this purpose we close galaxy within a 50 kpc physical search and assume that these close pairs are diagnostics mergers. The close pair fraction is then used to determine how much stellar mass growth have experience over the range of interest to mergers. We a weak that the spectroscopically corrected BCG pair fraction increases with decreasing redshift, that major mergers may become more important towards the present evolution the pair fractions \[eq:fm\_fit\]) within 30 and kpc respectively given by $m \pm and \pm - $z=0.32$ we find the contribution made by major mergers the stellar mass of be $24 \pm 14 $ per cent, on - Using data from SALT we extend study to $z=0.45$. From this redshift, we find the fractional contribution made major mergers stellar mass of a present day BCG to $29 \pm | =0.3-0) = 0.75$. These are the predicted stellAr mass growTh facTorS of ThE ICL BetwEen $z=0.5-0$ and $z=0.3-0$ from [@CoNTini2014]. substituting these valueS into ThE ReleVAnT equaTions abOVe, WE FinD $(\DElTa M_{\StAR})_{\mAthrm{iCL}} \Sim 2 \timeS 10^{11} \, \mathrm{M_{\odOt}}$.
WE mEasure the amoUNt Of stellar mAss Contained in tHe cLose coMpAniONs of BcGs To be $\sIm 4 \timeS 10^{11} \, \Mathrm{m_{\odot}}$, of whIcH Half is TRansferRED tO the iCL. Compared to the Icl sTEllar mass estimAte givEn ABoVE, We fInd There is sufFiCient STellar mATeRIAL in THe galaxies thaT are likely tO MerGe with ThE BCgS as majOr merGeRS to Account for tHe stEllar mass Growth OF the ICL BEtween $0.0 \lEq z \leq 0.3$. thiS maY alsO ImPlY thAt OUr aSSuMptIOn oF 50 per cent MaSs TransFer tO THE bCG dUriNg meRgers Is enough to groW thE ICL, ARguIng agAinst A higHeR mass TransfEr.
ConClUsions {#sec:concl}
===========
IN thiS paper, we eXamInE thE rOle thAT mergeRs pLay In the stEllar maSS buIlD-UP Of bCGs between $0.08 \leq z \leq 0.50$. foR THiS purpose We idenTIfY cLOse galaxY pAirS witHIN a 50 kpc PhysICaL search rAdius aND aSsUme that ThEse cloSe PaiRs aRe diaGNostIcs of mErgers. ThE closE Pair fraction is THen used to deteRMiNE HoW Much SteLlar mass groWth tHE BCGS havE ExPerIEnce oVer thE rEDsHIft range of interest dUe To majoR mergErs.
- We observe a Weak trend tHAT The spectRoscOPiCAlly corrected BcG paiR fraction iNCreases wIth deCreasing Redshift, sUGGesting tHat MajOr mErgERS mAy become more iMPOrtaNt Towards The Present Day. the EvoLutIoN of the paiR fractioNs (eq. \[Eq:Fm\_Fit\]) WithiN 30 And 50 kpc arE rEspEcTivEly giVEn by $m = -6.42 \pM 2.58$ and $m = -2.30 \Pm 1.27$.
- SiNcE $z=0.32$ WE fiNd the frACtIONal cOnTrIbutIon MaDe by mAjor MErgErs to thE stellar mAss OF a prEsEnT day BCG To be $24 \pm 14 $ per cent, On Average.
- UsiNg DatA from Salt we extenD our study to $z=0.45$. From this redSHift, we fInd The frActiOnal contrIbuTion maDe bY Major mErgers To the StEllAR Mass oF A PrEseNt Day BCG to be $29 \PM | =0.3-0) = 0.75$. These are the predi ctedste lla rmass gro wth factors of theICL between $z=0.5-0$and $ z= 0 .3-0 $ f rom [ @Contin i 20 1 4 ].Su bs tit ut i ng thes e v alues i nto the re lev an t equationsa bo ve, we fin d $ (\Delta M_{\ sta r})_{\ ma thr m {ICL} } \ sim 2 \time s 10^{1 1} \, \ma th r m{M_{\ o dot}}$. We mea sure the amount o f s t ellar mass con tained i n t h e cl ose companion sof BC G s to be $\ s i m 4\ times 10^{11} \, \mathrm { M_{ \odot} }$ , o f which half i s tr ansferred t o th e ICL. Co mpared to theI CL stel lar ma ssest imat e g iv enab o ve, we fi n d t here issu ff icien t st e l l a r ma ter ialin th e galaxies th atarel ike ly to merg e wi th theBCGs a s maj or mergers to acc ount for theste ll arma ss gr o wth of th e I CL betw een $0. 0 \l eq z \l eq 0.3$. This mayal s o i mply tha t oura ss um p tion of50 pe r ce n t mass tra n sf er to th e BCGd ur in g merge rs is en ou ghtogrowt he I CL, ar guing ag ainst a higher masst ransfer.
Con c lu s i on s {#s ec: concl}
==== ==== = ==
In t h is pa p er, w e exa mi n et he role that merger splay i n the stellar mass build-upo f BCGs bet ween $0 . 08 \leq z \leq 0.50 $. For thi s purpose we i dentifyclose gal a x y pairswit hin a50k p cphysical sear c h rad iu s and a ssu me that th ese cl ose p airs arediagnost ic sof m erg ers.T he close p air f rac tioni s then used tode te r min e how m u ch s tell ar m assgro wt h the BCG s ha ve expe rience ov ert he r ed sh ift ran ge of interes tdue to maj or me rgers. - We o bserve a weak trend tha t the sp ect rosco pica lly corre cte d BCGpai r fract ion in creas es wi t h decr e a si ngre dshift, su g g est ing t ha t ma jor mer gers may become mo r e i mportant towa rds the p re sen t d a y.Th e ev o l ution of the pa ir fractio ns (E q. \[eq:fm \ _fi t\ ]) with in 30 a nd 50 kpc are respecti vely give nby $ m = - 6.42 \pm 2 .58$ and $m = -2. 3 0 \pm 1. 27$.
- Sinc e$z= 0.32$ we fi n d t he fr action al contr ibuti on made by major mergers to the s tellar mass of a presen t d a y B CG to be$24\pm 14 $ p ercen t, on av e rage.
- U sin g data fro m SALT wee xt end o ur study to $ z = 0 .45 $. Fr omt his re dshi ft, we find the f r actional contr ibut i o n m ade by m aj or mergers tothe s t e llar mas sof a presen t day BC Gt o be$29 \p m | =0.3-0) =_0.75$. These_are the predicted stellar_mass growth_factors_of the_ICL_between $z=0.5-0$ and_$z=0.3-0$ from [@Contini2014]._Substituting these values into_the relevant equations_above,_we find $(\Delta M_{\star})_{\mathrm{ICL}} \sim 2 \times 10^{11} \, \mathrm{M_{\odot}}$.
We measure the amount of_stellar_mass contained_in_the_close companions of BCGs to_be $\sim 4 \times 10^{11}_\, \mathrm{M_{\odot}}$,_of which half is transferred to the ICL._Compared_to the ICL_stellar mass estimate given above, we find there is_sufficient stellar material in the galaxies_that are likely_to_merge_with the BCGs as_major mergers to account for the_stellar mass growth of the ICL_between $0.0 \leq z \leq 0.3$. This_may also imply that our assumption_of 50 per cent mass_transfer to_the BCG during mergers is_enough to grow_the ICL,_arguing against a_higher mass transfer.
Conclusions {#sec:concl}
===========
In this paper,_we examine the_role that mergers play in the_stellar_mass build-up of_BCGs_between_$0.08 \leq_z \leq 0.50$._For_this purpose_we_identify close galaxy pairs within a_50_kpc physical search radius and assume that_these close pairs are_diagnostics_of mergers. The close_pair fraction is then used_to determine how much stellar mass_growth the_BCGs have_experience over the redshift range of interest due to major mergers.
-_ We observe a weak_trend that the spectroscopically_corrected BCG_pair_fraction increases with_decreasing_redshift, suggesting_that major mergers may become more important_towards the_present day. The evolution of the_pair fractions (Eq. \[eq:fm\_fit\])_within_30 and 50 kpc are respectively_given by $m = -6.42 \pm_2.58$ and $m = -2.30_\pm_1.27$.
-_ Since $z=0.32$ we_find the fractional contribution made by_major mergers to_the stellar mass of a present day_BCG_to be $24 \pm 14 $_per_cent, on average.
- Using_data_from_SALT we extend our study_to $z=0.45$. From this redshift, we_find the fractional contribution made by major mergers to_the stellar mass_of a present day BCG_to_be_$29 \pm |
a_{11} & a_{12} & a_{13} & a_{14} & a_{15} \\
a_{21} & a_{22} & a_{23} & a_{24} & a_{25} \\
a_{31} & a_{32} & a_{33} & a_{34} & a_{35} \\
a_{41} & a_{42} & a_{43} & a_{44} & a_{45} \\
a_{51} & a_{52} & a_{53} & a_{54} & a_{55} \\
\end{array} \right)
\left(
\begin{array}{c}
f_1\\
f_3\\
f_{5+7} \\
f_{5-7} \\
f_9
\end{array} \right) \.$$ Similarly, by employing $\bar u(p^\prime,s^\prime)
( \rlap/q + \rlap/q^\prime) \gamma_5 u(p,s) $, we have for the $g$-type form factors, $$\left(
\begin{array}{c}
d_1\\
d_3\\
d_{5+7} \\
d_{5-7}
\end{array} \right)=\left(
\begin{array}{ccccc}
b_{11} & b_{12} & b_{13} & b_{14} \\
b_{21} & b_{22} & b_{23} & b_{24} \\
b_{31} & b_{32} & b_{33} & b_{34} \\
b_{41} & b_{42} & b_{43} & b_{44}
\end{array} \right)
\left(
\begin{array}{c}
g_1\\
g_3\\
g_{5+7} \\
g_{5-7}
\end{array} \right) \,$$ To save space, we omit | a_{11 } & a_{12 } & a_{13 } & a_{14 } & a_{15 } \\
a_{21 } & a_{22 } & a_{23 } & a_{24 } & a_{25 } \\
a_{31 } & a_{32 } & a_{33 } & a_{34 } & a_{35 } \\
a_{41 } & a_{42 } & a_{43 } & a_{44 } & a_{45 } \\
a_{51 } & a_{52 } & a_{53 } & a_{54 } & a_{55 } \\
\end{array } \right)
\left (
\begin{array}{c }
f_1\\
f_3\\
f_{5 + 7 } \\
f_{5 - 7 } \\
f_9
\end{array } \right) \.$$ Similarly, by employing $ \bar u(p^\prime, s^\prime)
(\rlap / q + \rlap / q^\prime) \gamma_5 u(p, s) $, we have for the $ g$-type form factors, $ $ \left (
\begin{array}{c }
d_1\\
d_3\\
d_{5 + 7 } \\
d_{5 - 7 }
\end{array } \right)=\left (
\begin{array}{ccccc }
b_{11 } & b_{12 } & b_{13 } & b_{14 } \\
b_{21 } & b_{22 } & b_{23 } & b_{24 } \\
b_{31 } & b_{32 } & b_{33 } & b_{34 } \\
b_{41 } & b_{42 } & b_{43 } & b_{44 }
\end{array } \right)
\left (
\begin{array}{c }
g_1\\
g_3\\
g_{5 + 7 } \\
g_{5 - 7 }
\end{array } \right) \,$$ To spare quad, we omit |
a_{11} & a_{12} & a_{13} & a_{14} & a_{15} \\
a_{21} & a_{22} & a_{23} & a_{24} & a_{25} \\
a_{31} & a_{32} & a_{33} & a_{34} & a_{35} \\
a_{41} & a_{42} & a_{43} & e_{44} & a_{45} \\
a_{51} & a_{52} & a_{53} & a_{54} & a_{55} \\
\enf{array} \rught)
\owft(
\begin{arczy}{c}
f_1\\
f_3\\
f_{5+7} \\
f_{5-7} \\
f_9
\ehf{arrcy} \right) \.$$ Similatly, by emplofing $\bar u(p^\prike,r^\pxime)
( \rlap/q + \rlap/q^\prime) \gamma_5 u(p,s) $, we havr vor the $g$-type sorm sactkgs, $$\left(
\begin{array}{c}
d_1\\
d_3\\
d_{5+7} \\
d_{5-7}
\end{arrzy} \righu)=\left(
\begin{array}{cccvc}
b_{11} & b_{12} & b_{13} & b_{14} \\
h_{21} & b_{22} & b_{23} & b_{24} \\
h_{31} & b_{32} & v_{33} & f_{34} \\
b_{41} & b_{42} & b_{43} & b_{44}
\end{array} \rjght)
\left(
\begin{array}{c}
g_1\\
g_3\\
g_{5+7} \\
g_{5-7}
\end{afray} \xight) \,$$ To sqvw sodce, we omit | a_{11} & a_{12} & a_{13} & a_{14} \\ & a_{22} a_{23} & a_{24} a_{32} a_{33} & a_{34} a_{35} \\ a_{41} a_{42} & a_{43} & a_{44} & \\ a_{51} & a_{52} & a_{53} & a_{54} & a_{55} \\ \end{array} \right) \begin{array}{c} f_1\\ f_3\\ f_{5+7} \\ f_{5-7} \\ f_9 \end{array} \right) \.$$ Similarly, by $\bar ( + \gamma_5 u(p,s) $, we have for the $g$-type form factors, $$\left( \begin{array}{c} d_1\\ d_3\\ d_{5+7} \\ \end{array} \right)=\left( \begin{array}{ccccc} b_{11} & b_{12} & b_{13} b_{14} \\ b_{21} & & b_{23} & b_{24} \\ & & b_{33} b_{34} b_{41} b_{42} & b_{43} b_{44} \end{array} \right) \left( \begin{array}{c} g_1\\ g_3\\ g_{5+7} \\ g_{5-7} \end{array} \right) \,$$ To save space, we |
a_{11} & a_{12} & a_{13} & a_{14} & a_{15} \\
a_{21} & a_{22} & a_{23} & a_{24} & a_{25} \\
a_{31} & a_{32} & a_{33} & a_{34} & a_{35} \\
a_{41} & a_{42} & a_{43} & a_{44} & a_{45} \\
a_{51} & a_{52} & a_{53} & a_{54} & a_{55} \\
\End{array} \riGht)
\leFt(
\bEgiN{aRray}{C}
f_1\\
f_3\\
f_{5+7} \\
F_{5-7} \\
f_9
\end{array} \righT) \.$$ simiLarly, by employing $\bar u(p^\pRime,s^\PrIMe)
( \rlAP/q + \Rlap/q^\Prime) \gaMMa_5 U(P,S) $, we HaVe For ThE $G$-tYpe foRm fActors, $$\lEft(
\begin{arRay}{C}
d_1\\
D_3\\
d_{5+7} \\
d_{5-7}
\end{array} \rIGhT)=\left(
\begin{ArrAy}{ccccc}
b_{11} & b_{12} & b_{13} & b_{14} \\
b_{21} & B_{22} & b_{23} & b_{24} \\
B_{31} & b_{32} & b_{33} & b_{34} \\
b_{41} & b_{42} & B_{43} & b_{44}
\End{ARray} \rIghT)
\left(
\Begin{aRRay}{c}
g_1\\
g_3\\
G_{5+7} \\
g_{5-7}
\end{arraY} \rIGht) \,$$ To sAVe space, WE OmIt |
a_{11} & a_{12} & a_ {13} & a_ {14} & a_ {1 5} \\
a_{21} & a _{22 } & a_{23} & a_{24} & a _ {25} \ \
a_ {31} & a _ {32 }& a _{ 3 3} & a _{3 4} & a_{35} \ \ a_ {41} & a_ { 42 } & a_{4 3}& a_{44} & a_{45} \\
a_{5 1}& a_{52} & a_ {53} & a _{ 5 4} & a_{55} \\ \end{array} \righ t )\ left(
\begin{a rray}{ c}
f _ 1 \\f_3 \\
f_{5+7} \ \
f_{ 5 -7} \\f _9 \ end { array} \right ) \.$$ Simi l arl y, byem plo y ing $\ bar u (p ^ \pr ime,s^\prim e) ( \rlap/q + \rl a p/q^\pr i me) \ga mma_5 u (p, s) $ , w ehav ef ort he $g $ -ty pe formfa ct ors,$$\l e f t (
\be gin {arr ay}{c }
d_1\\
d_3\ \
d _{5+ 7 } \ \
d_{ 5-7}
\en d{ array } \rig ht)=\ le ft(
\begin{arra y}{c cccc}
b_ {11 }& b_{12 } & b _{1 3}& b_{1 4} \\
b _ { 2 1} & b_{22} & b _{ 2 3 }& b_{24 } \ \
b _ {31} & b_{ 32}& b_{ 33}& b_{34} \\ b _{ 41 } & b_ {42} & b_ {43 } & b _{44 }
\ end{arra y} \r i ght)
\left(
\ b egin{array}{c } g _ 1\ \
g_3 \\g_{5+7} \\g_{5 - 7} \end { ar ray } \rig ht) \ ,$ $ T o save space, we omi t |
a_{11} &_ _ a_{12} & _ a_{13}_&_ a_{14}__& a_{15}_ _\\
a_{21} & _ a_{22}_&_ a_{23} & a_{24} & a_{25} \\
a_{31}_&_ __a_{32}_& a_{33} &_ a_{34} & _ a_{35}_ \\
a_{41} & a_{42}_&_ a_{43}_& a_{44} & a_{45} _ \\
a_{51} & _ a_{52} &___a_{53} & a_{54}_ & a_{55} _ \\
\end{array} \right)
\left(
\begin{array}{c}
f_1\\
f_3\\
f_{5+7}_\\
f_{5-7} \\
f_9
\end{array} \right) \.$$ Similarly, by_employing $\bar u(p^\prime,s^\prime)
( \rlap/q +_\rlap/q^\prime) \gamma_5 u(p,s)_$, we_have for the $g$-type form_factors, $$\left(
\begin{array}{c}
d_1\\
d_3\\
d_{5+7}_\\
d_{5-7}
\end{array}_\right)=\left(
\begin{array}{ccccc}
b_{11} &_ b_{12} & _ b_{13} &_ b_{14} __\\
b_{21} &___ b_{22}_& _b_{23}_& _b_{24}_ \\
b_{31}_&_ b_{32} & _b_{33} & b_{34}__ \\
b_{41}_& b_{42}_& b_{43} & _b_{44} _
\end{array}_\right)
\left(
\begin{array}{c}
g_1\\
g_3\\
g_{5+7} \\
g_{5-7}
\end{array} \right) \,$$ To save space, we_omit |
underset{( r_k\ge r_{ k+ 1})}{\longmapsto} { \ifthenelse{\equal{r_{ k+
2}}{}}{\renewcommand*{}{accepting}}{\renewcommand*{}{}}
\settowidth{\myautolength}{$ \max( b_k, b_{ k+ 1}- p_k)$}
\addtolength{\myautolength}{1em}
{\ifthenelse{\lengthtest{\the\myautolength>\minautolength}}
{\setlength{\myautolength}{\myautolength}}
{\setlength{\myautolength}{\minautolength}}}
\begin{tikzpicture}[baseline, shorten >=1pt]
\node[state, initial] (1) {$r_k$};
\node[state, {}, right=\myautolength of 1] (end) {$r_{ k+
2}$};
\path[->] (1) edge node[above] {$ p_k+ p_{ k+ 1}$} node[below] {$ \max( b_k, b_{ k+ 1}- p_k)$} (end);
\end{tikzpicture}
}$$
Informally, any run through the RTEA for $f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$ which maximizes output energy will spend no time in the state with rate $r_{ i+ 1}$, as this time may as well be spent in the state with rate $r_i$ without lowering output energy. To make this argument precise, we prove that this transformation does not change the values of $f$.
Let $f'$ denote the function which results from the transformation. Let $x\in L$ and $t\in {[ 0, \infty]}$. We show first that $f( x, t)\le f'( x, t)$, which is clear if $f( x, t)= \bot$. If $f( x, t)\ne \bot$, then there is an accepting run through the RTEA corresponding to $f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$. Hence we have $t | underset { (r_k\ge r _ { k+ 1})}{\longmapsto } { \ifthenelse{\equal{r _ { k+
2}}{}}{\renewcommand*{}{accepting}}{\renewcommand * { } { } }
\settowidth{\myautolength}{$ \max (b_k, b _ { k+ 1}- p_k)$ }
\addtolength{\myautolength}{1em }
{ \ifthenelse{\lengthtest{\the\myautolength>\minautolength } }
{ \setlength{\myautolength}{\myautolength } }
{ \setlength{\myautolength}{\minautolength } } }
\begin{tikzpicture}[baseline, shorten > = 1pt ]
\node[state, initial ] (1) { $ r_k$ };
\node[state, { }, right=\myautolength of 1 ] (end) { $ r _ { k+
2}$ };
\path[- > ] (1) border node[above ] { $ p_k+ p _ { k+ 1}$ } node[below ] { $ \max (b_k, b _ { k+ 1}- p_k)$ } (goal);
\end{tikzpicture }
} $ $
Informally, any run through the RTEA for $ f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$ which maximizes output department of energy will spend no time in the state with rate $ roentgen _ { i+ 1}$, as this time may as well be spend in the state with rate $ r_i$ without lower output energy. To seduce this argument precise, we prove that this transformation does not change the values of $ f$.
Let $ f'$ announce the function which results from the transformation. Let $ x\in L$ and $ t\in { [ 0, \infty]}$. We testify first that $ f (x, t)\le f' (x, t)$, which is clear if $ f (ten, t)= \bot$. If $ f (x, t)\ne \bot$, then there is an accept run through the RTEA corresponding to $ f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$. Hence we have $ t | undfrset{( r_k\ge r_{ k+ 1})}{\longmapstu} { \ifthendlse{\equal{r_{ k+
2}}{}}{\renewcommaid*{}{acxeptibg}}{\renewcommand*{}{}}
\settowidgh{\myautolvngth}{$ \max( b_k, u_{ k+ 1}- p_k)$}
\addtolengvg{\myautolength}{1sl}
{\iftkeielse{\lengthtest{\jhe\myautolencth>\minautolengdh}}
{\setlength{\myautolength}{\myautjlength}}
{\setleggth{\krautklength}{\minautolength}}}
\begin{tikzpictude}[baselpne, shorten >=1pt]
\npde[state, initial] (1) {$r_k$};
\jode[state, {}, tjghe=\nyautolength of 1] (end) {$r_{ k+
2}$};
\path[->] (1) edge node[above] {$ p_k+ p_{ k+ 1}$} npde[below] {$ \naz( b_n, b_{ k+ 1}- p_k)$} (env);
\end{tihzpicture}
}$$
Informally, dny run through the RBEA fmr $d_1{\mathrel{\mathop{\trianglxright}}}\dotsm{\mathrel{\majhop{\triangnexight}}}f_n$ which maximizws outpot enargy qilu skenv nk time in the state with rate $e_{ i+ 1}$, as this time msy qs well be spsnt in tre state with rate $r_i$ without lowering mutlut energy. To make this argument precise, we krove that this transformation does not change the values ox $f$.
Lev $w'$ dtnite tfw vunction which results from the transformatiog. Ltt $q\in L$ and $t\in {[ 0, \ikfty]}$. We show first tjay $f( x, t)\le f'( x, j)$, which is clear if $f( x, t)= \boh$. If $f( v, t)\ne \bot$, then thete is an accepting run throygh the RTEA xorresponding to $f_1{\lathrel{\mathup{\troanglrright}}}\dotsm{\mathrel{\matho'{\trianfleright}}}f_n$. Jence we gxve $t | underset{( r_k\ge r_{ k+ 1})}{\longmapsto} { \ifthenelse{\equal{r_{ \settowidth{\myautolength}{$ b_k, b_{ 1}- p_k)$} \addtolength{\myautolength}{1em} >=1pt] initial] (1) {$r_k$}; {}, right=\myautolength of (end) {$r_{ k+ 2}$}; \path[->] (1) node[above] {$ p_k+ p_{ k+ 1}$} node[below] {$ \max( b_k, b_{ k+ 1}- (end); \end{tikzpicture} }$$ Informally, any run through the RTEA for $f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$ which maximizes energy spend time the state with rate $r_{ i+ 1}$, as this time may as well be spent in state with rate $r_i$ without lowering output energy. make this argument precise, prove that this transformation does change values of Let denote function which results the transformation. Let $x\in L$ and $t\in {[ 0, \infty]}$. We show first that $f( x, t)\le x, t)$, clear if x, \bot$. $f( x, t)\ne there is an accepting run through to $f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$. Hence we have $t | underset{( r_k\ge r_{ k+ 1})}{\longmapsto} { \iFthenelse{\eQual{r_{ K+
2}}{}}{\reNewCoMmanD*{}{accEpting}}{\renewcomMAnd*{}{}}
\sEttowidth{\myautolength}{$ \mAx( b_k, b_{ K+ 1}- p_K)$}
\AddtOLeNgth{\mYautoleNGtH}{1EM}
{\ifThEnElsE{\lENgThtesT{\thE\myautoLength>\minaUtoLeNgth}}
{\setlengtH{\MyAutolength}{\MyaUtolength}}
{\setLenGth{\myaUtOleNGth}{\miNauTolenGth}}}
\begIN{tikzpIcture}[basElINe, shorTEn >=1pt]
\nodE[STaTe, inItial] (1) {$r_k$};
\node[state, {}, rIGhT=\Myautolength of 1] (End) {$r_{ k+
2}$};
\pAtH[->] (1) EdGE NodE[abOve] {$ p_k+ p_{ k+ 1}$} nodE[bElow] {$ \mAX( b_k, b_{ k+ 1}- p_k)$} (ENd);
\END{TikZPicture}
}$$
InformAlly, any run tHRouGh the RtEa foR $F_1{\mathrEl{\matHoP{\TriAngleright}}}\dOtsm{\Mathrel{\maThop{\trIAngleriGHt}}}f_n$ whiCh maxiMizEs oUtpuT EnErGy wIlL SpeND nO tiME in The state WiTh Rate $r_{ I+ 1}$, as tHIS TIme mAy aS welL be spEnt in the state WitH ratE $R_i$ wIthouT loweRing OuTput eNergy. TO make ThIs argument preciSe, we Prove that ThiS tRanSfOrmatIOn does Not ChaNge the vAlues of $F$.
let $F'$ dENOTe The function which reSuLTS fRom the trAnsforMAtIoN. let $x\in L$ aNd $T\in {[ 0, \InftY]}$. wE show FirsT ThAt $f( x, t)\le f'( X, t)$, whicH Is ClEar if $f( x, T)= \bOt$. If $f( x, T)\nE \boT$, thEn theRE is aN accepTing run tHrougH The RTEA correspONding to $f_1{\mathrEL{\mATHoP{\TriaNglEright}}}\dotsm{\MathREl{\maThop{\TRiAngLErighT}}}f_n$. HeNcE We HAve $t | underset{( r_k\ge r_{ k+ 1 })}{\longm apsto } { \ift henelse{\equal{r_{ k+ 2 } }{}} { \r enewc ommand* { }{ a c cep ti ng }}{ \r e ne wcomm and *{}{}}\settowidt h{\ my autolength}{ $ \ max( b_k,b_{ k+ 1}- p_k) $}\addto le ngt h {\mya uto lengt h}{1em }
{\ift henelse{\ le n gthtes t {\the\m y a ut olen gth>\minautolengt h }} {\s etleng th { \m y a uto len gth}{\myau to lengt h }}
{\ s etlength{\mya utolength}{ \ min autole ng th} } }
\beg in{ti kz p ict ure}[baseli ne,shorten > =1pt] \node[ s tate, i nitial ] ( 1 ) {$ r_k$};
\ no de[st ate, { } , rig ht= \mya utole ngth of 1] (e nd) { $r_ { k+ 2}$ };
\p ath[-> ] (1) edge node[ab ove] {$ p_k+p_{ k + 1 }$ } nod e [below ] { $ \ max( b_ k, b_{k + 1 }- p _ k) $} (end);
\end{tik zp i c tu re}
}$$
Infor m al ly , any run t hro ught h e RTE A fo r $ f_1{\mat hrel{\ m at ho p{\tria ng lerigh t} }}\ dot sm{\m a thre l{\mat hop{\tri angle r ight}}}f_n$ wh i ch maximizeso ut p u te nerg y w ill spend n o ti m e in the st ate withrate$r _ {i + 1}$, as this time m ay aswellbe spent in t he state w i t h rate $r _i$w it h out lowering o utput energy. T o make th is ar gument p recise, w e prove th atthi s t ran s f or mation does n o t cha ng e the v alu es of $ f$.
L et$f' $denote th e functi on w hi ch re sults from the t ran sf orm ation . Let $ x\inL$ a nd $ t \in {[ 0,\ in f t y]}$ .We sho w f ir st th at $ f ( x , t)\le f'( x, t )$, whic his clearif $f( x, t)= \ bot$. If $ f( x, t)\ne \ bot$, th en there is an acceptin g run th rou gh th e RT EA corres pon ding t o $ f _1{\ma threl{ \math op {\t r i angle r i gh t}} }\ dotsm{\mat h r el{ \math op {\tr iangler ight}}}f_n$. Hence wehave $t | underset{( r_k\ge_r_{ k+_1})}{\longmapsto} { _ __ __ _ _ _ __ \ifthenelse{\equal{r_{ k+
2}}{}}{\renewcommand*{}{accepting}}{\renewcommand*{}{}}
\settowidth{\myautolength}{$ \max(_b_k,_b_{ k+_1}-_p_k)$}
\addtolength{\myautolength}{1em}
{\ifthenelse{\lengthtest{\the\myautolength>\minautolength}}
_ _ _{\setlength{\myautolength}{\myautolength}}
_ __{\setlength{\myautolength}{\minautolength}}}
\begin{tikzpicture}[baseline, shorten >=1pt]
_ \node[state, initial] _ _ ___ _(1) _{$r_k$};
\node[state, {}, right=\myautolength of_1] (end) {$r_{ k+
_ 2}$};
_ \path[->] (1) edge_ _ node[above] {$ p_k+ p_{_k+ 1}$} node[below]_{$ \max(_b_k, b_{ k+_1}- p_k)$} (end);
\end{tikzpicture}
}$$
Informally, any run through_the RTEA for_$f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$ which maximizes output energy will_spend_no time in_the_state_with rate_$r_{ i+ 1}$,_as_this time_may_as well be spent in the_state_with rate $r_i$ without lowering output energy._To make this argument_precise,_we prove that this_transformation does not change the_values of $f$.
Let $f'$ denote the_function which_results from_the transformation. Let $x\in L$ and $t\in {[ 0, \infty]}$. We_show first that $f( x, t)\le_f'( x, t)$, which_is clear_if_$f( x, t)=_\bot$._If $f(_x, t)\ne \bot$, then there is an_accepting run_through the RTEA corresponding to $f_1{\mathrel{\mathop{\triangleright}}}\dotsm{\mathrel{\mathop{\triangleright}}}f_n$._Hence we have $t |
79 & 0.047 & -20 03 35.439 & 0.044 & 0.90 &0.03 & 9.22\
G10.34-0.14 & 18 09 01.456 & 0.010 & -20 05 07.703 & 0.010 & 5.33 &0.04 & 9.36\
G10.34-0.14 & 18 08 59.983 & 0.059 & -20 03 35.502 & 0.054 & 0.81 &0.03 & 9.36\
G10.34-0.14 & 18 09 01.457 & 0.007 & -20 05 07.698 & 0.007 & 9.37 &0.05 & 9.50\
G10.34-0.14 & 18 08 59.986 & 0.105 & -20 03 35.460 & 0.097 & 0.54 &0.04 & 9.50\
G10.34-0.14 & 18 09 01.457 & 0.005 & -20 05 07.701 & 0.005 & 15.97 &0.06 & 9.64\
G10.34-0.14 & 18 08 59.977 & 0.154 & -20 03 35.389 & 0.142 & 0.48 &0.05 & 9.64\
G10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.698 & 0.004 & 33.55 &0.12 & 9.77\
G10.34-0.14 & 18 08 59.961 & 0.167 & -20 03 35.165 & 0.154 & 0.84 &0.10 & 9.77\
G10.34-0.14 & 18 09 01.456 & 0.006 & -20 05 07.709 & 0.005 & 59.72 &0.25 & 9.91\
G10.34-0.14 & 18 08 59.978 & 0.112 & -20 03 35.217 & 0.103 & 2.61 &0.21 & 9.91\
G10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.730 & 0.005 & 71.76 &0.28 &10.05\
G10.34- | 79 & 0.047 & -20 03 35.439 & 0.044 & 0.90 & 0.03 & 9.22\
G10.34 - 0.14 & 18 09 01.456 & 0.010 & -20 05 07.703 & 0.010 & 5.33 & 0.04 & 9.36\
G10.34 - 0.14 & 18 08 59.983 & 0.059 & -20 03 35.502 & 0.054 & 0.81 & 0.03 & 9.36\
G10.34 - 0.14 & 18 09 01.457 & 0.007 & -20 05 07.698 & 0.007 & 9.37 & 0.05 & 9.50\
G10.34 - 0.14 & 18 08 59.986 & 0.105 & -20 03 35.460 & 0.097 & 0.54 & 0.04 & 9.50\
G10.34 - 0.14 & 18 09 01.457 & 0.005 & -20 05 07.701 & 0.005 & 15.97 & 0.06 & 9.64\
G10.34 - 0.14 & 18 08 59.977 & 0.154 & -20 03 35.389 & 0.142 & 0.48 & 0.05 & 9.64\
G10.34 - 0.14 & 18 09 01.456 & 0.005 & -20 05 07.698 & 0.004 & 33.55 & 0.12 & 9.77\
G10.34 - 0.14 & 18 08 59.961 & 0.167 & -20 03 35.165 & 0.154 & 0.84 & 0.10 & 9.77\
G10.34 - 0.14 & 18 09 01.456 & 0.006 & -20 05 07.709 & 0.005 & 59.72 & 0.25 & 9.91\
G10.34 - 0.14 & 18 08 59.978 & 0.112 & -20 03 35.217 & 0.103 & 2.61 & 0.21 & 9.91\
G10.34 - 0.14 & 18 09 01.456 & 0.005 & -20 05 07.730 & 0.005 & 71.76 & 0.28 & 10.05\
G10.34- | 79 & 0.047 & -20 03 35.439 & 0.044 & 0.90 &0.03 & 9.22\
G10.34-0.14 & 18 09 01.456 & 0.010 & -20 05 07.703 & 0.010 & 5.33 &0.04 & 9.36\
G10.34-0.14 & 18 08 59.983 & 0.059 & -20 03 35.502 & 0.054 & 0.81 &0.03 & 9.36\
J10.34-0.14 & 18 09 01.457 & 0.007 & -20 05 07.698 & 0.007 & 9.37 &0.05 & 9.50\
G10.34-0.14 & 18 08 59.986 & 0.105 & -20 03 35.460 & 0.097 & 0.54 &0.04 & 9.50\
G10.34-0.14 & 18 09 01.457 & 0.005 & -20 05 07.701 & 0.005 & 15.97 &0.06 & 9.64\
G10.34-0.14 & 18 08 59.977 & 0.154 & -20 03 35.389 & 0.142 & 0.48 &0.05 & 9.64\
E10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.698 & 0.004 & 33.55 &0.12 & 9.77\
G10.34-0.14 & 18 08 59.961 & 0.167 & -20 03 35.165 & 0.154 & 0.84 &0.10 & 9.77\
G10.34-0.14 & 18 09 01.456 & 0.006 & -20 05 07.709 & 0.005 & 59.72 &0.25 & 9.91\
G10.34-0.14 & 18 08 59.978 & 0.112 & -20 03 35.217 & 0.103 & 2.61 &0.21 & 9.91\
G10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.730 & 0.005 & 71.76 &0.28 &10.05\
G10.34- | 79 & 0.047 & -20 03 35.439 & &0.03 & G10.34-0.14 & 18 -20 07.703 & 0.010 5.33 &0.04 & G10.34-0.14 & 18 08 59.983 & & -20 03 35.502 & 0.054 & 0.81 &0.03 & 9.36\ G10.34-0.14 & 09 01.457 & 0.007 & -20 05 07.698 & 0.007 & 9.37 &0.05 9.50\ & 08 & 0.105 & -20 03 35.460 & 0.097 & 0.54 &0.04 & 9.50\ G10.34-0.14 & 18 01.457 & 0.005 & -20 05 07.701 & & 15.97 &0.06 & G10.34-0.14 & 18 08 59.977 0.154 -20 03 & & &0.05 & 9.64\ & 18 09 01.456 & 0.005 & -20 05 07.698 & 0.004 & 33.55 &0.12 & 9.77\ & 18 & 0.167 -20 35.165 0.154 & 0.84 9.77\ G10.34-0.14 & 18 09 01.456 -20 05 07.709 & 0.005 & 59.72 &0.25 9.91\ G10.34-0.14 18 08 59.978 & 0.112 & 03 35.217 & 0.103 & 2.61 &0.21 & G10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.730 & 0.005 & 71.76 G10.34- | 79 & 0.047 & -20 03 35.439 & 0.044 & 0.90 &0.03 & 9.22\
G10.34-0.14 & 18 09 01.456 & 0.010 & -20 05 07.703 & 0.010 & 5.33 &0.04 & 9.36\
G10.34-0.14 & 18 08 59.983 & 0.059 & -20 03 35.502 & 0.054 & 0.81 &0.03 & 9.36\
G10.34-0.14 & 18 09 01.457 & 0.007 & -20 05 07.698 & 0.007 & 9.37 &0.05 & 9.50\
G10.34-0.14 & 18 08 59.986 & 0.105 & -20 03 35.460 & 0.097 & 0.54 &0.04 & 9.50\
G10.34-0.14 & 18 09 01.457 & 0.005 & -20 05 07.701 & 0.005 & 15.97 &0.06 & 9.64\
G10.34-0.14 & 18 08 59.977 & 0.154 & -20 03 35.389 & 0.142 & 0.48 &0.05 & 9.64\
G10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.698 & 0.004 & 33.55 &0.12 & 9.77\
G10.34-0.14 & 18 08 59.961 & 0.167 & -20 03 35.165 & 0.154 & 0.84 &0.10 & 9.77\
G10.34-0.14 & 18 09 01.456 & 0.006 & -20 05 07.709 & 0.005 & 59.72 &0.25 & 9.91\
G10.34-0.14 & 18 08 59.978 & 0.112 & -20 03 35.217 & 0.103 & 2.61 &0.21 & 9.91\
G10.34-0.14 & 18 09 01.456 & 0.005 & -20 05 07.730 & 0.005 & 71.76 &0.28 &10.05\
G10.34- | 79 & 0.047 & -20 03 35.439 & 0.044 & 0.90 &0 .03 & 9.2 2\
G 10.34-0.14 & 1 8 0901.456 & 0.010 & -20 0 5 07. 70 3 & 0 . 01 0 & 5 .33 &0. 0 4& 9.3 6\
G 10. 34 - 0. 14 &1808 59.9 83 & 0.059 &-2 0 03 35.502& 0 .054 & 0.8 1 & 0.03 & 9.36\
G1 0.34-0 .1 4 & 18 09 01 .457& 0.00 7 & -20 05 07.69 8& 0.007 & 9.37& 0 .0 5 &9.50\
G10.34-0.14 &1 8 08 59.986 &0.105&- 20 0 3 3 5.4 60 & 0.097 & 0.54 &0.04 & 9. 5 0 \
G1 0 .34-0.14 & 18 09 01.457& 0. 005 &-2 0 0 5 07.70 1 & 0 .0 0 5 & 15.97 &0.0 6 &9.64\
G10 .34-0. 1 4 & 180 8 59.97 7 & 0. 154 &-200 335 .38 9& 0. 1 42 &0 .48 &0.05 & 9 .6 4\
G1 0.34 - 0 . 1 4 &1809 0 1.456 & 0.005 & -2 0 0 5 07 . 698 & 0. 004 & 33. 55 &0.1 2 & 9. 77\
G 10 .34-0.14 & 18 0 8 59 .961 & 0. 167 & -2 003 35 . 165 &0.1 54& 0.84&0.10 & 9.7 7\ G 1 0. 34-0.14 & 18 09 01 .4 5 6 & 0.006 & -20 0 5 0 7. 7 09 & 0.0 05 &59.7 2 &0.25 & 9 . 91 \
G10.34 -0.14& 1 808 59.9 78 & 0.1 12 &-20 03 3 5 .217 & 0.1 03 & 2.6 1 &0. 2 1 & 9.91\
G10. 3 4-0.14 & 18 0 9 0 1 . 45 6 & 0 .00 5 & -20 0507.7 3 0 &0.00 5 & 71 . 76 &0 .28 & 10 . 05 \
G10.34- | 79 &_0.047 &_-20 03 35.439 &_0.044 &_0.90_&0.03 &_9.22\
G10.34-0.14_& 18 09_01.456 & 0.010_& -20 05 07.703_& 0.010 &_5.33_&0.04 & 9.36\
G10.34-0.14 & 18 08 59.983 & 0.059 & -20 03 35.502 &_0.054_& 0.81_&0.03_&_9.36\
G10.34-0.14 & 18 09 01.457_& 0.007 & -20 05_07.698 &_0.007 & 9.37 &0.05 & 9.50\
G10.34-0.14 & 18_08_59.986 & 0.105_& -20 03 35.460 & 0.097 & 0.54 &0.04_& 9.50\
G10.34-0.14 & 18 09 01.457_& 0.005 &_-20_05_07.701 & 0.005 &_15.97 &0.06 & 9.64\
G10.34-0.14 & 18_08 59.977 & 0.154 & -20_03 35.389 & 0.142 & 0.48 &0.05_& 9.64\
G10.34-0.14 & 18 09 01.456_& 0.005 & -20 05_07.698 &_0.004 & 33.55 &0.12 &_9.77\
G10.34-0.14 & 18_08 59.961_& 0.167 &_-20 03 35.165 & 0.154 &_0.84 &0.10 &_9.77\
G10.34-0.14 & 18 09 01.456 &_0.006_& -20 05_07.709_&_0.005 &_59.72 &0.25 &_9.91\
G10.34-0.14_& 18_08_59.978 & 0.112 & -20 03_35.217_& 0.103 & 2.61 &0.21 & 9.91\
G10.34-0.14_& 18 09 01.456_&_0.005 & -20 05_07.730 & 0.005 & 71.76_&0.28 &10.05\
G10.34- |
matrices $\sigma_i$ ($\tau_i$) operate in spin (particle-hole) space. Choosing the impurity spin ${\bf S}$ to point along the $z$ direction, the $4\times 4$ Hamiltonian separates into independent $2\times2$ Hamiltonians $${\cal H}_\pm = \xi_{\bf p} \tau_z (V\tau_z \pm J S) \delta({\bf r}) + \Delta \tau_x.
\label{Hpm}$$ A standard calculation [@SPientka2013] shows that $H_+$ ($H_-$) has one subgap solution $[u_\epsilon({\bf r}),v_\epsilon({\bf r}]$ ($[u_{-\epsilon}({\bf r}),v_{-\epsilon}({\bf r})]$), whose energy we denote by $\epsilon$ ($-\epsilon$). Since the BdG formalism doubles the degrees of freedom, it is sufficient to consider only the solutions of one of these Hamiltonians, say $H_-$. Its subgap (YSR) state starts out at positive energies at small exchange couplings and crosses to negative energies for strong coupling. Keeping only the contribution of the subgap state with Bogoliubov operator $$\gamma_\epsilon = \int d{\bf r} [u^*_\epsilon({\bf r}) \psi_\downarrow({\bf r}) - v^*_\epsilon({\bf r}) \psi_\uparrow^\dagger({\bf r}) ] ,$$ we can write the electron operators as $$\psi_\downarrow({\bf r}) = u_\epsilon({\bf r}) \gamma_\epsilon+\ldots$$ and $$\psi_\uparrow({\bf r}) = -v^*_\epsilon({\bf r})\gamma_\epsilon^\dagger + \ldots,$$ where the ellipses indicate the contributions of above-gap quasiparticles. The even parity state satisfies $\gamma_\epsilon|{\rm even}\rangle =0$ and the odd-parity state is $|{\rm odd}\rangle = \gamma_\epsilon^\dagger|{\rm even}\rangle$. Note that $\gamma_\epsilon^\dagger$ removes an electron spin of 1/2 from the electron system, so that the spin of the odd parity state is lower by 1/2 compared to the (spinless) even parity state.
In the regime of single-electron tunneling, the differential conductance at $eV=\pm (\Delta_t+\epsilon)$ (for a superconducting tip with gap $\Delta_t$) provides access to the Bogoliubov | matrices $ \sigma_i$ ($ \tau_i$) operate in spin (particle - hole) quad. choose the impurity spin $ { \bf S}$ to point along the $ z$ direction, the $ 4\times 4 $ Hamiltonian separates into autonomous $ 2\times2 $ Hamiltonians $ $ { \cal H}_\pm = \xi_{\bf p } \tau_z (V\tau_z \pm J S) \delta({\bf r }) + \Delta \tau_x.
\label{Hpm}$$ A standard calculation [ @SPientka2013 ] shows that $ H_+$ ($ H_-$) has one subgap solution $ [ u_\epsilon({\bf r}),v_\epsilon({\bf r}]$ ($ [ u_{-\epsilon}({\bf r}),v_{-\epsilon}({\bf r})]$), whose department of energy we announce by $ \epsilon$ ($ -\epsilon$). Since the BdG formalism doubles the degrees of freedom, it is sufficient to consider only the solution of one of these Hamiltonians, say $ H_-$. Its subgap (YSR) state start out at positivist energies at humble exchange couplings and crosse to negative energies for strong yoke. Keeping only the contribution of the subgap state of matter with Bogoliubov operator $ $ \gamma_\epsilon = \int d{\bf roentgen } [ u^*_\epsilon({\bf r }) \psi_\downarrow({\bf r }) - v^*_\epsilon({\bf r }) \psi_\uparrow^\dagger({\bf r }) ] , $ $ we can write the electron operators as $ $ \psi_\downarrow({\bf radius }) = u_\epsilon({\bf r }) \gamma_\epsilon+\ldots$$ and $ $ \psi_\uparrow({\bf r }) = -v^*_\epsilon({\bf r})\gamma_\epsilon^\dagger + \ldots,$$ where the ellipses indicate the contributions of above - gap quasiparticles. The even parity state satisfies $ \gamma_\epsilon|{\rm even}\rangle = 0 $ and the odd - parity state is $ |{\rm odd}\rangle = \gamma_\epsilon^\dagger|{\rm even}\rangle$. Note that $ \gamma_\epsilon^\dagger$ removes an electron spin of 1/2 from the electron system, so that the tailspin of the curious parity department of state is lower by 1/2 compared to the (spinless) even parity state.
In the government of single - electron tunneling, the differential conductance at $ eV=\pm (\Delta_t+\epsilon)$ (for a superconducting tip with gap $ \Delta_t$) provides entree to the Bogoliubov | mahrices $\sigma_i$ ($\tau_i$) operaue in spin (particle-hole) vpace. Dhoosing the impurity spin ${\bf S}$ to plibt aling the $z$ direction, thd $4\times 4$ Jamiltonuan wwparates iifo indeizndenf $2\timzs2$ Hamiltonians $${\gal H}_\pm = \xi_{\tf p} \tau_z (V\tau_s \om J S) \delta({\bf r}) + \Delta \tau_x.
\label{Hpm}$$ A stancagd calculation [@SPitntha2013] sgows that $H_+$ ($H_-$) has one subgap solutjon $[u_\epvilon({\bf r}),v_\epsolon({\bf r}]$ ($[u_{-\epsilon}({\bf r}),v_{-\epsipon}({\bv r})]$), whose energy wf denote by $\epsyoon$ ($-\epsilon$). Rince the BdG formalisj doubles the degrees of freedoo, it ns sufficiebt to wonsider onoy thv solutions on one ox these Hamiltonians, xay $H_-$. Uts subgap (YSR) state vtarts out at posijive energhea at small exchante coupningv ana crussts vo hegatige xnergies fod strong coypling. Keeping only trv contributioh of tre subgap state with Bogoliubov operator $$\camja_\epsilon = \int d{\bf r} [u^*_\epsilon({\bf r}) \psi_\downartow({\bf r}) - v^*_\qpsilon({\bf r}) \psi_\uparrow^\dagger({\bf r}) ] ,$$ we can wride thx dleebvon upegators as $$\psi_\downarrow({\bf r}) = u_\epsilon({\bf r}) \gammw_\spxikon+\ldots$$ and $$\pfi_\uparrow({\bf r}) = -f^*_\gpsilon({\bf r})\gamoa_\epsilon^\sagger + \ldots,$$ wherf the ejlipsws indicaue thr contributions of above-gap quasiparticjws. The even parity state satirfiex $\gamka_\epsilon|{\rm even}\rangle =0$ and the odd-parlty state ks $|{\rm odd}\rangle = \gakmd_\epsilon^\dagger|{\rm even}\ranglq$. Note thet $\gakma_\epsiuon^\dsgger$ wemoves an elecbson spin of 1/2 from hhe epewtron systfm, so that the spin of the odd 'erity state ix nowvr by 1/2 coiparec to the (spinjess) even parijy state.
Iu the fegime of aingle-enectron tungeling, the dixverential coiductance at $wV=\pm (\Delta_t+\dosilon)$ (for a sipercondubtnng tip wuth gap $\Delta_t$) proyides zccess to the Bigooiubov | matrices $\sigma_i$ ($\tau_i$) operate in spin (particle-hole) the spin ${\bf to point along 4$ separates into independent Hamiltonians $${\cal H}_\pm \xi_{\bf p} \tau_z (V\tau_z \pm J \delta({\bf r}) + \Delta \tau_x. \label{Hpm}$$ A standard calculation [@SPientka2013] shows that $H_+$ has one subgap solution $[u_\epsilon({\bf r}),v_\epsilon({\bf r}]$ ($[u_{-\epsilon}({\bf r}),v_{-\epsilon}({\bf r})]$), whose energy we by ($-\epsilon$). the formalism doubles the degrees of freedom, it is sufficient to consider only the solutions of one these Hamiltonians, say $H_-$. Its subgap (YSR) state out at positive energies small exchange couplings and crosses negative for strong Keeping the of the subgap with Bogoliubov operator $$\gamma_\epsilon = \int d{\bf r} [u^*_\epsilon({\bf r}) \psi_\downarrow({\bf r}) - v^*_\epsilon({\bf r}) \psi_\uparrow^\dagger({\bf r}) ,$$ we the electron as r}) u_\epsilon({\bf r}) \gamma_\epsilon+\ldots$$ r}) = -v^*_\epsilon({\bf r})\gamma_\epsilon^\dagger + \ldots,$$ indicate the contributions of above-gap quasiparticles. The even state satisfies even}\rangle =0$ and the odd-parity state $|{\rm odd}\rangle = \gamma_\epsilon^\dagger|{\rm even}\rangle$. Note that $\gamma_\epsilon^\dagger$ an electron spin of 1/2 from the electron system, so that the spin of the state is lower by compared to the even state. the of single-electron the differential conductance at $eV=\pm (\Delta_t+\epsilon)$ (for a superconducting tip with $\Delta_t$) provides access to the Bogoliubov | matrices $\sigma_i$ ($\tau_i$) operate In spin (partIcle-hOle) SpaCe. chooSing The impurity spiN ${\Bf S}$ tO point along the $z$ directiOn, the $4\TiMEs 4$ HaMIlToniaN separaTEs INTo iNdEpEndEnT $2\TiMes2$ HaMilTonians $${\Cal H}_\pm = \xi_{\bf P} \taU_z (v\tau_z \pm J S) \delTA({\bF r}) + \Delta \tau_X.
\laBel{Hpm}$$ A standArd CalculAtIon [@spientKa2013] sHows tHat $H_+$ ($H_-$) hAS one suBgap solutIoN $[U_\epsilON({\bf r}),v_\epSILoN({\bf r}]$ ($[U_{-\epsilon}({\bf r}),v_{-\epsilON}({\bF R})]$), whose energy we Denote By $\EPsILOn$ ($-\ePsiLon$). Since thE BDG forMAlism doUBlES THe dEGrees of freedoM, it is sufficIEnt To consIdEr oNLy the sOlutiOnS Of oNe of these HaMiltOnians, say $h_-$. Its suBGap (YSR) sTAte starTs out aT poSitIve eNErGiEs aT sMAll EXcHanGE coUplings aNd CrOsses To neGATIVe enErgIes fOr strOng coupling. KeEpiNg onLY thE contRibutIon oF tHe subGap staTe witH BOgoliubov operatOr $$\gaMma_\epsiloN = \inT d{\Bf r} [U^*_\ePsiloN({\Bf r}) \psi_\DowNarRow({\bf r}) - v^*_\Epsilon({\BF r}) \pSi_\UPARrOw^\dagger({\bf r}) ] ,$$ we can wrItE THe Electron OperatORs As $$\PSi_\downarRoW({\bf R}) = u_\epSILon({\bf R}) \gamMA_\ePsilon+\ldOts$$ and $$\PSi_\UpArrow({\bf R}) = -v^*_\EpsiloN({\bF r})\gAmmA_\epsiLOn^\daGger + \ldOts,$$ where The elLIpses indicate tHE contributionS Of ABOvE-Gap qUasIparticles. THe evEN parIty sTAtE saTIsfieS $\gammA_\ePSiLOn|{\rm even}\rangle =0$ and thE oDd-pariTy staTe is $|{\rm odd}\rangLe = \gamma_\epsILON^\dagger|{\rM eveN}\RaNGle$. Note that $\gamMa_\epsIlon^\dagger$ REmoves an ElectRon spin oF 1/2 from the eLECtron sysTem, So tHat The SPIn Of the odd paritY STate Is Lower by 1/2 ComPared to The (SpiNleSs) eVeN parity stAte.
In the ReGiMe Of SinGle-elECtron tunNeLinG, tHe dIfferENtial cOnducTancE aT $ev=\Pm (\DElta_t+\epSIlON)$ (For a SuPeRconDucTiNg tip With GAp $\DElta_t$) prOvides accEss TO the boGoLiubov | matrices $\sigma_i$ ($\ta u_i$) oper ate i n s pin ( part icle -hole) space.C hoos ing the impurity spin${\bf S } $ to po int a long th e $ z $ di re ct ion ,t he $4\t ime s 4$ Ha miltoniansep ar ates into in d ep endent $2\ tim es2$ Hamilto nia ns $${ \c alH }_\pm =\xi_{ \bf p} \tau_z (V\tau_z \ p m J S) \delta( { \ bf r}) + \Delta \tau_x. \ l abel{Hpm}$$ Astanda rd ca l c ula tio n [@SPient ka 2013] shows t h at $ H _+$ ($H_-$) has o ne subgap s o lut ion $[ u_ \ep s ilon({ \bf r }) , v_\ epsilon({\b f r} ]$ ($[u_{ -\epsi l on}({\b f r}),v_ {-\eps ilo n}( {\bf r} )] $), w h ose en erg y we denoteby $ \epsi lon$ ( $ - \eps ilo n$). Sinc e the BdG for mal ismd oub les t he de gree sof fr eedom, it i ssufficient to c onsi der onlythe s olu ti ons o f one o f t hes e Hamil tonians , sa y$ H _ -$ . Its subgap (YSR) s t a te startsout at po si t ive ener gi esat s m a ll ex chan g ecoupling s andc ro ss es to n eg ativeen erg ies fors tron g coup ling. Ke eping only the contr i bution of the su b g ap stat e w ith Bogoliu bovo pera tor$ $\ gam m a_\ep silon = \i n t d{\bf r} [u^*_\ep si lon({\ bf r} ) \psi_\down arrow({\bf r } ) - v^*_ \eps i lo n ({\bf r}) \ps i_\up arrow^\dag g er({\bfr}) ] ,$$we can wr i t e the el ect ron op era t o rs as $$\psi_\d o w narr ow ({\bf r })= u_\ep sil on( {\b f r }) \gamma_\ epsilon+ \l do ts $$ an d $$\ p si_\upar ro w({ \b f r }) =- v^*_\e psilo n({\ bf r } )\g amma_\e p si l o n^\d ag ge r +\ld ot s,$$wher e th e ellip ses indic ate theco nt ributio ns of above-g ap quasipart ic les . Thee v en parit y state satisfies $\gam m a_\epsi lon |{\rm eve n}\rangle =0 $ andthe odd-pa rity s tateis $| { \ rm od d } \r ang le = \gamma_ \ e psi lon^\ da gger |{\rm e ven}\rangle$. Note tha t $\gamma_\ep sil on^\ d a gg er$ re m ove sa n e l e ctron spin of 1 /2 from th ee le ctron syst e m,so that t he spin of t h e odd p arity sta te is low er by1 / 2 c ompared to the (sp inless) e v en pa r it y sta te.
In t he re gimeof sin g le- elect ron tu nn eling, thedi fferenti al conductance at $eV=\ pm (\D elta_ t+\ epsilon)$ (f o r a supercon duct ing tip wi thgap $\De lta _ t$) p rovi d es ac c ess t o th e Bogoliub o v | matrices_$\sigma_i$ ($\tau_i$)_operate in spin (particle-hole)_space. Choosing_the_impurity spin_${\bf_S}$ to point_along the $z$_direction, the $4\times 4$_Hamiltonian separates into_independent_$2\times2$ Hamiltonians $${\cal H}_\pm = \xi_{\bf p} \tau_z (V\tau_z \pm J S) \delta({\bf r})_+_\Delta \tau_x.
_\label{Hpm}$$_A_standard calculation [@SPientka2013] shows that_$H_+$ ($H_-$) has one subgap_solution $[u_\epsilon({\bf_r}),v_\epsilon({\bf r}]$ ($[u_{-\epsilon}({\bf r}),v_{-\epsilon}({\bf r})]$), whose energy we_denote_by $\epsilon$ ($-\epsilon$)._Since the BdG formalism doubles the degrees of freedom,_it is sufficient to consider only_the solutions of_one_of_these Hamiltonians, say $H_-$._Its subgap (YSR) state starts out_at positive energies at small exchange_couplings and crosses to negative energies for_strong coupling. Keeping only the contribution_of the subgap state with_Bogoliubov operator_$$\gamma_\epsilon = \int d{\bf r}_[u^*_\epsilon({\bf r}) _\psi_\downarrow({\bf r})_- v^*_\epsilon({\bf r})_ \psi_\uparrow^\dagger({\bf r}) ] _ ,$$ we_can write the electron operators as_$$\psi_\downarrow({\bf_r}) = u_\epsilon({\bf_r})_\gamma_\epsilon+\ldots$$_and $$\psi_\uparrow({\bf_r}) = -v^*_\epsilon({\bf_r})\gamma_\epsilon^\dagger_+ \ldots,$$_where_the ellipses indicate the contributions of_above-gap_quasiparticles. The even parity state satisfies $\gamma_\epsilon|{\rm_even}\rangle =0$ and the_odd-parity_state is $|{\rm odd}\rangle_= \gamma_\epsilon^\dagger|{\rm even}\rangle$. Note that_$\gamma_\epsilon^\dagger$ removes an electron spin of_1/2 from_the electron_system, so that the spin of the odd parity state is_lower by 1/2 compared to the_(spinless) even parity state.
In_the regime_of_single-electron tunneling, the_differential_conductance at_$eV=\pm (\Delta_t+\epsilon)$ (for a superconducting tip with_gap $\Delta_t$)_provides access to the Bogoliubov |
i & 0 & 1
\end{pmatrix} N^{-1}(z),$$ which gives us.
This completes the proof of Proposition \[prop:p-1\].
We conclude this section by evaluating $E(-1)$ and $E'(-1)$ for later use. The calculations are straightforward and cumbersome by combining and the asymptotics of $N(z)$ and $f(z)$ given in and. We omit the details but present the results below. $$\label{eq: E(-1)}
E(-1) = \begin{pmatrix}
-i 2^{\frac{1}{12}} 3^{\frac14} C_1^{\frac12} \pi^{\frac12} s^{\frac23} & - \frac{2^{\frac13}}{3^{\frac12}} &
- \frac{1}{2^{\frac5{12}} 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23} } \\
- i 2^{-\frac14} 3^{\frac14} C_1^{\frac12} \pi^{\frac12} s^{\frac23} & \frac{2}{3^{\frac12}} &
\frac{5}{ 2^{\frac34} 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23}} \\
i\frac{3^{\frac14} C_1^{\frac12} \pi^{\frac12} s^{\frac23} }{
2^{\frac{19}{12}}} & -\frac{5}{2^{\frac43} 3^{\frac12}} & \frac{25}{
2^{\frac{25}{12}} 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23}}
\end{pmatrix},$$ and $$\label{eq: Ep(-1)}
E'(-1) = \begin{pmatrix}
-i\frac{ (13 C_1 + 108 C_3) \pi^{\frac12} s^{\frac23}}{36 \cdot 2^{\frac{11}{12}} 3^{\frac34} C_1^{\frac12}} & * & * \\
i \frac{(35 C_1 - 108 C_3) \pi^{\frac12} s^{\frac23}}{72 \cdot 2 | i & 0 & 1
\end{pmatrix } N^{-1}(z),$$ which gives us.
This completes the proof of Proposition \[prop: p-1\ ].
We conclude this incision by measure $ E(-1)$ and $ E'(-1)$ for later use. The calculations are aboveboard and cumbersome by combining and the asymptotics of $ N(z)$ and $ f(z)$ give in and. We omit the details but salute the results below. $ $ \label{eq: E(-1) }
E(-1) = \begin{pmatrix }
-i 2^{\frac{1}{12 } } 3^{\frac14 } C_1^{\frac12 } \pi^{\frac12 } s^{\frac23 } & - \frac{2^{\frac13}}{3^{\frac12 } } &
- \frac{1}{2^{\frac5{12 } } 3^{\frac54 } C_1^{\frac12 } \pi^{\frac12 } s^{\frac23 } } \\
- i 2^{-\frac14 } 3^{\frac14 } C_1^{\frac12 } \pi^{\frac12 } s^{\frac23 } & \frac{2}{3^{\frac12 } } &
\frac{5 } { 2^{\frac34 } 3^{\frac54 } C_1^{\frac12 } \pi^{\frac12 } s^{\frac23 } } \\
i\frac{3^{\frac14 } C_1^{\frac12 } \pi^{\frac12 } s^{\frac23 } } {
2^{\frac{19}{12 } } } & -\frac{5}{2^{\frac43 } 3^{\frac12 } } & \frac{25 } {
2^{\frac{25}{12 } } 3^{\frac54 } C_1^{\frac12 } \pi^{\frac12 } s^{\frac23 } }
\end{pmatrix},$$ and $ $ \label{eq: Ep(-1) }
E'(-1) = \begin{pmatrix }
-i\frac { (13 C_1 + 108 C_3) \pi^{\frac12 } s^{\frac23}}{36 \cdot 2^{\frac{11}{12 } } 3^{\frac34 } C_1^{\frac12 } } & * & * \\
i \frac{(35 C_1 - 108 C_3) \pi^{\frac12 } s^{\frac23}}{72 \cdot 2 | i & 0 & 1
\end{pmatrix} N^{-1}(z),$$ which nives us.
This comklwtes tie proor of Prooosition \[prop:p-1\].
We conclude thms swctiob by evaluating $E(-1)$ and $D'(-1)$ for latvr use. Thw caoxulations ede stralyhtfodaard end cumbersome ny combininc and the asym[tutncs of $N(z)$ and $f(z)$ given in and. We omie the drtwils but presegt tnq reallus below. $$\label{eq: E(-1)}
E(-1) = \begin{pmatrjx}
-i 2^{\frac{1}{12}} 3^{\frac14} C_1^{\grac12} \pi^{\frac12} s^{\frac23} & - \frac{2^{\fgac13}}{3^{\fgac12}} &
- \frac{1}{2^{\frac5{12}} 3^{\vrac54} C_1^{\frac12} \kj^{\frwx12} s^{\frac23} } \\
- i 2^{-\frac14} 3^{\frac14} C_1^{\frac12} \pi^{\ftac12} s^{\frac23} & \frac{2}{3^{\frac12}} &
\frac{5}{ 2^{\fraz34} 3^{\frae54} C_1^{\frac12} \pi^{\ftcx12} s^{\vtac23}} \\
i\frac{3^{\hrac14} C_1^{\srac12} \pi^{\frac12} s^{\frac23} }{
2^{\xrac{19}{12}}} & -\ftac{5}{2^{\frac43} 3^{\frac12}} & \nrac{25}{
2^{\frqc{25}{12}} 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23}}
\end{pmatrix},$$ and $$\laben{es: Ep(-1)}
E'(-1) = \begin{pmarrux}
-i\frdc{ (13 X_1 + 108 C_3) \ki^{\fcac12} s^{\frac23}}{36 \cdkt 2^{\frac{11}{12}} 3^{\frzc34} C_1^{\frac12}} & * & * \\
i \frac{(35 C_1 - 108 V_3) \io^{\frac12} s^{\frac23}}{72 \ddot 2 | i & 0 & 1 \end{pmatrix} N^{-1}(z),$$ us. completes the of Proposition \[prop:p-1\]. evaluating and $E'(-1)$ for use. The calculations straightforward and cumbersome by combining and asymptotics of $N(z)$ and $f(z)$ given in and. We omit the details but the results below. $$\label{eq: E(-1)} E(-1) = \begin{pmatrix} -i 2^{\frac{1}{12}} 3^{\frac14} C_1^{\frac12} \pi^{\frac12} & \frac{2^{\frac13}}{3^{\frac12}} - 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23} } \\ - i 2^{-\frac14} 3^{\frac14} C_1^{\frac12} \pi^{\frac12} s^{\frac23} & \frac{2}{3^{\frac12}} & 2^{\frac34} 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23}} \\ i\frac{3^{\frac14} C_1^{\frac12} s^{\frac23} }{ 2^{\frac{19}{12}}} & 3^{\frac12}} & \frac{25}{ 2^{\frac{25}{12}} 3^{\frac54} \pi^{\frac12} \end{pmatrix},$$ and Ep(-1)} = -i\frac{ (13 C_1 108 C_3) \pi^{\frac12} s^{\frac23}}{36 \cdot 2^{\frac{11}{12}} 3^{\frac34} C_1^{\frac12}} & * & * \\ i \frac{(35 C_1 - C_3) \pi^{\frac12} 2 | i & 0 & 1
\end{pmatrix} N^{-1}(z),$$ which gives us.
this compleTes thE prOof Of propOsitIon \[prop:p-1\].
We concLUde tHis section by evaluating $e(-1)$ and $E'(-1)$ FoR LateR UsE. The cAlculatIOnS ARe sTrAiGhtFoRWaRd and CumBersome By combininG anD tHe asymptoticS Of $n(z)$ and $f(z)$ givEn iN and. We omit thE deTails bUt PreSEnt thE reSults Below. $$\lABel{eq: E(-1)}
e(-1) = \begin{pmaTrIX}
-i 2^{\frac{1}{12}} 3^{\FRac14} C_1^{\fraC12} \PI^{\fRac12} s^{\Frac23} & - \frac{2^{\frac13}}{3^{\frac12}} &
- \fRAc{1}{2^{\FRac5{12}} 3^{\frac54} C_1^{\frac12} \pi^{\Frac12} s^{\fRaC23} } \\
- I 2^{-\fRAC14} 3^{\frAc14} C_1^{\Frac12} \pi^{\frac12} S^{\fRac23} & \frAC{2}{3^{\frac12}} &
\frAC{5}{ 2^{\fRAC34} 3^{\FraC54} c_1^{\frac12} \pi^{\frac12} s^{\fRac23}} \\
i\frac{3^{\fraC14} c_1^{\frAc12} \pi^{\frAc12} S^{\frAC23} }{
2^{\frac{19}{12}}} & -\fRac{5}{2^{\frAc43} 3^{\FRac12}} & \Frac{25}{
2^{\frac{25}{12}} 3^{\fraC54} C_1^{\frAc12} \pi^{\frac12} s^{\Frac23}}
\enD{Pmatrix},$$ ANd $$\label{Eq: Ep(-1)}
E'(-1) = \bEgiN{pmAtriX}
-I\fRaC{ (13 C_1 + 108 C_3) \Pi^{\FRac12} S^{\FrAc23}}{36 \cDOt 2^{\fRac{11}{12}} 3^{\frac34} C_1^{\FrAc12}} & * & * \\
I \frac{(35 c_1 - 108 C_3) \pi^{\FRAC12} S^{\fraC23}}{72 \cdOt 2 | i & 0 & 1
\end{pmatrix}N^{-1}(z), $$ wh ich gi ve s us .
T his completest he p roof of Proposition \[ prop: p- 1 \].We conc lude th i ss e cti on b y e va l ua ting$E( -1)$ an d $E'(-1)$ fo rlater use. T h ecalculatio nsare straight for ward a nd cu m berso meby co mbinin g and t he asympt ot i cs of$ N(z)$ a n d $ f(z) $ given in and. W e o m it the details but p re s en t the re sults belo w. $$\l a bel{eq: E( - 1 ) }
E(-1) = \begi n{pmatrix} -i 2^ {\ fra c {1}{12 }} 3^ {\ f rac 14} C_1^{\f rac1 2} \pi^{\ frac12 } s^{\f r ac23} & - \fr ac{ 2^{ \fra c 13 }} {3^ {\ f rac 1 2} } & - \frac {1 }{ 2^{\f rac5 { 1 2 } } 3^ {\f rac5 4} C_ 1^{\frac12} \ pi^ {\fr a c12 } s^{ \frac 23}}\\
- i 2 ^{-\f ra c14} 3^{\frac14 } C_ 1^{\frac1 2}\p i^{ \f rac12 } s^{\f rac 23} & \fra c{2}{3^ { \fr ac 1 2 } } &
\frac{5}{ 2^ {\ f r ac 34} 3^{\ frac54 } C _1 ^ {\frac12 }\pi ^{\f r a c12}s^{\ f ra c23}} \\
i\ f ra c{ 3^{\fra c1 4} C_1 ^{ \fr ac1 2} \p i ^{\f rac12} s^{\fra c23}} {
2^{\frac{1 9 }{12}}} & -\f r ac { 5 }{ 2 ^{\f rac 43} 3^{\fra c12} } & \ frac { 25 }{ 2^{\ frac{ 25 } {1 2 }} 3^{\frac54} C_1 ^{ \frac1 2} \p i^{\frac12} s ^{\frac23} } \end{pm atri x }, $ $ and $$\label {eq:Ep(-1)}
E '(-1) =\begi n{pmatri x}
-i \ f rac{ (1 3 C _1+ 1 08C _ 3) \pi^{\frac12 } s^{\ fr ac23}}{ 36\cdot 2 ^{\ fra c{1 1}{ 12 }} 3^{\fr ac34} C _1 ^{ \f ra c12 }} &* & * \\
i\fr ac{(3 5 C_1 - 108C_3) \p i ^{\ frac12} s^ { \ frac 23 }} {72\cd ot 2 | i &_0 &_1
\end{pmatrix} N^{-1}(z),$$_which gives_us.
This_completes the_proof_of Proposition \[prop:p-1\].
We_conclude this section_by evaluating $E(-1)$ and_$E'(-1)$ for later_use._The calculations are straightforward and cumbersome by combining and the asymptotics of $N(z)$ and_$f(z)$_given in_and._We_omit the details but present_the results below. $$\label{eq: E(-1)}
_ E(-1)_= \begin{pmatrix}
-i 2^{\frac{1}{12}}_3^{\frac14}_C_1^{\frac12} \pi^{\frac12} _s^{\frac23} & - \frac{2^{\frac13}}{3^{\frac12}} &
-_\frac{1}{2^{\frac5{12}} 3^{\frac54} C_1^{\frac12} \pi^{\frac12} s^{\frac23} }_\\
_-_i_2^{-\frac14} 3^{\frac14} C_1^{\frac12} \pi^{\frac12}_s^{\frac23} & \frac{2}{3^{\frac12}} &
_ \frac{5}{ 2^{\frac34} 3^{\frac54} C_1^{\frac12} \pi^{\frac12}_s^{\frac23}} \\
i\frac{3^{\frac14} C_1^{\frac12} \pi^{\frac12}_s^{\frac23} }{
2^{\frac{19}{12}}} & -\frac{5}{2^{\frac43}_3^{\frac12}} & \frac{25}{
2^{\frac{25}{12}}_3^{\frac54} _C_1^{\frac12} \pi^{\frac12} s^{\frac23}}
\end{pmatrix},$$_and $$\label{eq: Ep(-1)}
_ E'(-1)_= \begin{pmatrix}
_ -i\frac{ (13 C_1_+ 108 C_3)_\pi^{\frac12} s^{\frac23}}{36 \cdot 2^{\frac{11}{12}} 3^{\frac34} _C_1^{\frac12}}_& * &__*_\\
_ i_\frac{(35_C_1 -_108_C_3) \pi^{\frac12} s^{\frac23}}{72 \cdot 2 |
wavelength]{} multiplexed channels simultaneously via one single mode fibre and 32 different entangled states are shared between the 28 different pairs of users.
{width="90.00000%"}
The experiment was performed in two stages. In the first stage the QNSP, MU, the 8 users each connected to the QNSP/MU with a single fibre $\sim$10m in length, and the 16 detectors were situated in a single laboratory in the Nano Science and Quantum Information (NSQI) building in Bristol. To demonstrate the stability of our network we recorded data for [18.4]{}hours as shown in Fig \[fig:stability\]. To be able to account for finite key effects with a security parameter of [$10^{-5}$]{}, we computed the private key once every 10 minutes and the figure shows the average secure key generation rate per second in each 10min period for each of the 28 links (see section \[sec:methods\]). The total secure key obtained is shown in extended data Table \[tab:short\]. Users A through H used superconducting nanowire detectors from Photon Spot while Ivan used a combination of one SNSPD and one InGaAs Single Photon Avalanche photoDiode (SPAD). We note that the use of heterogeneous detectors did not significantly impact the key generation rates.
Alice Bob Chloe Dave Feng Gopi Heidi Ivan
------- ------- ------- ------- ------- ------- ------- ------- -------
Alice 31621 9102 6081 15529 39054 7159 979
Bob 23798 24208 83645 42062 8893 10400
Chloe 16480 14931 18735 9495 13155
Dave 42 17170 12544 3999
Feng 19925 10470 593
Gopi 10448 3953
Heidi 1700
: **Total secure key (bits) over long distance links for the city wide metropolitan network demonstration.** We connected 4 locations/users across the city of Bristol as shown in Fig \[fig:source+wdms\] via deployed fibre in a loop back configuration. Two other users were sent signals through fibre spools and the remaining two were connected via short (10m) fibres. The distances | wavelength ] { } multiplexed channels simultaneously via one single mode fibre and 32 unlike embroiled department of state are shared between the 28 unlike pairs of exploiter.
! [ image](Fig04_Stability.pdf){width="90.00000% " }
The experiment was performed in two stagecoach. In the first stage the QNSP, MU, the 8 user each connected to the QNSP / MU with a single fiber $ \sim$10 m in length, and the 16 detectors were situate in a individual laboratory in the Nano Science and Quantum Information (NSQI) building in Bristol. To demonstrate the stability of our network we record data for [ 18.4]{}hours as shown in Fig \[fig: stability\ ]. To be able to account for finite cardinal effects with a security argument of [ $ 10^{-5}$ ] { }, we computed the secret key once every 10 minute and the figure shows the average impregnable key generation rate per second in each 10min period for each of the 28 links (see section \[sec: methods\ ]). The total secure key obtained is shown in extended data Table \[tab: short\ ]. Users A through H use superconducting nanowire detectors from Photon Spot while Ivan used a combination of one SNSPD and one InGaAs Single Photon Avalanche photoDiode (SPAD). We notice that the habit of heterogenous detectors did not significantly impact the key generation rates.
Alice Bob Chloe Dave Feng Gopi Heidi Ivan
------- ------- ------- ------- ------- ------- ------- ------- -------
Alice 31621 9102 6081 15529 39054 7159 979
Bob 23798 24208 83645 42062 8893 10400
Chloe 16480 14931 18735 9495 13155
Dave 42 17170 12544 3999
Feng 19925 10470 593
Gopi 10448 3953
Heidi 1700
: * * Total secure keystone (bits) over long distance links for the city wide metropolitan network demonstration. * * We connect 4 locations / users across the city of Bristol as shown in Fig \[fig: source+wdms\ ] via deployed fibre in a loop back shape. Two early drug user were sent signals through fibre bobbin and the remaining two were connected via short (10 m) fibres. The distances | wavflength]{} multiplexed chankels simultaneously via one sjngle moae fibre and 32 different entaiglee stauvs are shared between the 28 difverent pqirs if users.
{wibti="90.00000%"}
The experiment was perfosmed in two stdgds. In the first stage the QNSP, MU, the 8 users ewch connected jo tht QGSP/MH with a single fibre $\sim$10m in lengfh, and uhe 16 detectors werr situated in a single lablratlry in the Nano Sclence and Qoznttn Informatiov (NSQI) building in Briatol. To demonstrate the stabiligy of our netwotk wf recorded deta fog [18.4]{}hours as shown in Fhg \[fig:stsbility\]. To be sblx to account for finite kxy effects with a sesurity pasajeter of [$10^{-5}$]{}, we compytwd thg prieate jey onde ebery 10 linntes and ths figure shiws the average secirq key generatikn ratq [er second in each 10min period for each mf fhe 28 links (see section \[swc:methods\]). The total sgcure key jbtained is shown in extended data Table \[tab:short\]. Gsers X tkvjjth H used superconducting nanowire detectors frjj Khonon Spot while Ivcn used a combinstlom of one SNSPD and ous JnGaAs Single Photln Avalwnche photoDioqe (SLAD). We note that the use of heterogeneols dwtectors did not sngnificantly impsct tne key generation rates.
Alicf Bob Chloe Dave Fvng Gopi Heidi Ivan
------- ------- ------- ------- ------- ------- ------- ------- -------
Aliee 31621 9102 6081 15529 39054 7159 979
Bob 23798 24208 83645 42062 8893 10400
Chloe 16480 14931 18735 9495 13155
Davc 42 17170 12544 3999
Feng 19925 10470 593
Gipi 10448 3953
Hgisi 1700
: **Tuyal sdcure ktv (blts) ovet long distance linkv fod the city wide meyrjpolitan network demonstratiom.** We connected 4 lofatiois/userr acrpss the city of Bristol as shown jn Fig \[fig:douvce+wdms\] via dqploned sibre in a loop back configuration. Two other userw were sent signalw through fibre spopks and the cemainyng two ware connected via shirt (10m) fibres. The cistances | wavelength]{} multiplexed channels simultaneously via one single and different entangled are shared between users. The experiment was in two stages. the first stage the QNSP, MU, 8 users each connected to the QNSP/MU with a single fibre $\sim$10m in and the 16 detectors were situated in a single laboratory in the Nano and Information building Bristol. To demonstrate the stability of our network we recorded data for [18.4]{}hours as shown in \[fig:stability\]. To be able to account for finite effects with a security of [$10^{-5}$]{}, we computed the key every 10 and figure the average secure generation rate per second in each 10min period for each of the 28 links (see section \[sec:methods\]). total secure is shown extended Table Users A through superconducting nanowire detectors from Photon Spot a combination of one SNSPD and one InGaAs Photon Avalanche (SPAD). We note that the use heterogeneous detectors did not significantly impact the key rates. Alice Bob Chloe Dave Feng Gopi Heidi Ivan ------- ------- ------- ------- ------- ------- ------- Alice 31621 9102 15529 39054 7159 Bob 24208 42062 10400 Chloe 14931 18735 9495 13155 Dave 42 17170 12544 3999 Feng 19925 593 Gopi 10448 3953 Heidi 1700 : **Total secure key long links for the wide metropolitan network demonstration.** connected locations/users across the city as in deployed in loop back configuration. Two users were sent signals through spools and the remaining (10m) fibres. The distances | wavelength]{} multiplexed chanNels simultAneouSly Via OnE sinGle mOde fibre and 32 difFErenT entangled states are shaRed beTwEEn thE 28 DiFfereNt pairs OF uSERs.
{wIdth="90.00000%"}
The Experiment Was PeRformed in two STaGes. In the fiRst Stage the QNSP, mU, tHe 8 userS eAch COnnecTed To the qNSP/MU WIth a siNgle fibre $\SiM$10M in lenGTh, and thE 16 DEtEctoRs were situated in a SInGLe laboratory in The NanO SCIeNCE anD QuAntum InforMaTion (NsqI) buildINg IN bRisTOl. To demonstraTe the stabilITy oF our neTwOrk WE recorDed daTa FOr [18.4]{}hOurs as shown In FiG \[fig:stabiLity\]. To BE able to ACcount fOr finiTe kEy eFfecTS wItH a sEcURitY PaRamETer Of [$10^{-5}$]{}, we compUtEd The prIvatE KEY Once EveRy 10 miNutes And the figure sHowS the AVerAge seCure kEy geNeRatioN rate pEr secOnD in each 10min perioD for Each of the 28 LinKs (See SeCtion \[SEc:methOds\]). the Total seCure key OBtaInED IS sHown in extended data taBLE \[tAb:short\]. USers A tHRoUgH h used supErConDuctING nanoWire DEtEctors frOm PhotON SPoT while IVaN used a CoMbiNatIon of ONe SNsPD and One InGaAS SingLE Photon AvalancHE photoDiode (SPad). WE NOtE That The Use of heteroGeneOUs deTectORs Did NOt sigNificAnTLy IMpact the key generatiOn Rates.
ALice BOb Chloe Dave FeNg Gopi HeidI iVAn
------- ------- ------- ------- ------- ------- ------- ------- -------
Alice 31621 9102 6081 15529 39054 7159 979
BOb 23798 24208 83645 42062 8893 10400
ChLOe 16480 14931 18735 9495 13155
dAve 42 17170 12544 3999
Feng 19925 10470 593
Gopi 10448 3953
HeiDi 1700
: **TotAl secure keY (Bits) over Long dIstance lInks for thE CIty wide mEtrOpoLitAn nETWoRk demonstratiON.** we coNnEcted 4 loCatIons/useRs aCroSs tHe cItY of BristoL as shown In fiG \[fIg:SouRce+wdMS\] via deplOyEd fIbRe iN a looP Back coNfiguRatiOn. twO OthEr users WErE SEnt sIgNaLs thRouGh Fibre SpooLS anD the remAining two WerE ConnEcTeD via shoRt (10m) fibres. The dIsTances | wavelength]{} multiplexedchannels s imult ane ous ly via one single mode f i breand 32 different entan gledst a tesa re shar ed betw e en t he28 d iff er e nt pair s o f users .
{width=" 90. 00000%"}
Th e e xperim en t w a s per for med i n twos tages. In the f ir s t stag e the QN S P ,MU,the 8 users eachc on n ected to the Q NSP/MU w i th a si ngl e fibre $\ si m$10m in leng t h, a n d t h e 16 detector s were situ a ted in asi ngl e labor atory i n th e Nano Scie nceand Quant um Inf o rmation (NSQI)buildi nginBris t ol .Tode m ons t ra tet hestabilit yof ournetw o r k we r eco rded data for [18.4]{} hou rs a s sh own i n Fig \[f ig :stab ility\ ]. To b e able to accou nt f or finite ke yeff ec ts wi t h a se cur ity parame ter of[ $10 ^{ - 5 } $] {}, we computed th ep r iv ate keyonce e v er y1 0 minute sand the f igure sho w sthe aver age se c ur ekey gen er ationra teper seco n d in each10min pe riodf or each of the 28 links (see se c t io n \[s ec: methods\]). The tota l se c ur e k e y obt ained i s s h own in extended dat aTable\[tab :short\]. Use rs A throu g h H used s uper c on d ucting nanowir e det ectors fro m PhotonSpotwhile Iv an used a c ombinati onofone SN S P Dand one InGaA s Sing le Photon Av alanche ph oto Dio de(S PAD). Wenote tha tth eus e o f het e rogeneou sdet ec tor s did not si gnifi cant ly i m pac t the k e yg e nera ti on rat es.
Al ice Bob Ch loe Da ve Feng Gopi He id i Ivan
- --- --- -- - - --- ---- --- ------- ------- --- - --- --- --- - --- ---- ------- Alice 31621 910 2 6 0 81 1 55 29 39054 71 5 9 979
B ob 237 9 8 24208 8364 5 420 6 2 88 9 3 1 04 0 0
Chloe 16480 1 493 1 18735 9495 1 3 155
Dave 42 17170 1 2544 3 999
F eng 199 25 10470 593
Gopi 10448 3 953
Heid i 1 7 0 0
: * *To t al sec urekey (bits) over l o ng distance li nksf o r t hec itywi de metropolita n n et w o rk demon st ration.** W e connec te d 4 lo cation s/user s acros s th e cityof B ris tol as sh own i n Fig \[ fi g: s ource+ wdms \] via d eploye d fib r e in a loop backconfi g u ratio n . T wo ot he r users were sent sign als through fibre spo ols a nd there mainin g t wo were conn e cted viashort (10m)fi bres . T he dis tanc e s | wavelength]{} multiplexed_channels simultaneously_via one single mode_fibre and_32_different entangled_states_are shared between_the 28 different_pairs of users.
{width="90.00000%"}
The experiment_was performed in_two_stages. In the first stage the QNSP, MU, the 8 users each connected to_the_QNSP/MU with_a_single_fibre $\sim$10m in length, and_the 16 detectors were situated_in a_single laboratory in the Nano Science and Quantum_Information_(NSQI) building in_Bristol. To demonstrate the stability of our network we_recorded data for [18.4]{}hours as shown_in Fig \[fig:stability\]. To_be_able_to account for finite_key effects with a security parameter_of [$10^{-5}$]{}, we computed the private_key once every 10 minutes and the_figure shows the average secure key_generation rate per second in_each 10min_period for each of the_28 links (see_section \[sec:methods\]). The_total secure key_obtained is shown in extended data_Table \[tab:short\]. Users A_through H used superconducting nanowire detectors_from_Photon Spot while_Ivan_used_a combination_of one SNSPD_and_one InGaAs_Single_Photon Avalanche photoDiode (SPAD). We note_that_the use of heterogeneous detectors did not_significantly impact the key_generation_rates.
_ _ Alice _Bob _ _Chloe Dave Feng _Gopi Heidi _ Ivan
-------_------- -------_-------_------- ------- -------_-------_-------
_ Alice _ _ 31621 _9102 _6081_ 15529 _39054 7159 _ 979
__Bob_ _ _ _ 23798 __24208 83645 _42062_ 8893 _10400
__ Chloe _ _ _ _ _16480__ 14931 18735 9495 _ 13155
_ Dave_ _ _ _ _42 17170 _ 12544 _3999
_ Feng ____ _ _ _ _ _ _ __ 19925 _ 10470 593
_ Gopi __ _ _ _ __ _ _ _ _ _ _ _ _ 10448 _3953
_ Heidi ____ _ _ _ __ _ _ _ _ 1700
: **Total_secure key_(bits) over long distance_links for_the_city wide metropolitan network demonstration.** We_connected 4_locations/users across the city of Bristol as shown_in_Fig \[fig:source+wdms\] via deployed fibre in a loop back configuration. Two other users_were_sent_signals through fibre_spools_and the remaining two were connected_via_short_(10m)_fibres._The distances |
would increase estimates of stellar temperature and luminosity, shifting them onto the evolutionary tracks of rapidly evolving higher-mass central stars, and decreasing the estimated PN ages.
[^3]: There are Merrett numbers for 101 of their 253 H II regions, suggesting that the M06 studies of the PNLF are contaminated by small numbers of tiny H II regions in the inner disk.
[^4]: We suspect that confusion of this line with the $\lambda$4358 sky line of Hg I is to blame when observing at moderate dispersion using a fiber spectrograph. This sky line is faint at the Roque de los Muchachos Observatory.
[^5]: It is worth noting that the PNe in all of these galaxies as well as and those in M31 (Paper I), the SMC and LMC (KH12), and the Milky Way (H10) –as well as the Sun and Orion – share the same Ne/O ratios. This indicates that O and Ne are universally enriched at the same rates and, presumably, by the same processes in all of these galaxies, implying that any oxygen enrichment process operating within NGC 300 (and apparently only NGC 300) also proportionally enriches Ne. See the discussion in section 7 of Stas13 for more details.
[^6]: A very instructive simulation of the stellar dynamics during and after this encounter by J. Dubinski and L. Widrow is available at www.youtube.com/watch?v=6eZm3LHlyrs; pericenter = 53 kpc at -2.6 GY.
---
abstract: 'We investigate the effect of the anisotropic spin-spin interaction on the ground state density distribution of the one dimensional spin-1 bosonic gases within a modified Gross-Pitaevskii theory both in the weakly interaction regime and in the Tonks-Girardeau (TG) regime. We find that for ferromagnetic spinor gas the phase separation occurs even for weak anisotropy of the spin-spin interaction, which becomes more and more obvious and the component of $m_F=0$ diminishes as the anisotropy increases. However, no phase separation is found for anti-ferromagnetic spinor gas in both regimes.'
author:
- Yajiang Hao
- Yunbo Zhang
- 'J. Q. Liang'
- Shu Chen
title: 'Phase separation in the trapped spinor gases with anisotropic spin-spin interaction' | would increase estimates of stellar temperature and luminosity, shifting them onto the evolutionary tracks of quickly develop higher - mass cardinal stars, and decrease the estimated PN ages.
[ ^3 ]: There be Merrett numbers for 101 of their 253 H II region, suggesting that the M06 studies of the PNLF are contaminate by small numbers of bantam H II regions in the inner disk.
[ ^4 ]: We suspect that confusion of this line with the $ \lambda$4358 sky line of Hg I is to blame when observing at moderate distribution using a fiber spectrograph. This sky cable is faint at the Roque de los Muchachos Observatory.
[ ^5 ]: It is worth note that the PNe in all of these galaxies equally well as and those in M31 (Paper I), the SMC and LMC (KH12), and the Milky Way (H10) – as well as the Sun and Orion – partake the same Ne / O ratios. This indicates that O and Ne are universally enrich at the same rates and, presumably, by the same processes in all of these galaxies, incriminate that any oxygen enrichment summons operating within NGC 300 (and apparently only NGC 300) also proportionally enriches Ne. See the discussion in section 7 of Stas13 for more detail.
[ ^6 ]: A very instructive simulation of the stellar moral force during and after this confrontation by J. Dubinski and L. Widrow is available at www.youtube.com/watch?v=6eZm3LHlyrs; pericenter = 53 kpc at -2.6 GY.
---
abstract:' We investigate the impression of the anisotropic spin - spin interaction on the flat coat state density distribution of the one dimensional spin-1 bosonic gases within a modified Gross - Pitaevskii hypothesis both in the weakly interaction regime and in the Tonks - Girardeau (TG) regime. We find that for ferromagnetic spinor gas the phase separation occurs even for weak anisotropy of the spin - spin interaction, which becomes more and more obvious and the component of $ m_F=0 $ decrease as the anisotropy increases. However, no phase interval is found for anti - ferromagnetic spinor gas in both regimes.'
author:
- Yajiang Hao
- Yunbo Zhang
-' J. Q. Liang'
- Shu Chen
title:' Phase interval in the trapped spinor gases with anisotropic spin - tailspin interaction' | wokld increase estimates on stellar tempercrure aid lumihosity, sfifting them onto the evolutmonaey trqcks of rapidly evolvivg higher-lass cenrral wtars, and vscreasiky the cstimctxd PN ages.
[^3]: Therg are Merretd numbers for 101 ow cheir 253 H II regions, suggesting that tre M06 stidles of the PNLS art cjntajpncted by small numbers of tiny H JI regimns in the inmer disk.
[^4]: We suspect that clnfudion of this line aith the $\lanbda$4358 wky line of Fg I is to blame when ogserving at moderate dispersion usiny a fiber skzxtrlcraph. This wky lpne is faint at the Rmque de los Muchachos Obvercatory.
[^5]: It is worth noving that the PNe in all of tveae galaxies as weol as atd tvose un O31 (Pzpxr I), the SLC end LMC (KH12), znd the Miljy Way (H10) –as well as trv Sun and Orikn – shwrq the same Ne/O ratios. This indicates thdt K and Ne are universallt enriched at the samg rates anq, presumably, by the same processes in all of thesa galexkes, iiooylng that any oxygen enrichment process operatyhg wpthin NGC 300 (and appcrently only NGC 300) apsp proportionaluy enrndhss Ne. See the disckssion yn sextion 7 of Stax13 for more details.
[^6]: A very ibstructive spmulqtion of the stellcr dynamics buring and sfter this encounter by J. Dhbinski and L. Widrow ks available at dww.joutgbe.com/wauzh?v=6eZm3LHlyrs; perisenter = 53 jpc ct -2.6 GY.
---
acstrsct: 'We investigahe thc effect of the anidotrokic sphn-spin intfraction on the ground state deixity distribotimn mf the oue dimcnsional spin-1 bjsonic gases wnthin a iodifked Gross-Pptaevskii theory botr in the weakni interaction regime wnd un tye Tonkr-Eirardeau (TG) rrgime. We ynnd that dor ferromagnetic xpivkr gas the phasz weparation occuts dveg voc weah anisotropy mf tfe rlin-spkn interactlon, whivh becomes more and kore obvious and the cpmionent of $m_F=0$ dimigishes as the anisotropy increades. Hmweter, no phwse separation is found for anfi-ferromahnebic spinor gaf in both regimex.'
author:
- Yajiang Hao
- Yunbo Zhang
- 'J. Q. Liaig'
- Shu Chen
title: 'Phase weparation in the ttapked spinor gaxes wyth anisodropic spin-spin inteeaction' | would increase estimates of stellar temperature and them the evolutionary of rapidly evolving the PN ages. [^3]: are Merrett numbers 101 of their 253 H II suggesting that the M06 studies of the PNLF are contaminated by small numbers tiny H II regions in the inner disk. [^4]: We suspect that confusion this with $\lambda$4358 line of Hg I is to blame when observing at moderate dispersion using a fiber spectrograph. sky line is faint at the Roque de Muchachos Observatory. [^5]: It worth noting that the PNe all these galaxies well and in M31 (Paper the SMC and LMC (KH12), and the Milky Way (H10) –as well as the Sun and Orion share the ratios. This that and are universally enriched same rates and, presumably, by the all of these galaxies, implying that any oxygen process operating NGC 300 (and apparently only NGC also proportionally enriches Ne. See the discussion in 7 of Stas13 for more details. [^6]: A very instructive simulation of the stellar dynamics after this encounter by Dubinski and L. is at pericenter 53 kpc -2.6 GY. --- abstract: 'We investigate the effect of the anisotropic interaction on the ground state density distribution of the one bosonic within a modified theory both in the interaction and in the Tonks-Girardeau We that gas phase occurs even for weak of the spin-spin interaction, which more and more obvious diminishes as the anisotropy increases. However, no phase is found for anti-ferromagnetic spinor gas in regimes.' author: - Yajiang Hao - Yunbo Zhang - 'J. Q. Liang' Shu Chen separation in the trapped spinor gases with anisotropic interaction' | would increase estimates of sTellar tempEratuRe aNd lUmInosIty, sHifting them ontO The eVolutionary tracks of rapIdly eVoLVing HIgHer-maSs centrAL sTARs, aNd DeCreAsINg The esTimAted PN aGes.
[^3]: There arE MeRrEtt numbers foR 101 Of Their 253 H II reGioNs, suggesting ThaT the M06 sTuDieS Of the pNLf are cOntamiNAted by Small numbErS Of tiny h iI regioNS In The iNner disk.
[^4]: We suspect THaT Confusion of thiS line wItH ThE $\LAmbDa$4358 sKy line of Hg i iS to blAMe when oBSeRVINg aT Moderate dispeRsion using a FIbeR spectRoGraPH. This sKy linE iS FaiNt at the RoquE de lOs MuchachOs ObseRVatory.
[^5]: IT Is worth Noting ThaT thE PNe IN aLl Of tHeSE gaLAxIes AS weLl as and tHoSe In M31 (PaPer I), THE smC anD LMc (KH12), aNd the milky Way (H10) –as weLl aS the sUn aNd OriOn – shaRe thE sAme Ne/o ratioS. This InDicates that O and ne arE universaLly EnRicHeD at thE Same raTes And, PresumaBly, by thE SamE pROCEsSes in all of these galAxIES, iMplying tHat any OXyGeN EnrichmeNt ProCess OPEratiNg wiTHiN NGC 300 (and aPparenTLy OnLy NGC 300) alSo ProporTiOnaLly EnricHEs Ne. see the DiscussiOn in sECtion 7 of Stas13 for MOre details.
[^6]: A veRY iNSTrUCtivE siMulation of tHe stELlar DynaMIcS duRIng anD afteR tHIs ENcounter by J. Dubinski AnD L. WidrOw is aVailable at www.Youtube.com/WATCh?v=6eZm3LHLyrs; PErICenter = 53 kpc at -2.6 GY.
---
aBstraCt: 'We investIGate the eFfect Of the aniSotropic sPIN-spin intEraCtiOn oN thE GRoUnd state densiTY DistRiBution oF thE one dimEnsIonAl sPin-1 BoSonic gaseS within a MoDiFiEd groSs-PitAEvskii thEoRy bOtH in The weAKly intEractIon rEgImE And In the ToNKs-gIRardEaU (Tg) regIme. we Find tHat fOR feRromagnEtic spinoR gaS The pHaSe SeparatIon occurs even FoR weak anisoTrOpy Of the sPIN-spin intEraction, which becomes morE And more ObvIous aNd thE componenT of $M_F=0$ dimiNisHEs as thE anisoTropy InCreASEs. HowEVEr, No pHaSe separatiON Is fOund fOr Anti-FerromaGnetic spinor gas in bOTh rEgimes.'
author:
- YAjiAng HAO
- yuNbo zHaNG
- 'J. Q. liANg'
- SHU chen
title: 'Phase sEparation iN tHE tRapped spinOR gaSeS with anIsotropIc spiN-Spin intEraction' | would increase estimatesof stellar temp era tur eandlumi nosity, shifti n g th em onto the evolutiona ry tr ac k s of ra pidly evolvi n gh i ghe r- ma ssce n tr al st ars , and d ecreasingthe e stimated PNa ge s.
[^3]:The re are Merre ttnumber sfor 101 o f t heir253 HI I regi ons, sugg es t ing th a t the M 0 6 s tudi es of the PNLF ar e c o ntaminated bysmallnu m be r s of ti ny H II re gi ons i n the in n er d i sk.
[^4]: We sus pect that c o nfu sion o fthi s linewithth e $\ lambda$4358 sky line ofHg I i s to bla m e whenobserv ing at mod e ra te di sp e rsi o nusi n g a fiber s pe ct rogra ph.T h i s sky li ne i s fai nt at the Roq uede l o s M uchac hos O bser va tory.
[^5] : Itis worth noting t hatthe PNe i n a ll of t heseg alaxie s a s w ell asand tho s e i nM 3 1 ( Paper I), the SMCan d LM C (KH12) , andt he M i lky Way(H 10) –as w ell a s th e S un and O rion – sh ar e the s am e Ne/O r ati os. This indi catesthat O a nd Ne are universall y enriched att he s am e rat esand, presum ably , bythes am e p r ocess es in a l lo f these galaxies, i mp lyingthatany oxygen en richment p r o c ess oper atin g w i thin NGC 300 ( and a pparentlyo nly NGC300)also pro portional l y enriche s N e.See th e di scussion in s e c tion 7 of Sta s13 for mo redet ail s.
[ ^6]: A ve ry instr uc ti ve s imu latio n of thest ell ar dy namic s durin g and aft er t h isencount e rb y J.Du bi nski an dL. Wi drow isavailab le at www .yo u tube .c om /watch? v=6eZm3LHlyrs ;pericenter = 53 kpc a t -2.6 GY.
---
abstract: 'We inv e stigate th e eff ectof the an iso tropic sp i n-spin inter actio nont h e gro u n dsta te density d i s tri butio nof t he onedimensional spin-1 bos onic gases wi thi n am o di fie d G r oss -P i tae v s kii theory both in the we ak l yinteractio n re gi me andin theTonks - Girarde au (TG) r egime. We f indt h atfor ferrom agneticspinor ga s thep ha se se par ationoc cur s eve n forw eak anis otropy o f thespin- sp in inter action, which becomes m ore an d mor e o bvious an d t h e c omponentof $ m_F=0$ dim ini she s asthe aniso trop y i ncr e ases. How e ver, no p h as e s e p ar ation is fo u n d fo r ant i-f e rromag neti c spinor gas in b o th regimes.'
a utho r :
-Yaj i angHa o
- Yunbo Zhan g
- ' J . Q. Lian g'
- Shu Chen
title:'P h ase s eparat ion in the tr a p pe d spino r ga ses with ani sot ro p ic spin -s pi n inter acti on ' | would_increase estimates_of stellar temperature and_luminosity, shifting_them_onto the_evolutionary_tracks of rapidly_evolving higher-mass central_stars, and decreasing the_estimated PN ages.
[^3]:_There_are Merrett numbers for 101 of their 253 H II regions, suggesting that the M06_studies_of the_PNLF_are_contaminated by small numbers of_tiny H II regions in the_inner disk.
[^4]:_We suspect that confusion of this line with_the_$\lambda$4358 sky line_of Hg I is to blame when observing at moderate_dispersion using a fiber spectrograph. This_sky line is_faint_at_the Roque de los_Muchachos Observatory.
[^5]: It is worth noting_that the PNe in all of_these galaxies as well as and those_in M31 (Paper I), the SMC and_LMC (KH12), and the Milky_Way (H10)_–as well as the Sun_and Orion –_share the_same Ne/O ratios._This indicates that O and Ne_are universally enriched_at the same rates and, presumably,_by_the same processes_in_all_of these_galaxies, implying that_any_oxygen enrichment_process_operating within NGC 300 (and apparently only_NGC 300)_also proportionally enriches Ne. See the discussion_in section 7 of_Stas13_for more details.
[^6]: A_very instructive simulation of the_stellar dynamics during and after this_encounter by_J. Dubinski_and L. Widrow is available at www.youtube.com/watch?v=6eZm3LHlyrs; pericenter = 53 kpc_at -2.6 GY.
---
abstract: 'We investigate_the effect of the_anisotropic spin-spin_interaction_on the ground_state_density distribution_of the one dimensional spin-1 bosonic gases_within a_modified Gross-Pitaevskii theory both in the_weakly interaction regime and_in_the Tonks-Girardeau (TG) regime. We find_that for ferromagnetic spinor gas the_phase separation occurs even for_weak_anisotropy_of the spin-spin interaction, which_becomes more and more obvious and_the component of_$m_F=0$ diminishes as the anisotropy increases. However,_no_phase separation is found for anti-ferromagnetic_spinor_gas in both regimes.'
author:
- Yajiang Hao
-_Yunbo_Zhang
-_'J. Q. Liang'
- Shu Chen
title:_'Phase separation in the trapped spinor_gases with anisotropic spin-spin interaction' |
}^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right)
+\frac{2\xi s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr
{\gamma}^{(n)}=&-\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2+s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)},\end{aligned}$$ and $$\begin{aligned}
\tilde{{\alpha}}^{(0)}\,{}_{s}B_{\ell m}^{(1)}(\xi)+\tilde{{\beta}}^{(0)}\,{}_{s}B_{\ell m}^{(0)}(\xi)&=0, \cr
\tilde{{\alpha}}^{(n)}\,{}_{s}B_{\ell m}^{(n+1)}(\xi)+\tilde{{\beta}}^{(n)}\,{}_{s}B_{\ell m}^{(n)}
(\xi)+\tilde{{\gamma}}^{(n)}\,{}_{s}B_{\ell m}^{(n-1)}(\xi)&=0, \quad (n\ge 1)
\label{eq:3termElm2}\end{aligned}$$ with $$\begin{aligned}
\tilde{{\alpha}}^{(n)}=&-\frac{4\xi(n+\alpha+1)(n+\beta+1)(n+(\alpha+\beta)/2+1+s)}{(2n+\alpha+\beta+2)
(2n+\alpha+\beta+3)},\cr
\tilde{{\beta}}^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right)
+\frac{2\xi s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr
\tilde{{\gamma}}^{(n)}=&\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2-s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)}.\end{aligned}$$ Note that, for deriving Eq. (\[eq:3termElm\]) and Eq. | } ^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right)
+ \frac{2\xi s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr
{ \gamma}^{(n)}=&-\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2+s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)},\end{aligned}$$ and $ $ \begin{aligned }
\tilde{{\alpha}}^{(0)}\,{}_{s}B_{\ell m}^{(1)}(\xi)+\tilde{{\beta}}^{(0)}\,{}_{s}B_{\ell m}^{(0)}(\xi)&=0, \cr
\tilde{{\alpha}}^{(n)}\,{}_{s}B_{\ell m}^{(n+1)}(\xi)+\tilde{{\beta}}^{(n)}\,{}_{s}B_{\ell m}^{(n) }
(\xi)+\tilde{{\gamma}}^{(n)}\,{}_{s}B_{\ell m}^{(n-1)}(\xi)&=0, \quad (n\ge 1)
\label{eq:3termElm2}\end{aligned}$$ with $ $ \begin{aligned }
\tilde{{\alpha}}^{(n)}=&-\frac{4\xi(n+\alpha+1)(n+\beta+1)(n+(\alpha+\beta)/2 + 1+s)}{(2n+\alpha+\beta+2)
(2n+\alpha+\beta+3)},\cr
\tilde{{\beta}}^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right)
+ \frac{2\xi s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr
\tilde{{\gamma}}^{(n)}=&\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2 - s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)}.\end{aligned}$$ Note that, for deriving Eq. (\[eq:3termElm\ ]) and Eq. | }^{(n)}=&\,_{s}E_{\fll m}(\xi)+\xi^2-\left(n+\frac{\alpha+\btta}{2}\right)\left(n+\frac{\copha+\beva}{2}+1\right)
+\rrac{2\xi s(\xlpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2i+\alpya+\betq+2)},\cr
{\gamma}^{(n)}=&-\frac{4\xi n(n+\alphx+\beta)(n+(\alpja+\beta)/2+s)}{(2n+\qlphe+\beta-1)(2n+\alpha+\beta)},\eis{aligned}$$ and $$\gcgin{anmgned}
\tilde{{\alpha}}^{(0)}\,{}_{x}B_{\ell m}^{(1)}(\xi)+\thlde{{\beta}}^{(0)}\,{}_{s}B_{\ell k}^{(0)}(\xk)&=0, \er
\tilde{{\alpha}}^{(n)}\,{}_{s}B_{\ell m}^{(n+1)}(\xi)+\tilde{{\beta}}^{(n)}\,{}_{s}B_{\ejl m}^{(n)}
(\xi)+\yipde{{\gamma}}^{(n)}\,{}_{s}B_{\ell m}^{(n-1)}(\xp)&=0, \zuad (n\ge 1)
\label{eq:3termElm2}\end{aligned}$$ with $$\begin{anigned}
\tilde{{\allha}}^{(n)}=&-\frac{4\xi(n+\alpha+1)(n+\beta+1)(n+(\alphw+\betw)/2+1+s)}{(2n+\alpha+\beta+2)
(2n+\alpha+\heta+3)},\cr
\tilde{{\veta}}^{(g)}=&\,_{w}E_{\ell m}(\xi)+\xi^2-\ldft(n+\frac{\alkhc+\beta}{2}\right)\lgft(n+\frac{\alpha+\beta}{2}+1\right)
+\frac{2\xi s(\aloha-\beca)(\alpha+\beta)}{(2b+\aophw+\teta)(2n+\alpha+\bxta+2)},\cr
\tplde{{\gamma}}^{(n)}=&\frag{4\qi n(n+\al[ha+\beta)(m+(\alpha+\beta)/2-s)}{(2n+\aliha+\beva-1)(2n+\aopha+\beta)}.\end{aligned}$$ Nove that, for deriving Eq. (\[eq:3termAlj\]) and Eq. | }^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right) +\frac{2\xi s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr {\gamma}^{(n)}=&-\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2+s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)},\end{aligned}$$ and m}^{(1)}(\xi)+\tilde{{\beta}}^{(0)}\,{}_{s}B_{\ell \cr \tilde{{\alpha}}^{(n)}\,{}_{s}B_{\ell m}^{(n)} (\xi)+\tilde{{\gamma}}^{(n)}\,{}_{s}B_{\ell m}^{(n-1)}(\xi)&=0, $$\begin{aligned} (2n+\alpha+\beta+3)},\cr \tilde{{\beta}}^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right) s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr \tilde{{\gamma}}^{(n)}=&\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2-s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)}.\end{aligned}$$ that, for deriving Eq. (\[eq:3termElm\]) and | }^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\Beta}{2}\right)\lEft(n+\fRac{\AlpHa+\Beta}{2}+1\RighT)
+\frac{2\xi s(\alpha-\bETa)(\alPha+\beta)}{(2n+\alpha+\beta)(2n+\alphA+\beta+2)},\Cr
{\GAmma}^{(N)}=&-\FrAc{4\xi n(N+\alpha+\bETa)(N+(\ALphA+\bEtA)/2+s)}{(2n+\AlPHa+\Beta-1)(2n+\AlpHa+\beta)},\eNd{aligned}$$ aNd $$\bEgIn{aligned}
\tilDE{{\aLpha}}^{(0)}\,{}_{s}B_{\ell m}^{(1)}(\Xi)+\tIlde{{\beta}}^{(0)}\,{}_{s}B_{\elL m}^{(0)}(\xI)&=0, \cr
\tilDe{{\AlpHA}}^{(n)}\,{}_{s}B_{\eLl m}^{(N+1)}(\xi)+\tiLde{{\betA}}^{(N)}\,{}_{s}B_{\ell M}^{(n)}
(\xi)+\tilde{{\GaMMa}}^{(n)}\,{}_{s}B_{\eLL m}^{(n-1)}(\xi)&=0, \quAD (N\gE 1)
\labEl{eq:3termElm2}\end{aliGNeD}$$ With $$\begin{alignEd}
\tildE{{\aLPhA}}^{(N)}=&-\FraC{4\xi(N+\alpha+1)(n+\betA+1)(n+(\Alpha+\BEta)/2+1+s)}{(2n+\alPHa+\BETA+2)
(2n+\aLPha+\beta+3)},\cr
\tildE{{\beta}}^{(n)}=&\,_{s}E_{\ell M}(\Xi)+\xI^2-\left(n+\FrAc{\aLPha+\betA}{2}\righT)\lEFt(n+\Frac{\alpha+\beTa}{2}+1\riGht)
+\frac{2\xi S(\alpha-\BEta)(\alphA+\Beta)}{(2n+\alPha+\betA)(2n+\aLphA+\betA+2)},\Cr
\TiLde{{\GaMMa}}^{(n)}=&\FRaC{4\xi N(N+\alPha+\beta)(n+(\AlPhA+\beta)/2-S)}{(2n+\alPHA+\BEta-1)(2n+\AlpHa+\beTa)}.\end{Aligned}$$ Note thAt, fOr deRIviNg Eq. (\[eQ:3termelm\]) aNd eq. | }^{(n)}=&\,_{s}E_{\ell m}( \xi)+\xi^2 -\lef t(n +\f ra c{\a lpha +\beta}{2}\rig h t)\l eft(n+\frac{\alpha+\be ta}{2 }+ 1 \rig h t)
+\fr ac{2\xi s( \ a lph a- \b eta )( \ al pha+\ bet a)}{(2n +\alpha+\b eta )( 2n+\alpha+\b e ta +2)},\cr
{ \ga mma}^{(n)}=& -\f rac{4\ xi n( n +\alp ha+ \beta )(n+(\ a lpha+\ beta)/2+s )} { (2n+\a l pha+\be t a -1 )(2n +\alpha+\beta)},\ e nd { aligned}$$ and $$\be gi n {a l i gne d}\tilde{{\a lp ha}}^ { (0)}\,{ } _{ s } B _{\ e ll m}^{(1)}(\ xi)+\tilde{ { \be ta}}^{ (0 )}\ , {}_{s} B_{\e ll m}^ {(0)}(\xi)& =0,\cr
\tild e{{\al p ha}}^{( n )}\,{}_ {s}B_{ \el l m }^{( n +1 )} (\x i) + \ti l de {{\ b eta }}^{(n)} \, {} _{s}B _{\e l l m }^{( n)}
(\x i)+\t ilde{{\gamma} }^{ (n)} \ ,{} _{s}B _{\el l m} ^{ (n-1) }(\xi) &=0,\q uad (n\ge 1)
\l abel {eq:3term Elm 2} \en d{ align e d}$$ w ith $$ \begin{ aligned }
\t il d e { {\ alpha}}^{(n)}=&-\f ra c { 4\ xi(n+\al pha+1) ( n+ \b e ta+1)(n+ (\ alp ha+\ b e ta)/2 +1+s ) }{ (2n+\alp ha+\be t a+ 2)
(2n+\a lp ha+\be ta +3) },\ cr
\t i lde{ {\beta }}^{(n)} =&\,_ { s}E_{\ell m}(\ x i)+\xi^2-\lef t (n + \ fr a c{\a lph a+\beta}{2} \rig h t)\l eft( n +\ fra c {\alp ha+\b et a }{ 2 }+1\right)
+\frac{2 \x i s(\a lpha- \beta)(\alpha +\beta)}{( 2 n + \alpha+\ beta ) (2 n +\alpha+\beta+ 2)},\ cr
\tilde{ { \gamma}} ^{(n) }=&\frac {4\xi n(n + \ alpha+\b eta )(n +(\ alp h a +\ beta)/2-s)}{( 2 n +\al ph a+\beta -1) (2n+\al pha +\b eta )}. \e nd{aligne d}$$ Not eth at ,for deri v ing Eq.(\ [eq :3 ter mElm\ ] ) andEq. | }^{(n)}=&\,_{s}E_{\ell m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right)
+\frac{2\xi_s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr
{\gamma}^{(n)}=&-\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2+s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)},\end{aligned}$$_and $$\begin{aligned}
\tilde{{\alpha}}^{(0)}\,{}_{s}B_{\ell m}^{(1)}(\xi)+\tilde{{\beta}}^{(0)}\,{}_{s}B_{\ell m}^{(0)}(\xi)&=0,_\cr
\tilde{{\alpha}}^{(n)}\,{}_{s}B_{\ell m}^{(n+1)}(\xi)+\tilde{{\beta}}^{(n)}\,{}_{s}B_{\ell_m}^{(n)}
(\xi)+\tilde{{\gamma}}^{(n)}\,{}_{s}B_{\ell_m}^{(n-1)}(\xi)&=0, \quad_(n\ge_1)
\label{eq:3termElm2}\end{aligned}$$ with $$\begin{aligned}
\tilde{{\alpha}}^{(n)}=&-\frac{4\xi(n+\alpha+1)(n+\beta+1)(n+(\alpha+\beta)/2+1+s)}{(2n+\alpha+\beta+2)
(2n+\alpha+\beta+3)},\cr
\tilde{{\beta}}^{(n)}=&\,_{s}E_{\ell_m}(\xi)+\xi^2-\left(n+\frac{\alpha+\beta}{2}\right)\left(n+\frac{\alpha+\beta}{2}+1\right)
+\frac{2\xi s(\alpha-\beta)(\alpha+\beta)}{(2n+\alpha+\beta)(2n+\alpha+\beta+2)},\cr
\tilde{{\gamma}}^{(n)}=&\frac{4\xi n(n+\alpha+\beta)(n+(\alpha+\beta)/2-s)}{(2n+\alpha+\beta-1)(2n+\alpha+\beta)}.\end{aligned}$$_Note that, for deriving_Eq. (\[eq:3termElm\]) and Eq. |
_{-\infty}^{+\infty}\!\!\!\!\!\!\!\exp\!\Big[\frac{({\bar{v}}+\alpha\Gamma-y)^2+({\bar{u}}-\alpha\Sigma-y')^2}{-4\eta t}\Big]\tilde{g}_0\;dydy'\;,\nonumber\end{aligned}$$ where
\[eq.11\] $$\begin{aligned}
&\Gamma=k_x\sin(\omega_ct)+k_y\cos(\omega_ct)\;,\\
&\Sigma=k_y\sin(\omega_ct)-k_x\cos(\omega_ct)\;,\\
&\alpha=2i\eta/\omega_c^2\;.\end{aligned}$$
The complete solution must rely on the explicit form of the initial condition $\tilde{g}_0(k_x,k_y,{\bar{u}},{\bar{v}}$). Since we are interested in determining the density threshold for the electron deconfinement, a suitable choice for the initial condition is the density profile corresponding to a quasi-stable configuration of the plasma. This way, we can investigate the proper conditions for which the model exhibits a disruption. In the real space, an example of such an initial profile is shown in Ref. and can be reasonably approximated with a Gaussian of both space and velocity variables. Thus, the function $\tilde{g}_0$ reads as $$\label{eq.12}
\tilde{g}_0(k_x,k_y,{\bar{u}},{\bar{v}})=\frac{m}{2\pi k_BT}\frac{1}{2\pi \sigma^2}
\;e^{-\frac{m{\bar{u}}^{2}+m{\bar{v}}^{2}}{2 k_B T}-\frac{k_x^2+k_y^2}{2\sigma^2}}\;,$$ where $\sigma$ is the variance in the $k$-space.
The physical phenomena we are going to describe concern the inner part of the confined plasma as shown in Ref. . Therefore, we reasonably set $1/\sigma$ equal to one third of the minor radius of the FTU Tokamak device, [*i.e.,* ]{}$\sigma=8.33$m$^{-1}$. Moreover, being focused on the role played by the plasma density, the temperature is here and in the following considered as a fixed parameter. However, it is | _ { -\infty}^{+\infty}\!\!\!\!\!\!\!\exp\!\Big[\frac{({\bar{v}}+\alpha\Gamma - y)^2+({\bar{u}}-\alpha\Sigma - y')^2}{-4\eta t}\Big]\tilde{g}_0\;dydy'\;,\nonumber\end{aligned}$$ where
\[eq.11\ ] $ $ \begin{aligned }
& \Gamma = k_x\sin(\omega_ct)+k_y\cos(\omega_ct)\;,\\
& \Sigma = k_y\sin(\omega_ct)-k_x\cos(\omega_ct)\;,\\
& \alpha=2i\eta/\omega_c^2\;.\end{aligned}$$
The complete solution must rely on the explicit form of the initial circumstance $ \tilde{g}_0(k_x, k_y,{\bar{u}},{\bar{v}}$). Since we are concerned in determine the density brink for the electron deconfinement, a desirable choice for the initial condition is the density visibility represent to a quasi - stable shape of the plasma. This direction, we can investigate the proper conditions for which the mannequin exhibits a disruption. In the substantial quad, an example of such an initial profile is shown in Ref. and can be reasonably approximated with a Gaussian of both outer space and velocity variables. Thus, the routine $ \tilde{g}_0 $ reads as $ $ \label{eq.12 }
\tilde{g}_0(k_x, k_y,{\bar{u}},{\bar{v}})=\frac{m}{2\pi k_BT}\frac{1}{2\pi \sigma^2 }
\;e^{-\frac{m{\bar{u}}^{2}+m{\bar{v}}^{2}}{2 k_B T}-\frac{k_x^2+k_y^2}{2\sigma^2}}\;,$$ where $ \sigma$ is the variance in the $ k$-space.
The physical phenomena we are die to describe concern the inner part of the confined plasma as prove in Ref. . Therefore, we reasonably set $ 1/\sigma$ equal to one third of the minor spoke of the FTU Tokamak device, [ * i.e., * ] { } $ \sigma=8.33$m$^{-1}$. furthermore, being focus on the role played by the plasma density, the temperature is here and in the following considered as a fixed argument. However, it is | _{-\infhy}^{+\infty}\!\!\!\!\!\!\!\exp\!\Big[\frac{({\bar{v}}+\alpma\Gamma-y)^2+({\bar{u}}-\alphc\Wigma-y')^2}{-4\xta t}\Bif]\tilde{g}_0\;dhdy'\;,\nonumber\end{aligned}$$ where
\[ee.11\] $$\vegin{qligned}
&\Gamma=k_x\sin(\omega_zt)+k_y\cos(\omvga_ct)\;,\\
&\Sigmq=k_y\smn(\omega_ct)-k_x\cos(\omxfa_ct)\;,\\
&\alpmc=2i\eta/\klega_e^2\;.\eid{aligned}$$
The cokplete solgtion must relf un the explicit form of the initial cjnditiom $\hilde{g}_0(k_x,k_y,{\bar{u}},{\far{v}}$). Fincs we are interested in determining the deisity threshold for the electron deconfinfmenh, a suitable choicf for the ibitiwo condition ks the density profile corresponding to a quasi-stable confnguration od rhe klasma. This xay, we can investinste tha propet conditions fpr xhicy the model exhibits e disruption. In the teal space, au example of such an unutial proxile us rhosn ih Ref. wnd can be rezsonably approximated with a Gsufwian of both apace wnq velocity variables. Thus, the function $\uilde{f}_0$ reads as $$\label{eq.12}
\tilde{t}_0(k_x,k_y,{\bar{u}},{\bar{v}})=\frac{m}{2\pi n_BT}\frac{1}{2\pi \sigma^2}
\;e^{-\frac{m{\bar{u}}^{2}+m{\bar{v}}^{2}}{2 k_B T}-\frac{k_x^2+k_y^2}{2\sigma^2}}\;,$$ where $\sigka$ is ghe vwfuajce in the $k$-space.
The physical phenomena we arq gpikg to describe cjncern the onjet part of the zonfinzs llasma as shown in Ref. . Thgreforw, we reasjnabky set $1/\sigma$ equal to one tyird of the iunor radius of the FTU Tokamay defice, [*o.e.,* ]{}$\sigma=8.33$m$^{-1}$. Moreover, beiny focuaed on the gole playsa by the plasma aenxidy, the ttoperature is here and in tie foklowing conxidereq as a fixfd pavdmeter. However, it ls | _{-\infty}^{+\infty}\!\!\!\!\!\!\!\exp\!\Big[\frac{({\bar{v}}+\alpha\Gamma-y)^2+({\bar{u}}-\alpha\Sigma-y')^2}{-4\eta t}\Big]\tilde{g}_0\;dydy'\;,\nonumber\end{aligned}$$ where \[eq.11\] $$\begin{aligned} &\Gamma=k_x\sin(\omega_ct)+k_y\cos(\omega_ct)\;,\\ &\Sigma=k_y\sin(\omega_ct)-k_x\cos(\omega_ct)\;,\\ complete must rely the explicit form Since are interested in the density threshold the electron deconfinement, a suitable choice the initial condition is the density profile corresponding to a quasi-stable configuration of plasma. This way, we can investigate the proper conditions for which the model a In real an example of such an initial profile is shown in Ref. and can be reasonably approximated a Gaussian of both space and velocity variables. the function $\tilde{g}_0$ reads $$\label{eq.12} \tilde{g}_0(k_x,k_y,{\bar{u}},{\bar{v}})=\frac{m}{2\pi k_BT}\frac{1}{2\pi \sigma^2} \;e^{-\frac{m{\bar{u}}^{2}+m{\bar{v}}^{2}}{2 T}-\frac{k_x^2+k_y^2}{2\sigma^2}}\;,$$ $\sigma$ is variance the The physical phenomena are going to describe concern the inner part of the confined plasma as shown in Ref. . we reasonably equal to third the radius of the device, [*i.e.,* ]{}$\sigma=8.33$m$^{-1}$. Moreover, being focused played by the plasma density, the temperature is and in following considered as a fixed parameter. it is | _{-\infty}^{+\infty}\!\!\!\!\!\!\!\exp\!\Big[\frac{({\bar{v}}+\aLpha\Gamma-y)^2+({\Bar{u}}-\aLphA\SiGmA-y')^2}{-4\etA t}\BiG]\tilde{g}_0\;dydy'\;,\nonUMber\End{aligned}$$ where
\[eq.11\] $$\begin{AlignEd}
&\gAmma=K_X\sIn(\omeGa_ct)+k_y\cOS(\oMEGa_cT)\;,\\
&\SIgMa=k_Y\sIN(\oMega_cT)-k_x\Cos(\omegA_ct)\;,\\
&\alpha=2i\eTa/\oMeGa_c^2\;.\end{aligneD}$$
thE complete sOluTion must rely On tHe explIcIt fORm of tHe iNitiaL condiTIon $\tilDe{g}_0(k_x,k_y,{\baR{u}},{\BAr{v}}$). SinCE we are iNTErEsteD in determining the DEnSIty threshold foR the elEcTRoN DEcoNfiNement, a suiTaBle chOIce for tHE iNITIal COndition is the Density profILe cOrrespOnDinG To a quaSi-staBlE ConFiguration oF the Plasma. ThiS way, we CAn invesTIgate thE propeR coNdiTionS FoR wHicH tHE moDEl ExhIBitS a disrupTiOn. in the Real SPACE, an eXamPle oF such An initial profIle Is shOWn iN Ref. aNd can Be reAsOnablY approXimatEd With a Gaussian of Both Space and vEloCiTy vArIableS. thus, thE fuNctIon $\tildE{g}_0$ reads AS $$\laBeL{EQ.12}
\TiLde{g}_0(k_x,k_y,{\bar{u}},{\bar{v}})=\frAc{M}{2\PI k_bT}\frac{1}{2\pi \Sigma^2}
\;e^{-\FRaC{m{\BAr{u}}^{2}+m{\bar{v}}^{2}}{2 K_B t}-\frAc{k_x^2+K_Y^2}{2\Sigma^2}}\;,$$ WherE $\SiGma$ is the VarianCE iN tHe $k$-spacE.
THe physIcAl pHenOmena WE are Going tO describE concERn the inner part OF the confined pLAsMA As SHown In REf. . Therefore, We reASonaBly sET $1/\sIgmA$ Equal To one ThIRd OF the minor radius of thE FtU TokaMak deVice, [*i.e.,* ]{}$\sigma=8.33$m$^{-1}$. MOreover, beiNG FOcused on The rOLe PLayed by the plasMa denSity, the temPErature iS here And in the Following CONsidered As a FixEd pAraMETeR. However, it is | _{-\infty}^{+\infty}\!\!\! \!\!\!\!\e xp\!\ Big [\f ra c{({ \bar {v}}+\alpha\Ga m ma-y )^2+({\bar{u}}-\alpha\ Sigma -y ' )^2} { -4 \etat}\Big] \ ti l d e{g }_ 0\ ;dy dy ' \; ,\non umb er\end{ aligned}$$ wh er e
\[eq.11\] $$ \begin{ali gne d}
&\Gamma=k _x\ sin(\o me ga_ c t)+k_ y\c os(\o mega_c t )\;,\\
&\Sigma= k_ y \sin(\ o mega_ct ) - k_ x\co s(\omega_ct)\;,\\ &\ a lpha=2i\eta/\o mega_c ^2 \ ;. \ e nd{ ali gned}$$
T he comp l ete sol u ti o n mus t rely on theexplicit fo r m o f thein iti a l cond ition $ \ til de{g}_0(k_x ,k_y ,{\bar{u} },{\ba r {v}}$). Since w e areint ere sted in d ete rm i nin g t hed ens ity thre sh ol d for the e l e ctro n d econ finem ent, a suitab lechoi c e f or th e ini tial c ondit ion is thede nsity profile c orre spondingtoaqua si -stab l e conf igu rat ion ofthe pla s ma. T h i s w ay, we can investi ga t e t he prope r cond i ti on s for whi ch th e mo d e l exh ibit s a disrupt ion. I n t he real s pa ce, an e xam ple of s u ch a n init ial prof ile i s shown in Ref. and can be r e as o n ab l y ap pro ximated wit h aG auss iano fbot h spac e and v e lo c ity variables. Thus ,the fu nctio n $\tilde{g}_ 0$ reads a s $ $\label{ eq.1 2 }\ tilde{g}_0(k_x ,k_y, {\bar{u}}, { \bar{v}} )=\fr ac{m}{2\ pi k_BT}\ f r ac{1}{2\ pi\si gma ^2} \ ;e ^{-\frac{m{\b a r {u}} ^{ 2}+m{\b ar{ v}}^{2} }{2 k_ B T }-\ fr ac{k_x^2+ k_y^2}{2 \s ig ma ^2 }}\ ;,$$w here $\s ig ma$ i s t he va r iancein th e $k $- sp a ce.
The p h ys i c al p he no mena we a re go ingt o d escribe concernthe inne rpa rt of t he confined p la sma as sho wn in Ref.. Therefor e, we reasonably set $1 / \sigma$ eq ual t o on e third o f t he min orr adiusof the FTUTo kam a k devi c e ,[*i .e .,* ]{}$\s i g ma= 8.33$ m$ ^{-1 }$. Mor eover, being focus e d o n the role pl aye d by t he pl a sm a de ns i ty, t he temperatureis here an di nthe follow i ngco nsidere d as afixed paramet er. Howev er, it is | _{-\infty}^{+\infty}\!\!\!\!\!\!\!\exp\!\Big[\frac{({\bar{v}}+\alpha\Gamma-y)^2+({\bar{u}}-\alpha\Sigma-y')^2}{-4\eta t}\Big]\tilde{g}_0\;dydy'\;,\nonumber\end{aligned}$$_where
\[eq.11\] $$\begin{aligned}
&\Gamma=k_x\sin(\omega_ct)+k_y\cos(\omega_ct)\;,\\
&\Sigma=k_y\sin(\omega_ct)-k_x\cos(\omega_ct)\;,\\
&\alpha=2i\eta/\omega_c^2\;.\end{aligned}$$
The_complete solution must rely_on the_explicit_form of_the_initial condition $\tilde{g}_0(k_x,k_y,{\bar{u}},{\bar{v}}$)._Since we are_interested in determining the_density threshold for_the_electron deconfinement, a suitable choice for the initial condition is the density profile corresponding_to_a quasi-stable_configuration_of_the plasma. This way, we_can investigate the proper conditions_for which_the model exhibits a disruption. In the real_space,_an example of_such an initial profile is shown in Ref. and_can be reasonably approximated with a_Gaussian of both_space_and_velocity variables. Thus, the_function $\tilde{g}_0$ reads as $$\label{eq.12}
\tilde{g}_0(k_x,k_y,{\bar{u}},{\bar{v}})=\frac{m}{2\pi k_BT}\frac{1}{2\pi_\sigma^2}
\;e^{-\frac{m{\bar{u}}^{2}+m{\bar{v}}^{2}}{2 k_B T}-\frac{k_x^2+k_y^2}{2\sigma^2}}\;,$$ where $\sigma$ is_the variance in the $k$-space.
The physical phenomena_we are going to describe concern_the inner part of the_confined plasma_as shown in Ref. . Therefore,_we reasonably set_$1/\sigma$ equal_to one third_of the minor radius of the_FTU Tokamak device,_[*i.e.,* ]{}$\sigma=8.33$m$^{-1}$. Moreover, being focused on_the_role played by_the_plasma_density, the_temperature is here_and_in the_following_considered as a fixed parameter. However,_it_is |
{\beta}_p)^T\sim N((\beta_1,\cdots,\beta_p)^T,{\mbox{\boldmath $\Sigma$}}^*)$, where the $(k,l)$th element in ${\mbox{\boldmath $\Sigma$}}^*$ is $\displaystyle{\mbox{\boldmath $\Sigma$}}_{kl}^*=\sigma^2s_{kl}/(ns_{kk}s_{ll})$.]{}
For ease of notation, let $Z_1,\cdots,Z_p$ be the standardized random variables of $\widehat{\beta}_1,\cdots,\widehat{\beta}_p$, that is, $$\label{b2}
Z_i=\frac{\widehat{\beta}_i}{{\mbox{SD}}(\widehat{\beta}_i)}=\frac{\widehat{\beta}_i}{\sigma/(\sqrt{n}s_{ii})}, \quad\quad i=1,\cdots,p.$$ In the above, we implicitly assume that $\sigma$ is known and the above standardized random variables are z-test statistics. The estimate of residual variance $\sigma^2$ will be discussed in Section 6 via refitted cross-validation (Fan, Guo & Hao, 2011). Then, conditioning on $\{X_j^i\}$, $$\label{c1}
(Z_1,\cdots,Z_p)^T\sim N((\mu_1,\cdots,\mu_p)^T,{\mbox{\boldmath $\Sigma$}}),$$ where $\mu_i=\sqrt{n}\beta_is_{ii}/\sigma$ and covariance matrix ${\mbox{\boldmath $\Sigma$}}$ has the $(k,l)$th element as $s_{kl}$. Simultaneously testing (2) based on $(\widehat{\beta}_1,\cdots,\widehat{\beta}_p)^T$ is thus equivalent to testing
$$\label{c2}
H_{0j}:\quad\mu_j=0\quad\text{vs}\quad H_{1j}:\quad \mu_j\neq0, \quad\quad j=1,\cdots,p$$
based on $(Z_1,\cdots,Z_p)^T$.
In (4), ${\mbox{\boldmath $\Sigma$}}$ is the population covariance matrix of $(Z_1,\cdots,Z_p)^T$, and is known based on the sample data $\{X_j^i\}$. The covariance matrix ${\mbox{\boldmath $\Sigma$}}$ can have arbitrary dependence structure. We would | { \beta}_p)^T\sim N((\beta_1,\cdots,\beta_p)^T,{\mbox{\boldmath $ \Sigma$}}^*)$, where the $ (k, l)$th element in $ { \mbox{\boldmath $ \Sigma$}}^*$ is $ \displaystyle{\mbox{\boldmath $ \Sigma$}}_{kl}^*=\sigma^2s_{kl}/(ns_{kk}s_{ll})$. ] { }
For ease of notation, let $ Z_1,\cdots, Z_p$ be the exchangeable random variable star of $ \widehat{\beta}_1,\cdots,\widehat{\beta}_p$, that is, $ $ \label{b2 }
Z_i=\frac{\widehat{\beta}_i}{{\mbox{SD}}(\widehat{\beta}_i)}=\frac{\widehat{\beta}_i}{\sigma/(\sqrt{n}s_{ii }) }, \quad\quad i=1,\cdots, p.$$ In the above, we implicitly assume that $ \sigma$ is known and the above exchangeable random variables are omega - test statistics. The appraisal of residual variance $ \sigma^2 $ will be discussed in Section 6 via refit crabbed - validation (Fan, Guo & Hao, 2011). Then, conditioning on $ \{X_j^i\}$, $ $ \label{c1 }
(Z_1,\cdots, Z_p)^T\sim N((\mu_1,\cdots,\mu_p)^T,{\mbox{\boldmath $ \Sigma$}}),$$ where $ \mu_i=\sqrt{n}\beta_is_{ii}/\sigma$ and covariance matrix $ { \mbox{\boldmath $ \Sigma$}}$ suffer the $ (k, l)$th element as $ s_{kl}$. Simultaneously examination (2) based on $ (\widehat{\beta}_1,\cdots,\widehat{\beta}_p)^T$ is thus equivalent to testing
$ $ \label{c2 }
H_{0j}:\quad\mu_j=0\quad\text{vs}\quad H_{1j}:\quad \mu_j\neq0, \quad\quad j=1,\cdots, p$$
based on $ (Z_1,\cdots, Z_p)^T$.
In (4), $ { \mbox{\boldmath $ \Sigma$}}$ is the population covariance matrix of $ (Z_1,\cdots, Z_p)^T$, and is known based on the sample datum $ \{X_j^i\}$. The covariance matrix $ { \mbox{\boldmath $ \Sigma$}}$ can have arbitrary dependence structure. We would | {\betw}_p)^T\sim N((\beta_1,\cdots,\beta_p)^T,{\mnox{\boldmath $\Sigmc$}}^*)$, where the $(k,m)$th elemdnt in ${\mbox{\boldmath $\Sigma$}}^*$ is $\dusplatstyle{\mbox{\boldmath $\Sigoa$}}_{kl}^*=\sigma^2d_{kl}/(ns_{kk}s_{ol})$.]{}
Foc ease of notation, let $Z_1,\gbots,Z_l$ be chx standardized tandom variatles of $\widehad{\bdtc}_1,\cdots,\widehat{\beta}_p$, that is, $$\label{b2}
Z_i=\fwac{\widenah{\beta}_i}{{\mbox{SD}}(\wiqehau{\beea}_i)}=\fdac{\widehat{\beta}_i}{\sigma/(\sqrt{n}s_{ii})}, \quad\qhad i=1,\cdmts,p.$$ In the anove, we implicitly assume hhat $\sigma$ is known anf the above stageardized ranaom variables are z-tesj statistics. The estimate of reskdual variance $\witma^2$ fill be disrussed in Section 6 via refhtted ctoss-validation (Fai, Gui & Hao, 2011). Then, conditioiing on $\{X_j^i\}$, $$\label{c1}
(Z_1,\cqots,Z_p)^T\sik U((\mu_1,\cdots,\mu_p)^T,{\mbox{\boldmqty $\Sigka$}}),$$ wvere $\nu_i=\rqru{n}\bxta_js_{ii}/\sihma$ and covarjance matriz ${\mbox{\boldmath $\Sigms$}}$ rqs the $(k,l)$th emement af $s_{kl}$. Simultaneously testing (2) based on $(\fidshat{\beta}_1,\cdots,\widehat{\betq}_p)^T$ is thus equivalenj to testigg
$$\label{c2}
H_{0j}:\quad\mu_j=0\quad\text{vs}\quad H_{1j}:\quad \mu_j\neq0, \quag\quad g=1,\cditf,o$$
vaded on $(Z_1,\cdots,Z_p)^T$.
In (4), ${\mbox{\boldmath $\Sigma$}}$ is the lokulstion covariange matrix of $(Z_1,\cdotx,Z_o)^T$, wnd is known cased on fhe sample data $\{X_j^l\}$. The cjvariqnce matryx ${\mnox{\boldmath $\Sigma$}}$ can have qrbitrary deiendwnce structure. We could | {\beta}_p)^T\sim N((\beta_1,\cdots,\beta_p)^T,{\mbox{\boldmath $\Sigma$}}^*)$, where the $(k,l)$th element $\Sigma$}}^*$ $\displaystyle{\mbox{\boldmath $\Sigma$}}_{kl}^*=\sigma^2s_{kl}/(ns_{kk}s_{ll})$.]{} ease of notation, random of $\widehat{\beta}_1,\cdots,\widehat{\beta}_p$, that $$\label{b2} Z_i=\frac{\widehat{\beta}_i}{{\mbox{SD}}(\widehat{\beta}_i)}=\frac{\widehat{\beta}_i}{\sigma/(\sqrt{n}s_{ii})}, \quad\quad In the above, we implicitly assume $\sigma$ is known and the above standardized random variables are z-test statistics. The of residual variance $\sigma^2$ will be discussed in Section 6 via refitted cross-validation Guo Hao, Then, on $\{X_j^i\}$, $$\label{c1} (Z_1,\cdots,Z_p)^T\sim N((\mu_1,\cdots,\mu_p)^T,{\mbox{\boldmath $\Sigma$}}),$$ where $\mu_i=\sqrt{n}\beta_is_{ii}/\sigma$ and covariance matrix ${\mbox{\boldmath $\Sigma$}}$ has the $(k,l)$th as $s_{kl}$. Simultaneously testing (2) based on $(\widehat{\beta}_1,\cdots,\widehat{\beta}_p)^T$ thus equivalent to testing H_{0j}:\quad\mu_j=0\quad\text{vs}\quad H_{1j}:\quad \mu_j\neq0, \quad\quad j=1,\cdots,p$$ on In (4), $\Sigma$}}$ the covariance matrix of and is known based on the sample data $\{X_j^i\}$. The covariance matrix ${\mbox{\boldmath $\Sigma$}}$ can have arbitrary structure. We | {\beta}_p)^T\sim N((\beta_1,\cdots,\beta_p)^T,{\Mbox{\boldmaTh $\SigMa$}}^*)$, wHerE tHe $(k,l)$Th elEment in ${\mbox{\bolDMath $\sigma$}}^*$ is $\displaystyle{\mboX{\boldMaTH $\SigMA$}}_{kL}^*=\sigmA^2s_{kl}/(ns_{kK}S_{lL})$.]{}
fOr eAsE oF noTaTIoN, let $Z_1,\CdoTs,Z_p$ be tHe standardIzeD rAndom variablES oF $\widehat{\beTa}_1,\cDots,\widehat{\bEta}_P$, that iS, $$\lAbeL{B2}
Z_i=\frAc{\wIdehaT{\beta}_i}{{\MBox{SD}}(\wIdehat{\betA}_i)}=\FRac{\widEHat{\beta}_I}{\SIgMa/(\sqRt{n}s_{ii})}, \quad\quad i=1,\cdOTs,P.$$ in the above, we imPlicitLy ASsUME thAt $\sIgma$ is knowN aNd the ABove staNDaRDIZed RAndom variableS are z-test stATisTics. ThE eStiMAte of rEsiduAl VAriAnce $\sigma^2$ wiLl be Discussed In SectIOn 6 via reFItted crOss-valIdaTioN (Fan, gUo & haO, 2011). ThEn, COndITiOniNG on $\{x_j^i\}$, $$\label{C1}
(Z_1,\CdOts,Z_p)^t\sim n((\MU_1,\CDots,\Mu_p)^t,{\mboX{\boldMath $\Sigma$}}),$$ wherE $\mu_I=\sqrT{N}\beTa_is_{iI}/\sigmA$ and CoVariaNce matRix ${\mbOx{\Boldmath $\Sigma$}}$ haS the $(K,l)$th elemeNt aS $s_{Kl}$. SImUltanEOusly tEstIng (2) Based on $(\Widehat{\BEta}_1,\CdOTS,\WiDehat{\beta}_p)^T$ is thus eQuIVAlEnt to tesTing
$$\laBEl{C2}
H_{0J}:\Quad\mu_j=0\qUaD\teXt{vs}\QUAd H_{1j}:\qUad \mU_J\nEq0, \quad\quAd j=1,\cdoTS,p$$
BaSed on $(Z_1,\cDoTs,Z_p)^T$.
IN (4), ${\mBox{\BolDmath $\sIgma$}}$ Is the pOpulatioN covaRIance matrix of $(Z_1,\CDots,Z_p)^T$, and is kNOwN BAsED on tHe sAmple data $\{X_j^I\}$. The COvarIancE MaTriX ${\Mbox{\bOldmaTh $\sIgMA$}}$ can have arbitrary dePeNdence StrucTure. We would | {\beta}_p)^T\sim N((\beta_ 1,\cdots,\ beta_ p)^ T,{ \m box{ \bol dmath $\Sigma$ } }^*) $, where the $(k,l)$th elem en t in$ {\ mbox{ \boldma t h$ \ Sig ma $} }^* $i s$\dis pla ystyle{ \mbox{\bol dma th $\Sigma$}}_ { kl }^*=\sigma ^2s _{kl}/(ns_{k k}s _{ll}) $. ]{}
Foreas e ofnotati o n, let $Z_1,\cd ot s ,Z_p$b e the s t a nd ardi zed random variab l es of $\widehat{\ beta}_ 1, \ cd o t s,\ wid ehat{\beta }_ p$, t h at is,$ $\ l a b el{ b 2}
Z_i=\frac{ \widehat{\b e ta} _i}{{\ mb ox{ S D}}(\w ideha t{ \ bet a}_i)}=\fra c{\w idehat{\b eta}_i } {\sigma / (\sqrt{ n}s_{i i}) },\qua d \q ua d i =1 , \cd o ts ,p. $ $ I n the ab ov e, we i mpli c i t l y as sum e th at $\ sigma$ is kno wnandt heabove stan dard iz ed ra ndom v ariab le s are z-test st atis tics. The es ti mat eof re s idualvar ian ce $\si gma^2$w ill b e d is cussed in Section6v i arefitted cross - va li d ation (F an , G uo & H ao, 2 011) . T hen, con dition i ng o n $\{X_ j^ i\}$,$$ \la bel {c1}( Z_1, \cdots ,Z_p)^T\ sim N ( (\mu_1,\cdots, \ mu_p)^T,{\mbo x {\ b o ld m ath$\S igma$}}),$$ whe r e $\ mu_i = \s qrt { n}\be ta_is _{ i i} / \sigma$ and covaria nc e matr ix ${ \mbox{\boldma th $\Sigma $ } } $ has th e $( k ,l ) $th element as $s_{ kl}$. Simu l taneousl y tes ting (2) based on $ (\wideha t{\ bet a}_ 1,\ c d ot s,\widehat{\b e t a}_p )^ T$ is t hus equiva len t t o t est in g
$$\lab el{c2}
H _{ 0j }: \q uad \mu_j = 0\quad\t ex t{v s} \qu ad H_ { 1j}:\q uad \ mu_j \n eq 0 , \ quad\qu a dj = 1,\c do ts ,p$$
b as ed on $(Z _ 1,\ cdots,Z _p)^T$.
In( 4),${ \m box{\bo ldmath $\Sigm a$ }}$ is the p opu lation c ovarianc e matrix of $(Z_1,\cdot s ,Z_p)^T $,and i s kn own based on the s amp l e data $\{X_ j^i\} $. Th e covar i a nc e m at rix ${\mbo x { \bo ldmat h$\Si gma$}}$ can have arbitrar y de pendence stru ctu re.W e w oul d | {\beta}_p)^T\sim N((\beta_1,\cdots,\beta_p)^T,{\mbox{\boldmath_$\Sigma$}}^*)$, where_the $(k,l)$th element in_${\mbox{\boldmath $\Sigma$}}^*$_is_$\displaystyle{\mbox{\boldmath $\Sigma$}}_{kl}^*=\sigma^2s_{kl}/(ns_{kk}s_{ll})$.]{}
For_ease_of notation, let_$Z_1,\cdots,Z_p$ be the_standardized random variables of_$\widehat{\beta}_1,\cdots,\widehat{\beta}_p$, that is,_$$\label{b2}
Z_i=\frac{\widehat{\beta}_i}{{\mbox{SD}}(\widehat{\beta}_i)}=\frac{\widehat{\beta}_i}{\sigma/(\sqrt{n}s_{ii})},_\quad\quad i=1,\cdots,p.$$ In the above, we implicitly assume that $\sigma$ is known and the_above_standardized random_variables_are_z-test statistics. The estimate of_residual variance $\sigma^2$ will be_discussed in_Section 6 via refitted cross-validation (Fan, Guo &_Hao,_2011). Then, conditioning_on $\{X_j^i\}$, $$\label{c1}
(Z_1,\cdots,Z_p)^T\sim N((\mu_1,\cdots,\mu_p)^T,{\mbox{\boldmath $\Sigma$}}),$$ where $\mu_i=\sqrt{n}\beta_is_{ii}/\sigma$ and covariance_matrix ${\mbox{\boldmath $\Sigma$}}$ has the $(k,l)$th_element as $s_{kl}$._Simultaneously_testing_(2) based on $(\widehat{\beta}_1,\cdots,\widehat{\beta}_p)^T$_is thus equivalent to testing
$$\label{c2}
H_{0j}:\quad\mu_j=0\quad\text{vs}\quad H_{1j}:\quad_\mu_j\neq0, \quad\quad j=1,\cdots,p$$
based on $(Z_1,\cdots,Z_p)^T$.
In (4),_${\mbox{\boldmath $\Sigma$}}$ is the population covariance matrix_of $(Z_1,\cdots,Z_p)^T$, and is known based_on the sample data $\{X_j^i\}$._The covariance_matrix ${\mbox{\boldmath $\Sigma$}}$ can have_arbitrary dependence structure._We would |
\]) at $z=0.1$ (upper panel), $z=0.2$ (middle panel) and $z=0.5$ (lower panel) in a ${\Omega_{\rm M}}=0.28,\,{\Omega_{\rm\Lambda}}=0.72$ universe, which is what we will use subsequently. We have performed a number of simulations using a wide range of cosmological parameter-values and found the oscillation effect to depend only weakly on cosmology. The most dramatic effect is the strong variation of attenuation with photon energy.
Since the attenuation varies very rapidly with photon energy in a similar manner over a broad energy range, we expect the frequency dependence to wash out to large extent when doing broad-band photometry.
In Fig. \[fig:n\], we show the rest-frame $B$-band magnitude attenuation for Type Ia SNe due to photon-axion oscillations for three different values of the electron density, in the redshift interval $0<z<2$, using values for the other input parameters from Eq. \[eq:values\]. Each point represents the average value and the error bars the dispersion for ten different lines-of-sight. In the upper panel, we have used $n_{\rm e}=10^{-7}\,{\rm cm}^{-3}(1+z)^3$, in the middle panel $n_{\rm e}=5\times 10^{-8}\,{\rm cm}^{-3}(1+z)^3$ and in the lower panel $n_{\rm e}=10^{-8}\,{\rm cm}^{-3}(1+z)^3$. Note that the effect is not necessarily increasing with increasing redshift. This is due to the fact that we are studying the [*rest-frame*]{} $B$-band magnitude. Since the amplitude of the oscillations scales roughly as $(\omega /n_{\rm e})^2$ (see Eq. \[eq:rho11approx\]), we need this combination to be large at some point in order to get close to maximal mixing. If the plasma density is high, the photon energy will be redshifted to too low energies before the plasma density is diluted due to the expansion.
In Fig. \[fig:b\], the rest-frame $B$-band magnitude attenuation for Type Ia SNe for three different values of the intergalactic field strength is shown. Again, each point represents the average value and the error bars the dispersion for ten different lines-of-sight. In the upper panel, $B_{\ | \ ]) at $ z=0.1 $ (upper panel), $ z=0.2 $ (middle panel) and $ z=0.5 $ (lower dialog box) in a $ { \Omega_{\rm M}}=0.28,\,{\Omega_{\rm\Lambda}}=0.72 $ population, which is what we will use subsequently. We have performed a numeral of simulation using a wide stove of cosmologic parameter - value and found the oscillation impression to depend only weakly on cosmology. The about dramatic effect is the strong variation of attenuation with photon energy.
Since the attenuation varies very rapidly with photon energy in a alike manner over a broad energy range, we have a bun in the oven the frequency dependence to wash out to large extent when doing wide - band photometry.
In Fig. \[fig: n\ ], we show the rest - frame $ B$-band magnitude attenuation for Type Ia SNe due to photon - axion cycle for three different values of the electron density, in the redshift interval $ 0 < z<2 $, using values for the other input parameters from Eq. \[eq: values\ ]. Each point represents the average value and the error bars the dispersion for ten different credit line - of - sight. In the upper panel, we have use $ n_{\rm e}=10^{-7}\,{\rm cm}^{-3}(1+z)^3 $, in the center panel $ n_{\rm e}=5\times 10^{-8}\,{\rm cm}^{-3}(1+z)^3 $ and in the lower dialog box $ n_{\rm e}=10^{-8}\,{\rm cm}^{-3}(1+z)^3$. Note that the effect is not necessarily increasing with increasing red shift. This is due to the fact that we are studying the [ * rest - frame * ] { } $ B$-band magnitude. Since the amplitude of the oscillations scales roughly as $ (\omega /n_{\rm e})^2 $ (see Eq. \[eq: rho11approx\ ]), we need this combination to be large at some point in orderliness to catch close to maximal mixing. If the plasma density is high, the photon energy will be redshifted to excessively low energies before the plasma density is dilute due to the expansion.
In Fig. \[fig: b\ ], the rest - frame $ B$-band magnitude attenuation for Type Ia SNe for three different value of the intergalactic field potency is shown. Again, each point represent the average value and the mistake bars the distribution for ten different lines - of - view. In the upper panel, $ B_{\ | \]) at $z=0.1$ (upper panel), $z=0.2$ (middle kanel) and $z=0.5$ (lower panel) mn a ${\Omsga_{\rm M}}=0.28,\,{\Ooega_{\rm\Lambda}}=0.72$ universe, which ms wyat wt will use subsequevtly. We hwve perfirmev a number of simulations usinf a wndx range of cosmplogical pdrameter-values avd found the oscillation effect to de[end onky weakly on cosiolobr. Ths most dramatic effect is the strohg varietion of attenustion with photon energy.
Sijce hhe attenuation vagies very rqpidjt with photov energy in a similar janner over a broad energy rangd, we zxpect the drwquftcy dependeice to wash out to large eftent wnen doing broac-baid pyotometry.
In Fig. \[fig:n\], wx show the rest-frame $B$-band macnntude attenuation for Ttpe Id SNa dud to phktpn-zxion lscmllations fkr three didferent values of tne vkectron densjty, in tre redshift interval $0<z<2$, using values for ths other input parameterw from Eq. \[eq:values\]. Eacj point rqpresents the average value and the error bars tha dis'efsiin wir ten different lines-of-sight. In the upper panej, wt hsve used $n_{\rm e}=10^{-7}\,{\vm cm}^{-3}(1+z)^3$, in the middke psgel $n_{\rm e}=5\timer 10^{-8}\,{\rm cm}^{-3}(1+z)^3$ znd in the lower pwnel $n_{\ri e}=10^{-8}\,{\rm cm}^{-3}(1+z)^3$. Note thay the effect is not necessaeily increaspng qith increasing rebshift. This ns due to tne fact that we are stubying fhe [*rest-frale*]{} $B$-band jxgnitude. Since tfe sm[litude of the oscillationf scales coughky as $(\ooega /n_{\rm e})^2$ (see Eq. \[eq:rjo11appvmx\]), we need this colbinajion tm be large at some point in order to get rkose to maxikan mpxing. If che plssma density ys high, the phpton enzrgy wkll be redahifted to too low energies befmte the plasma density is eilured due go the expansipn.
In Fig. \[fpg:y\], the resr-frame $B$-band magnibude xftenuation for Cvpt Ua SNe for thrre aifsegenv valtas of the indergxlazyic fkeld strengbh ks snown. Again, each poind relresents the averabe value abd the ewror bars the dispersion for tej difherent linex-of-fight. In the upper panel, $B_{\ | \]) at $z=0.1$ (upper panel), $z=0.2$ (middle $z=0.5$ panel) in ${\Omega_{\rm M}}=0.28,\,{\Omega_{\rm\Lambda}}=0.72$ universe, use We have performed number of simulations a wide range of cosmological parameter-values found the oscillation effect to depend only weakly on cosmology. The most dramatic is the strong variation of attenuation with photon energy. Since the attenuation varies rapidly photon in similar manner over a broad energy range, we expect the frequency dependence to wash out to extent when doing broad-band photometry. In Fig. \[fig:n\], show the rest-frame $B$-band attenuation for Type Ia SNe to oscillations for different of electron density, in redshift interval $0<z<2$, using values for the other input parameters from Eq. \[eq:values\]. Each point represents the value and bars the for different In the upper have used $n_{\rm e}=10^{-7}\,{\rm cm}^{-3}(1+z)^3$, in $n_{\rm e}=5\times 10^{-8}\,{\rm cm}^{-3}(1+z)^3$ and in the lower $n_{\rm e}=10^{-8}\,{\rm Note that the effect is not increasing with increasing redshift. This is due to fact that we are studying the [*rest-frame*]{} $B$-band magnitude. Since the amplitude of the oscillations as $(\omega /n_{\rm e})^2$ Eq. \[eq:rho11approx\]), we this to large some point order to get close to maximal mixing. If the plasma density high, the photon energy will be redshifted to too low the density is diluted to the expansion. In \[fig:b\], rest-frame $B$-band magnitude attenuation Ia for of intergalactic strength is shown. Again, point represents the average value the error bars the In the upper panel, $B_{\ | \]) at $z=0.1$ (upper panel), $z=0.2$ (middle panel) And $z=0.5$ (lower pAnel) iN a ${\OMegA_{\rM M}}=0.28,\,{\OmEga_{\rM\Lambda}}=0.72$ universE, WhicH is what we will use subseqUentlY. WE Have PErFormeD a numbeR Of SIMulAtIoNs uSiNG a Wide rAngE of cosmOlogical paRamEtEr-values and fOUnD the oscillAtiOn effect to dePenD only wEaKly ON cosmOloGy. The Most drAMatic eFfect is thE sTRong vaRIation oF ATtEnuaTion with photon eneRGy.
sInce the attenuaTion vaRiES vERY raPidLy with photOn EnergY In a simiLAr MANNer OVer a broad enerGy range, we exPEct The freQuEncY DependEnce tO wASh oUt to large exTent When doing Broad-bANd photoMEtry.
In FIg. \[fig:n\], We sHow The rESt-FrAme $b$-bANd mAGnItuDE atTenuatioN fOr type IA SNe DUE TO phoTon-AxioN osciLlations for thRee DiffERenT valuEs of tHe elEcTron dEnsity, In the ReDshift interval $0<z<2$, UsinG values foR thE oTheR iNput pARameteRs fRom eq. \[eq:valUes\]. Each POinT rEPREsEnts the average valuE aND ThE error baRs the dISpErSIon for teN dIffErenT LInes-oF-sigHT. IN the uppeR panel, WE hAvE used $n_{\rM e}=10^{-7}\,{\Rm cm}^{-3}(1+z)^3$, iN tHe mIddLe panEL $n_{\rm E}=5\times 10^{-8}\,{\Rm cm}^{-3}(1+z)^3$ and In the LOwer panel $n_{\rm e}=10^{-8}\,{\rM Cm}^{-3}(1+z)^3$. Note that thE EfFECt IS not NecEssarily incReasINg wiTh inCReAsiNG redsHift. THiS Is DUe to the fact that we arE sTudyinG the [*rEst-frame*]{} $B$-band Magnitude. SINCE the amplItudE Of THe oscillations ScaleS roughly as $(\OMega /n_{\rm e})^2$ (See Eq. \[Eq:rho11appRox\]), we need THIs combinAtiOn tO be LarGE At Some point in orDER to gEt Close to MaxImal mixIng. if tHe pLasMa Density is High, the pHoToN eNeRgy Will bE RedshiftEd To tOo Low EnergIEs befoRe the PlasMa DeNSitY is diluTEd DUE to tHe ExPansIon.
in fig. \[fiG:b\], thE ResT-frame $B$-Band magniTudE AtteNuAtIon for TYpe Ia SNe for thReE different VaLueS of the INTergalacTic field strength is shown. aGain, eacH poInt rePresEnts the avEraGe valuE anD The errOr bars The diSpErsION for tEN DiFfeReNt lines-of-sIGHt. IN the uPpEr paNel, $B_{\ | \]) at $z=0.1$ (upper pane l), $z=0.2 $ (mi ddl e p an el)and$z=0.5$ (lower pane l) in a ${\Omega_{\rmM}}=0 .2 8 ,\,{ \ Om ega_{ \rm\Lam b da } } =0. 72 $uni ve r se , whi chis what we will u sesu bsequently.W ehave perfo rme d a number o f s imulat io nsu singa w ide r ange o f cosmo logical p ar a meter- v alues a n d f ound the oscillatione ff e ct to depend o nly we ak l yo n co smo logy. Themo st dr a matic e f fe c t ist he strong var iation of a t ten uation w ith photon ener gy .
S ince the at tenu ation var ies ve r y rapid l y withphoton en erg y in asi mil ar man n er ov e r a broad e ne rg y ran ge,w e e xpec t t he f reque ncy dependenc e t o wa s h o ut to larg e ex te nt wh en doi ng br oa d-band photomet ry.
In Fig.\[f ig :n\ ], we s h ow the re st- frame $ B$-band mag ni t u d eattenuation for Ty pe I aSNe dueto pho t on -a x ion osci ll ati onsf o r thr ee d i ff erent va lues o f t he electr on densi ty , i n t he re d shif t inte rval $0< z<2$, using values f o r the other i n pu t pa r amet ers from Eq. \ [eq: v alue s\]. Ea chp ointrepre se n ts the average value a nd the e rrorbars the disp ersion for t e n differ entl in e s-of-sight. In theupper pane l , we hav e use d $n_{\r m e}=10^{ - 7 }\,{\rmcm} ^{- 3}( 1+z ) ^ 3$ , in the midd l e pan el $n_{\r m e }=5\tim es10^ {-8 }\, {\ rm cm}^{- 3}(1+z)^ 3$ a nd i n t he lo w er panel $ n_{ \r m e }=10^ { -8}\,{ \rm c m}^{ -3 }( 1 +z) ^3$. No t et h at t he e ffec t i snot n eces s ari ly incr easing wi thi ncre as in g redsh ift. This isdu e to the f ac t t hat we a re study ing the [*rest-frame*]{ } $B$-ba ndmagni tude . Since t heamplit ude of the oscil latio ns sc a l es ro u g hl y a s$(\omega / n _ {\r m e}) ^2 $ (s ee Eq.\[eq:rho11approx\] ) , w e need this c omb inat i o ntob el arg ea t s o m e point in orde r to get c lo s eto maximal mix in g. If t he plas ma de n sity is high, th e photonen ergy w ill be redshi fted totoo low e n ergie s b efore th e plas ma de nsity is di l ute d due to th eexpans ion.
I n Fig. \ [fig:b\], the rest-fram e $B$- bandmag nitude at ten u ati on for Ty pe I a SNe forthr eediffe ren t valu es o f t hei nterg alac t ic fields tr eng t h i s shown. Ag a i n , e ach p oin t repre sent s the average val u e and the erro r ba r s th e d i sper si on for ten dif fer en t lines-of -s ight. In th e upperpa n el, $ B_{\ | \]) at_$z=0.1$ (upper_panel), $z=0.2$ (middle panel)_and $z=0.5$_(lower_panel) in_a_${\Omega_{\rm M}}=0.28,\,{\Omega_{\rm\Lambda}}=0.72$ universe,_which is what_we will use subsequently._We have performed_a_number of simulations using a wide range of cosmological parameter-values and found the oscillation_effect_to depend_only_weakly_on cosmology. The most dramatic_effect is the strong variation_of attenuation_with photon energy.
Since the attenuation varies very rapidly_with_photon energy in_a similar manner over a broad energy range, we_expect the frequency dependence to wash_out to large_extent_when_doing broad-band photometry.
In Fig. \[fig:n\],_we show the rest-frame $B$-band magnitude_attenuation for Type Ia SNe due_to photon-axion oscillations for three different values_of the electron density, in the_redshift interval $0<z<2$, using values_for the_other input parameters from Eq. \[eq:values\]._Each point represents_the average_value and the_error bars the dispersion for ten_different lines-of-sight. In_the upper panel, we have used_$n_{\rm_e}=10^{-7}\,{\rm cm}^{-3}(1+z)^3$, in_the_middle_panel $n_{\rm_e}=5\times 10^{-8}\,{\rm cm}^{-3}(1+z)^3$_and_in the_lower_panel $n_{\rm e}=10^{-8}\,{\rm cm}^{-3}(1+z)^3$. Note that_the_effect is not necessarily increasing with increasing_redshift. This is due_to_the fact that we_are studying the [*rest-frame*]{} $B$-band_magnitude. Since the amplitude of the_oscillations scales_roughly as_$(\omega /n_{\rm e})^2$ (see Eq. \[eq:rho11approx\]), we need this combination to be_large at some point in order_to get close to_maximal mixing._If_the plasma density_is_high, the_photon energy will be redshifted to too_low energies_before the plasma density is diluted_due to the expansion.
In_Fig. \[fig:b\],_the rest-frame $B$-band magnitude attenuation for_Type Ia SNe for three different_values of the intergalactic field_strength_is_shown. Again, each point represents_the average value and the error_bars the dispersion_for ten different lines-of-sight. In the upper_panel,_$B_{\ |
$$\sigma= x^{{\frac{n}{2}}}({\rm Id}+S(0))\psi +O(x^{{\frac{n}{2}}+1}).$$
Remark from Proposition \[propofS\] that $S(0)^*=S(0)^{-1}=S(0)$ and so the operator $$\label{defc}
{\mathcal}{C}:={\frac{1}{2}}({\rm Id}+S(0))$$ is an orthogonal projector on a subspace of $L^2({\partial}{\overline}{X},{^0\Sigma})$ for the measure ${\rm dv}_{h_0}$ where $h_0=(x^{2}g)|_{T{\partial}{\overline}{X}}$. Notice from that, under a change of boundary defining function $\hat{x}=e^{\omega}x$, the operator ${\mathcal}{C}$ changes according to conjugation $\hat{{\mathcal}{C}}=e^{-{\frac{n}{2}}\omega_0}{\mathcal}{C}
e^{{\frac{n}{2}}\omega_0}$.
Now we want to prove that the range of $E(0)$ acting on $C^{\infty}({\partial}{\overline}{X},{^0\Sigma})$ is exactly the set of harmonic spinors in $x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$.
\[span\] Let $\phi\in x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$ such that $D_g\phi=0$ and let $\psi:=(x^{-\frac n2}\phi)|_{{\partial}{\overline}{X}}$. Then we have $E(0)\psi=2\phi$.
First let us write $\psi=\psi_++\psi_-$ with $\psi_\pm\in{^0\Sigma}_\pm$. Then we construct the approximate solution $\sigma_{\infty,+}({\lambda})$ of associated to $\psi_+$. Let us set $\phi_+({\lambda}):=\sigma_{\infty,+}({\lambda})$ and $\phi_-({\lambda}):=\phi-\phi_+({\lambda})$. One has $(x^{-{\frac{n}{2}}}\phi_-({0}))|_{x=0}=\psi_-\in{^0\Sigma}_-$ and $D_g\phi_-(0)=-D_g\phi_+(0)$. As in the proof of Proposition \[poisson\], we have $$\sigma_{+ | $ $ \sigma= x^{{\frac{n}{2}}}({\rm Id}+S(0))\psi + O(x^{{\frac{n}{2}}+1}).$$
Remark from Proposition \[propofS\ ] that $ S(0)^*=S(0)^{-1}=S(0)$ and so the operator $ $ \label{defc }
{ \mathcal}{C}:={\frac{1}{2}}({\rm Id}+S(0))$$ is an orthogonal projector on a subspace of $ L^2({\partial}{\overline}{X},{^0\Sigma})$ for the measure $ { \rm dv}_{h_0}$ where $ h_0=(x^{2}g)|_{T{\partial}{\overline}{X}}$. notification from that, under a variety of boundary defining function $ \hat{x}=e^{\omega}x$, the hustler $ { \mathcal}{C}$ changes accord to conjugation $ \hat{{\mathcal}{C}}=e^{-{\frac{n}{2}}\omega_0}{\mathcal}{C }
e^{{\frac{n}{2}}\omega_0}$.
Now we desire to raise that the range of $ E(0)$ dissemble on $ C^{\infty}({\partial}{\overline}{X},{^0\Sigma})$ is exactly the set of consonant spinors in $ x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$.
\[span\ ] Let $ \phi\in x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$ such that $ D_g\phi=0 $ and let $ \psi:=(x^{-\frac n2}\phi)|_{{\partial}{\overline}{X}}$. Then we have $ E(0)\psi=2\phi$.
First lease us write $ \psi=\psi_++\psi_-$ with $ \psi_\pm\in{^0\Sigma}_\pm$. Then we construct the approximate solution $ \sigma_{\infty,+}({\lambda})$ of associated to $ \psi_+$. Let us set $ \phi_+({\lambda}):=\sigma_{\infty,+}({\lambda})$ and $ \phi_-({\lambda}):=\phi-\phi_+({\lambda})$. One accept $ (x^{-{\frac{n}{2}}}\phi_-({0}))|_{x=0}=\psi_-\in{^0\Sigma}_-$ and $ D_g\phi_-(0)=-D_g\phi_+(0)$. As in the proof of Proposition \[poisson\ ], we have $ $ \sigma_{+ | $$\sihma= x^{{\frac{n}{2}}}({\rm Id}+S(0))\psi +O(x^{{\frxc{n}{2}}+1}).$$
Remark from Kripositmon \[prolofS\] thag $S(0)^*=S(0)^{-1}=S(0)$ and so the operator $$\lauel{dwfc}
{\maukcal}{C}:={\frac{1}{2}}({\rm Id}+S(0))$$ is av orthogojal projwctoc on a subspace of $L^2({\partlcl}{\ovedpine}{R},{^0\Smgma})$ for the messure ${\rm de}_{h_0}$ where $h_0=(x^{2}g)|_{T{\pdrgicl}{\overline}{X}}$. Notice from that, under a change ov boundary defyninb fundniin $\hat{x}=e^{\omega}x$, the operator ${\mzthcal}{C}$ changes accotding to conjugation $\hat{{\mahhcap}{C}}=e^{-{\frac{n}{2}}\omega_0}{\mathcwl}{C}
e^{{\frac{n}{2}}\omgfa_0}$.
Njq we want to prove thau che range or $E(0)$ acting on $C^{\infty}({\partial}{\overuine}{X},{^0\Xigma})$ is ezaxtlj the set of harmjnic spinors in $x^{{\fraw{n}{2}}}C^{\inftu}({\overline}{X},{^0\Sigms})$.
\[spen\] Lwt $\phi\in x^{{\frac{n}{2}}}C^{\infty}({\mverline}{X},{^0\Sigma})$ sucr that $D_g\[hn=0$ and let $\psi:=(x^{-\frac n2}\pyi)|_{{\partidl}{\ovarlivw}{X}}$. Gheh xe gave $E(0)\osi=2\'hi$.
First lef us write $\psi=\psi_++\psi_-$ with $\psi_\pk\ig{^0\Wigma}_\pm$. Then se conftwuct the approximate solution $\sigma_{\inftj,+}({\lamgda})$ of associated to $\psu_+$. Let us set $\phi_+({\lambdw}):=\sigma_{\infey,+}({\lambda})$ and $\phi_-({\lambda}):=\phi-\phi_+({\lambda})$. One has $(x^{-{\frac{n}{2}}}\[hi_-({0}))|_{x=0}=\pak_-\in{^0\Winma}_-$ xbd $D_g\phi_-(0)=-D_g\phi_+(0)$. As in the proof of Proposition \[poyaspn\], we have $$\sigma_{+ | $$\sigma= x^{{\frac{n}{2}}}({\rm Id}+S(0))\psi +O(x^{{\frac{n}{2}}+1}).$$ Remark from Proposition $S(0)^*=S(0)^{-1}=S(0)$ so the $$\label{defc} {\mathcal}{C}:={\frac{1}{2}}({\rm Id}+S(0))$$ a of $L^2({\partial}{\overline}{X},{^0\Sigma})$ for measure ${\rm dv}_{h_0}$ $h_0=(x^{2}g)|_{T{\partial}{\overline}{X}}$. Notice from that, under a of boundary defining function $\hat{x}=e^{\omega}x$, the operator ${\mathcal}{C}$ changes according to conjugation $\hat{{\mathcal}{C}}=e^{-{\frac{n}{2}}\omega_0}{\mathcal}{C} Now we want to prove that the range of $E(0)$ acting on $C^{\infty}({\partial}{\overline}{X},{^0\Sigma})$ exactly set harmonic in $x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$. \[span\] Let $\phi\in x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$ such that $D_g\phi=0$ and let $\psi:=(x^{-\frac n2}\phi)|_{{\partial}{\overline}{X}}$. Then we have First let us write $\psi=\psi_++\psi_-$ with $\psi_\pm\in{^0\Sigma}_\pm$. Then construct the approximate solution of associated to $\psi_+$. Let set and $\phi_-({\lambda}):=\phi-\phi_+({\lambda})$. has and As in the of Proposition \[poisson\], we have $$\sigma_{+ | $$\sigma= x^{{\frac{n}{2}}}({\rm Id}+S(0))\psi +O(x^{{\frac{N}{2}}+1}).$$
Remark froM PropOsiTioN \[pRopoFS\] thAt $S(0)^*=S(0)^{-1}=S(0)$ and so the oPEratOr $$\label{defc}
{\mathcal}{C}:={\fraC{1}{2}}({\rm Id}+s(0))$$ iS An orTHoGonal ProjectOR oN A SubSpAcE of $l^2({\pARtIal}{\ovErlIne}{X},{^0\SigMa})$ for the meAsuRe ${\Rm dv}_{h_0}$ where $h_0=(x^{2}G)|_{t{\pArtial}{\overLinE}{X}}$. Notice from ThaT, under A cHanGE of boUndAry deFining FUnctioN $\hat{x}=e^{\omeGa}X$, The opeRAtor ${\matHCAl}{c}$ chaNges according to coNJuGAtion $\hat{{\mathcaL}{C}}=e^{-{\fraC{n}{2}}\OMeGA_0}{\MatHcaL}{C}
e^{{\frac{n}{2}}\omEgA_0}$.
Now wE Want to pROvE THAt tHE range of $E(0)$ actiNg on $C^{\infty}({\pARtiAl}{\overLiNe}{X},{^0\sIgma})$ is ExactLy THe sEt of harmoniC spiNors in $x^{{\frAc{n}{2}}}C^{\inFTy}({\overlINe}{X},{^0\SigmA})$.
\[span\] LEt $\pHi\iN x^{{\frAC{n}{2}}}c^{\iNftY}({\oVErlINe}{x},{^0\SiGMa})$ sUch that $D_G\pHi=0$ And leT $\psi:=(X^{-\FRAC n2}\phI)|_{{\paRtiaL}{\overLine}{X}}$. Then we haVe $E(0)\Psi=2\pHI$.
FiRst leT us wrIte $\pSi=\Psi_++\psI_-$ with $\pSi_\pm\iN{^0\SIgma}_\pm$. Then we conStruCt the apprOxiMaTe sOlUtion $\SIgma_{\inFty,+}({\LamBda})$ of asSociateD To $\pSi_+$. lET Us Set $\phi_+({\lambda}):=\sigma_{\iNfTY,+}({\LaMbda})$ and $\pHi_-({\lambDA}):=\pHi-\PHi_+({\lambda})$. onE haS $(x^{-{\frAC{N}{2}}}\phi_-({0}))|_{x=0}=\Psi_-\iN{^0\siGma}_-$ and $D_g\Phi_-(0)=-D_g\pHI_+(0)$. AS iN the proOf Of PropOsItiOn \[pOissoN\], We haVe $$\sigmA_{+ | $$\sigma= x^{{\frac{n}{2} }}({\rm Id }+S(0 ))\ psi + O(x^ {{\f rac{n}{2}}+1}) . $$
Remark from Propositio n \[p ro p ofS\ ] t hat $ S(0)^*= S (0 ) ^ {-1 }= S( 0)$ a n dso th e o perator $$\label{ def c}
{\mathcal}{ C }: ={\frac{1} {2} }({\rm Id}+S (0) )$$ is a n o r thogo nal proj ectoro n a su bspace of $ L ^2({\p a rtial}{ \ o ve rlin e}{X},{^0\Sigma}) $ f o r the measure${\rmdv } _{ h _ 0}$ wh ere $h_0=( x^ {2}g) | _{T{\pa r ti a l } {\o v erline}{X}}$. Notice fro m th at, un de r a change of b ou n dar y definingfunc tion $\ha t{x}=e ^ {\omega } x$, the opera tor ${ \mat h ca l} {C} $c han g es ac c ord ing to c on ju gatio n $\ h a t { {\ma thc al}{ C}}=e ^{-{\frac{n}{ 2}} \ome g a_0 }{\ma thcal }{C}
e ^{{\f rac{n} {2}}\ om ega_0}$.
Now w e wa nt to pro veth atth e ran g e of $ E(0 )$actingon $C^{ \ inf ty } ( { \p artial}{\overline} {X } , {^ 0\Sigma} )$ ise xa ct l y the se tofharm o n ic sp inor s i n $x^{{\ frac{n } {2 }} }C^{\in ft y}({\o ve rli ne} {X},{ ^ 0\Si gma})$ .
\[spa n\] L e t $\phi\in x^{ { \frac{n}{2}}} C ^{ \ i nf t y}({ \ov erline}{X}, {^0\ S igma })$s uc h t h at $D _g\ph i= 0 $a nd let $\psi:=(x^{- \f rac n2 }\phi )|_{{\partial }{\overlin e } { X}}$. Th en w e h a ve $E(0)\psi=2 \phi$ .
First l e t us wri te $\ psi=\psi _++\psi_- $ with $\p si_ \pm \in {^0 \ S ig ma}_\pm$. The n we c on structthe approx ima tesol uti on $\sigma_ {\infty, +} ({ \l am bda })$ o f associa te d t o$\p si_+$ . Let u s set $\p hi _+ ( {\l ambda}) : =\ s i gma_ {\ in fty, +}( {\ lambd a})$ and $\phi_ -({\lambd a}) : =\ph i- \p hi_+({\ lambda})$. On ehas $(x^{- {\ fra c{n}{2 } } }\phi_-( {0}))|_{x=0}=\psi_-\in{ ^ 0\Sigma }_- $ and $D_ g\phi_-(0 )=- D_g\ph i_+ ( 0)$. A s in t he pr oo f o f Propo s i ti on\[ poisson\], w e h ave $ $\ sigm a_{+ | $$\sigma=_x^{{\frac{n}{2}}}({\rm Id}+S(0))\psi_+O(x^{{\frac{n}{2}}+1}).$$
Remark from Proposition \[propofS\]_that $S(0)^*=S(0)^{-1}=S(0)$_and_so the_operator_$$\label{defc}
{\mathcal}{C}:={\frac{1}{2}}({\rm Id}+S(0))$$ is_an orthogonal projector_on a subspace of_$L^2({\partial}{\overline}{X},{^0\Sigma})$ for the_measure_${\rm dv}_{h_0}$ where $h_0=(x^{2}g)|_{T{\partial}{\overline}{X}}$. Notice from that, under a change of boundary defining function_$\hat{x}=e^{\omega}x$,_the operator_${\mathcal}{C}$_changes_according to conjugation $\hat{{\mathcal}{C}}=e^{-{\frac{n}{2}}\omega_0}{\mathcal}{C}
e^{{\frac{n}{2}}\omega_0}$.
Now we_want to prove that the_range of_$E(0)$ acting on $C^{\infty}({\partial}{\overline}{X},{^0\Sigma})$ is exactly the set_of_harmonic spinors in_$x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$.
\[span\] Let $\phi\in x^{{\frac{n}{2}}}C^{\infty}({\overline}{X},{^0\Sigma})$ such that $D_g\phi=0$ and let_$\psi:=(x^{-\frac n2}\phi)|_{{\partial}{\overline}{X}}$. Then we have $E(0)\psi=2\phi$.
First_let us write_$\psi=\psi_++\psi_-$_with_$\psi_\pm\in{^0\Sigma}_\pm$. Then we construct_the approximate solution $\sigma_{\infty,+}({\lambda})$ of associated_to $\psi_+$. Let us set $\phi_+({\lambda}):=\sigma_{\infty,+}({\lambda})$_and $\phi_-({\lambda}):=\phi-\phi_+({\lambda})$. One has $(x^{-{\frac{n}{2}}}\phi_-({0}))|_{x=0}=\psi_-\in{^0\Sigma}_-$ and $D_g\phi_-(0)=-D_g\phi_+(0)$._As in the proof of Proposition_\[poisson\], we have $$\sigma_{+ |
superalgebras $sl(m;j|n;\epsilon) $ are consistent with the transformations of (super) vectors $${\cal X}^t(j,\epsilon)=(x_1,j_1x_2,\ldots,(1,m)x_m|\nu(x_{m+1},
\epsilon_1x_{m+2},\ldots,[1,n]x_{m+n}))^t,
\label{sv}$$ where the odd components are denote as $
x_{m+1}=\theta_1,\ldots, x_{m+n}=\theta_{n}
$ and $
\hat{\bar i}={\bar i}-m, \,
\hat{\bar k}={\bar k}-m=1,\ldots,n,\,
[\hat{\bar i},\hat{\bar k}]=
\prod^{\max(\hat{\bar i},\hat{\bar k})-1}_{l=\min(\hat{\bar i},\hat{\bar k})}
\epsilon_l, \, \epsilon_l=1,\xi_l,i, \,
\xi^2_l=0, \,
\xi_l\xi_p=\xi_p\xi_l \neq 0. $ The components of ${\cal X}(j;\epsilon) $ are choosen in such a way that the contraction parameters $\epsilon_l $ of the odd components were independent of the contraction parameters $j_l $ of the even ones. The transformations of the standart generators (\[gn\]) (marked with star) of the special linear superalgebra $ sl(m|n) $ to the generators of $sl(m;j|n,\epsilon)$ are given by $$H_I=H^*_I, \; E_{ij}=(i,j)E^*_{ij}, \;
E_{\bar i\bar j}=[\hat{\bar i},\hat{\bar j}]E^*_{\bar i\bar j}, \;
i \neq j, \; \bar i \neq \bar j,$$ $$E_{i\bar j}=\nu(1,i)[1,\hat{\bar j}]E^*_{i\bar j}, \quad
E_{\bar ij}=\nu(1,j)[1,\hat{\bar i}]E^*_{\bar ij}.
\label{5.1+}$$
Nonzero commutators and anticommutators are easily obtained from the corresponding | superalgebras $ sl(m;j|n;\epsilon) $ are consistent with the transformations of (super) vectors $ $ { \cal X}^t(j,\epsilon)=(x_1,j_1x_2,\ldots,(1,m)x_m|\nu(x_{m+1 },
\epsilon_1x_{m+2},\ldots,[1,n]x_{m+n}))^t,
\label{sv}$$ where the curious component are denote as $
x_{m+1}=\theta_1,\ldots, x_{m+n}=\theta_{n }
$ and $
\hat{\bar i}={\bar i}-m, \,
\hat{\bar k}={\bar k}-m=1,\ldots, n,\,
[ \hat{\bar i},\hat{\bar k}]=
\prod^{\max(\hat{\bar i},\hat{\bar k})-1}_{l=\min(\hat{\bar i},\hat{\bar k }) }
\epsilon_l, \, \epsilon_l=1,\xi_l, i, \,
\xi^2_l=0, \,
\xi_l\xi_p=\xi_p\xi_l \neq 0. $ The components of $ { \cal X}(j;\epsilon) $ are choosen in such a way that the contraction parameters $ \epsilon_l $ of the leftover components were independent of the compression parameters $ j_l $ of the even ones. The transformation of the standart generators (\[gn\ ]) (marked with star topology) of the special linear superalgebra $ sl(m|n) $ to the generators of $ sl(m;j|n,\epsilon)$ are give by $ $ H_I = H^*_I, \; E_{ij}=(i, j)E^*_{ij }, \;
E_{\bar i\bar j}=[\hat{\bar i},\hat{\bar j}]E^*_{\bar i\bar j }, \;
i \neq j, \; \bar i \neq \bar j,$$ $ $ E_{i\bar j}=\nu(1,i)[1,\hat{\bar j}]E^*_{i\bar j }, \quad
E_{\bar ij}=\nu(1,j)[1,\hat{\bar i}]E^*_{\bar ij }.
\label{5.1+}$$
Nonzero commutators and anticommutators are easily prevail from the corresponding | suoeralgebras $sl(m;j|n;\epsilon) $ are consistenj qith tie tranaformatiuns of (super) vectors $${\cal X}^t(j,\xpsioon)=(x_1,j_1z_2,\ldots,(1,m)x_m|\nu(x_{m+1},
\epsilon_1x_{m+2},\udots,[1,n]x_{m+n}))^n,
\label{sv}$$ qhert the odd componeifs are denote ws $
x_{k+1}=\vheta_1,\ldots, x_{m+n}=\tmeta_{n}
$ and $
\vat{\bar i}={\bar i}-m, \,
\hxt{\yar k}={\bar k}-m=1,\ldots,n,\,
[\hat{\bar i},\hat{\bar k}]=
\proq^{\max(\hat{\nag i},\hat{\bar k})-1}_{l=\mig(\hat{\nwr i},\gat{\bar k})}
\epsilon_l, \, \epsilon_l=1,\xi_l,i, \,
\xi^2_l=0, \,
\xi_l\xi_p=\qi_p\xi_l \neq 0. $ The vomponents of ${\cal X}(j;\epsiloj) $ age choosen in such a way that the xontraction oarameters $\epsilon_l $ or the odd components were indepdndenc of the cobteachhon parametxrs $j_l $ of the evek ones. Dhe tramsformations on the stqndart generators (\[gn\]) (karked with star) os the spewicl linear superalgebrq $ sl(m|n) $ to the tendrauorx kf $sl(m;u|n,\e'silon)$ are fiven by $$H_I=Y^*_I, \; E_{ij}=(i,j)E^*_{ij}, \;
E_{\bar i\naw j}=[\hat{\bar i},\hat{\gar j}]E^*_{\faw i\bar j}, \;
i \neq j, \; \bar i \neq \bar j,$$ $$E_{i\bar j}=\nh(1,i)[1,\hat{\bar j}]E^*_{i\bar j}, \quad
E_{\var ij}=\nu(1,j)[1,\hat{\bar i}]E^*_{\bar ij}.
\label{5.1+}$$
Njnzero commutators and anticommutators are easily obtamndd yvjo tje corresponding | superalgebras $sl(m;j|n;\epsilon) $ are consistent with the (super) $${\cal X}^t(j,\epsilon)=(x_1,j_1x_2,\ldots,(1,m)x_m|\nu(x_{m+1}, \label{sv}$$ where the $ x_{m+n}=\theta_{n} $ and \hat{\bar i}={\bar i}-m, \hat{\bar k}={\bar k}-m=1,\ldots,n,\, [\hat{\bar i},\hat{\bar k}]= i},\hat{\bar k})-1}_{l=\min(\hat{\bar i},\hat{\bar k})} \epsilon_l, \, \epsilon_l=1,\xi_l,i, \, \xi^2_l=0, \, \xi_l\xi_p=\xi_p\xi_l \neq 0. The components of ${\cal X}(j;\epsilon) $ are choosen in such a way that contraction $\epsilon_l of odd components were independent of the contraction parameters $j_l $ of the even ones. The transformations the standart generators (\[gn\]) (marked with star) of special linear superalgebra $ $ to the generators of are by $$H_I=H^*_I, E_{ij}=(i,j)E^*_{ij}, E_{\bar j}=[\hat{\bar i},\hat{\bar j}]E^*_{\bar j}, \; i \neq j, \; \bar i \neq \bar j,$$ $$E_{i\bar j}=\nu(1,i)[1,\hat{\bar j}]E^*_{i\bar j}, \quad E_{\bar i}]E^*_{\bar ij}. commutators and are obtained the corresponding | superalgebras $sl(m;j|n;\epsilon) $ Are consistEnt wiTh tHe tRaNsfoRmatIons of (super) vecTOrs $${\cAl X}^t(j,\epsilon)=(x_1,j_1x_2,\ldots,(1,m)x_M|\nu(x_{m+1},
\EpSIlon_1X_{M+2},\lDots,[1,n]X_{m+n}))^t,
\labEL{sV}$$ WHerE tHe Odd CoMPoNents Are Denote aS $
x_{m+1}=\theta_1,\ldOts, X_{m+N}=\theta_{n}
$ and $
\haT{\BaR i}={\bar i}-m, \,
\hat{\Bar K}={\bar k}-m=1,\ldots,n,\,
[\Hat{\Bar i},\haT{\bAr k}]=
\PRod^{\maX(\haT{\bar i},\Hat{\bar K})-1}_{L=\min(\haT{\bar i},\hat{\bAr K})}
\EpsiloN_L, \, \epsiloN_L=1,\Xi_L,i, \,
\xi^2_L=0, \,
\xi_l\xi_p=\xi_p\xi_l \neq 0. $ THE cOMponents of ${\cal X}(J;\epsilOn) $ ARe CHOosEn iN such a way tHaT the cONtractiON pARAMetERs $\epsilon_l $ of tHe odd componENts Were inDePenDEnt of tHe conTrACtiOn parameterS $j_l $ oF the even oNes. The TRansforMAtions oF the stAndArt GeneRAtOrS (\[gn\]) (MaRKed WItH stAR) of The speciAl LiNear sUperALGEBra $ sL(m|n) $ To thE geneRators of $sl(m;j|n,\EpsIlon)$ ARe gIven bY $$H_I=H^*_I, \; e_{ij}=(i,J)E^*_{Ij}, \;
E_{\baR i\bar j}=[\Hat{\baR i},\Hat{\bar j}]E^*_{\bar i\bar J}, \;
i \neQ j, \; \bar i \neq \Bar J,$$ $$E_{I\baR j}=\Nu(1,i)[1,\haT{\Bar j}]E^*_{i\Bar J}, \quAd
E_{\bar iJ}=\nu(1,j)[1,\hat{\BAr i}]e^*_{\bAR IJ}.
\lAbel{5.1+}$$
Nonzero commutaToRS AnD anticomMutatoRS aRe EAsily obtAiNed From THE corrEspoNDiNg | superalgebras $sl(m;j|n;\ epsilon) $ arecon sis te nt w iththe transforma t ions of (super) vectors $$ {\cal X } ^t(j , \e psilo n)=(x_1 , j_ 1 x _2, \l do ts, (1 , m) x_m|\ nu( x_{m+1} ,
\epsilon _1x _{ m+2},\ldots, [ 1, n]x_{m+n}) )^t ,
\label{sv} $$whereth e o d d com pon entsare de n ote as $
x_{m+1 }= \ theta_ 1 ,\ldots , x_ {m+n }=\theta_{n}
$ an d $
\hat{\bar i}= {\bari} - m, \ ,
\ hat {\bar k}={ \b ar k} - m=1,\ld o ts , n , \,[ \hat{\bar i}, \hat{\bar k } ]=\prod^ {\ max ( \hat{\ bar i }, \ hat {\bar k})-1 }_{l =\min(\ha t{\bar i},\hat { \bar k} )}
\ep sil on_ l, \ , \ ep sil on _ l=1 , \x i_l , i,\,
\xi^2 _l =0 , \,\xi_ l \ x i _p=\ xi_ p\xi _l \n eq 0. $ The c omp onen t s o f ${\ cal X }(j; \e psilo n) $ a re ch oo sen in such a w ay t hat the c ont ra cti on para m eters$\e psi lon_l $ of the odd c o m p on ents were independ en t of the con tracti o npa r ameters$j _l$ of t he ev en o n es . The tr ansfor m at io ns of t he stand ar t g ene rator s (\[ gn\])(markedwiths tar) of the sp e cial linear s u pe r a lg e bra$ s l(m|n) $ to the gene rato r sof$ sl(m; j|n,\ ep s il o n)$ are given by $$ H_ I=H^*_ I, \; E_{ij}=(i,j) E^*_{ij},\ ; E_{\bari\ba r j } =[\hat{\bar i} ,\hat {\bar j}]E ^ *_{\bari\bar j}, \;i \neq j, \ ; \bar i \n eq\ba r j , $ $$$E_{i\bar j} = \ nu(1 ,i )[1,\ha t{\ bar j}] E^* _{i \ba r j }, \quad
E_ {\bar ij }= \n u( 1, j)[ 1,\ha t {\bar i} ]E ^*_ {\ bar ij}. \label {5.1+ }$$
N on z ero commut a to r s and a nt icom mut at ors a re e a sil y obtai ned fromthe corr es po nding | superalgebras_$sl(m;j|n;\epsilon) $_are consistent with the_transformations of_(super)_vectors $${\cal_X}^t(j,\epsilon)=(x_1,j_1x_2,\ldots,(1,m)x_m|\nu(x_{m+1},
\epsilon_1x_{m+2},\ldots,[1,n]x_{m+n}))^t,
\label{sv}$$_where the odd_components are denote_as $
x_{m+1}=\theta_1,\ldots, x_{m+n}=\theta_{n}
$ and_$
\hat{\bar i}={\bar_i}-m,_\,
\hat{\bar k}={\bar k}-m=1,\ldots,n,\,
[\hat{\bar i},\hat{\bar k}]=
\prod^{\max(\hat{\bar i},\hat{\bar k})-1}_{l=\min(\hat{\bar i},\hat{\bar k})}
\epsilon_l, \, \epsilon_l=1,\xi_l,i, \,
\xi^2_l=0, \,
\xi_l\xi_p=\xi_p\xi_l \neq_0._$ The_components_of_${\cal X}(j;\epsilon) $ are choosen_in such a way that_the contraction_parameters $\epsilon_l $ of the odd components were_independent_of the contraction_parameters $j_l $ of the even ones. The transformations_of the standart generators (\[gn\]) (marked_with star) of_the_special_linear superalgebra $ sl(m|n)_$ to the generators of $sl(m;j|n,\epsilon)$_are given by $$H_I=H^*_I, \; E_{ij}=(i,j)E^*_{ij},_\;
E_{\bar i\bar j}=[\hat{\bar i},\hat{\bar j}]E^*_{\bar i\bar j},_\;
i \neq j, \; \bar i_\neq \bar j,$$ $$E_{i\bar j}=\nu(1,i)[1,\hat{\bar_j}]E^*_{i\bar j},_\quad
E_{\bar ij}=\nu(1,j)[1,\hat{\bar i}]E^*_{\bar ij}.
\label{5.1+}$$
Nonzero commutators_and anticommutators are_easily obtained_from the corresponding |
ii) in the definition of $t$ following equation (\[eq:Vdef\]). Thus $\mathcal{D}_1(f)$ can only intersect the component $C(v)$ of $\mathcal{Y}_v$. Now (\[eq:fiberint\]) shows $$\langle \mathcal{D}_1(f),n_{C(v)} C(v)\rangle_v = \langle \mathcal{D}_1(f),\mathcal{Y}_v\rangle_v = \mathrm{deg}(D_1(f)) = m.$$ This shows (\[eq:crank1\]) and completes the proof of (iv).
Finally, the inequalities in (\[eq:crank\]) of part (v) are a consequence of (\[eq:urp\]), (\[eq:bounds\]) and (\[eq:crank1\]).
Controlling vertical divisors {#s:Rumcapdiv}
=============================
\[lem:fixit\] Let $\mathcal{D}$, $M$ and $S_v$ be as in Proposition \[prop:divprop\]. Suppose $M' \subset M$ and that $C_0(v) \in S_v$ for $v \in M'$. Then there is a function $h \in K(\mathcal{Y})$ such that $$\label{eq:divisornow}
\mathrm{div}_{\mathcal{Y}}(h) = \mathcal{D}_1 - \mathcal{D}_2 + \sum_{v \in M'} E_v$$ where $\mathcal{D}_1$ and $\mathcal{D}_2$ are horizontal effective divisors which do not intersect, $\mathcal{D}_2$ has the same support as $\mathcal{D}$, $E_v$ is supported on $\mathcal{Y}_v$, and for $v \in M'$ and $C' \in S_v$ we have $$\label{eq:crank2}
\langle E_v, C' \rangle_v > 0 \quad \mathrm{if} \quad C_0(v) \ne C' \in S_v \quad \mathrm{and}\quad
\langle E_v, C_0(v)\rangle_v
< 0.$$
We use induction on the number of elements of $M - M'$. If $M = M'$, the Lemma is shown by Proposition \[prop:divprop\]. Suppose now that | ii) in the definition of $ t$ following equation (\[eq: Vdef\ ]). Thus $ \mathcal{D}_1(f)$ can only intersect the part $ C(v)$ of $ \mathcal{Y}_v$. nowadays (\[eq: fiberint\ ]) shows $ $ \langle \mathcal{D}_1(f),n_{C(v) } C(v)\rangle_v = \langle \mathcal{D}_1(f),\mathcal{Y}_v\rangle_v = \mathrm{deg}(D_1(f) ) = m.$$ This shows (\[eq: crank1\ ]) and dispatch the validation of (iv).
Finally, the inequalities in (\[eq: crank\ ]) of part (volt) are a consequence of (\[eq: urp\ ]), (\[eq: bounds\ ]) and (\[eq: crank1\ ]).
Controlling vertical divisor { # s: Rumcapdiv }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = =
\[lem: fixit\ ] Let $ \mathcal{D}$, $ M$ and $ S_v$ be as in Proposition \[prop: divprop\ ]. Suppose $ megabyte' \subset M$ and that $ C_0(v) \in S_v$ for $ v \in M'$. Then there is a routine $ planck's constant \in K(\mathcal{Y})$ such that $ $ \label{eq: divisornow }
\mathrm{div}_{\mathcal{Y}}(h) = \mathcal{D}_1 - \mathcal{D}_2 + \sum_{v \in M' } E_v$$ where $ \mathcal{D}_1 $ and $ \mathcal{D}_2 $ are horizontal effective divisors which do not intersect, $ \mathcal{D}_2 $ has the same support as $ \mathcal{D}$, $ E_v$ is supported on $ \mathcal{Y}_v$, and for $ volt \in M'$ and $ C' \in S_v$ we have $ $ \label{eq: crank2 }
\langle E_v, C' \rangle_v > 0 \quad \mathrm{if } \quad C_0(v) \ne C' \in S_v \quad \mathrm{and}\quad
\langle E_v, C_0(v)\rangle_v
< 0.$$
We use trigger on the number of elements of $ M - M'$. If $ M = M'$, the Lemma is shown by Proposition \[prop: divprop\ ]. think now that | ii) ln the definition of $t$ fullowing equation (\[eq:Vdxf\]). Thus $\mathcal{A}_1(f)$ can only intersect the colpinent $C(v)$ of $\mathcal{Y}_v$. Now (\[ed:fiberint\]) shows $$\lqnglt \mathcal{D}_1(f),n_{C(v)} C(v)\czngle_v = \langls \matkcel{D}_1(f),\mathcal{Y}_v\rakgle_v = \mathsm{deg}(D_1(f)) = m.$$ Thiv rhlws (\[eq:crank1\]) and completes the proof of (iv).
Fonwlly, the inequwlitpef in (\[vq:grank\]) of part (v) are a consequencs of (\[eq:lrp\]), (\[eq:bounds\]) and (\[rq:crank1\]).
Controlling verticap digisors {#s:Rumcapdiv}
=============================
\[lfm:fixit\] Let $\matrxal{D}$, $M$ and $S_x$ be as in Proposition \[prop:divprop\]. Suppose $M' \subset M$ and chat $C_0(v) \in W_v$ fog $v \in M'$. Thei therv is a function $h \in N(\mathcak{Y})$ such that $$\lsben{eq:eivisornow}
\mathrm{div}_{\methcal{Y}}(h) = \mathcal{D}_1 - \iathcal{D}_2 + \sbm_{v \in M'} E_v$$ where $\matycql{D}_1$ atd $\mdthcxo{D}_2$ xre hprjzontap ehfective dibisors whicy do not intersect, $\kaeycal{D}_2$ has the same fu[port as $\mathcal{D}$, $E_v$ is supported on $\mauhcal{G}_v$, and for $v \in M'$ and $C' \in S_v$ we have $$\label{ee:crank2}
\laggle E_v, C' \rangle_v > 0 \quad \mathrm{if} \quad C_0(v) \ne C' \in S_v \qnaa \nabhrm{xbd}\euad
\langle E_v, C_0(v)\rangle_v
< 0.$$
We use induction og tne number of eleients of $M - M'$. Ig $M = M'$, the Lemoa is shosn by Proposition \[orop:divkrop\]. Syppose nor thst | ii) in the definition of $t$ following Thus can only the component $C(v)$ $$\langle C(v)\rangle_v = \langle = \mathrm{deg}(D_1(f)) = This shows (\[eq:crank1\]) and completes the of (iv). Finally, the inequalities in (\[eq:crank\]) of part (v) are a consequence (\[eq:urp\]), (\[eq:bounds\]) and (\[eq:crank1\]). Controlling vertical divisors {#s:Rumcapdiv} ============================= \[lem:fixit\] Let $\mathcal{D}$, $M$ $S_v$ as Proposition Suppose $M' \subset M$ and that $C_0(v) \in S_v$ for $v \in M'$. Then there is function $h \in K(\mathcal{Y})$ such that $$\label{eq:divisornow} \mathrm{div}_{\mathcal{Y}}(h) \mathcal{D}_1 - \mathcal{D}_2 + \in M'} E_v$$ where $\mathcal{D}_1$ $\mathcal{D}_2$ horizontal effective which not $\mathcal{D}_2$ has the support as $\mathcal{D}$, $E_v$ is supported on $\mathcal{Y}_v$, and for $v \in M'$ and $C' \in S_v$ have $$\label{eq:crank2} C' \rangle_v 0 \mathrm{if} C_0(v) \ne C' \quad \mathrm{and}\quad \langle E_v, C_0(v)\rangle_v < induction on the number of elements of $M M'$. If = M'$, the Lemma is shown Proposition \[prop:divprop\]. Suppose now that | ii) in the definition of $t$ folloWing equatiOn (\[eq:VDef\]). thuS $\mAthcAl{D}_1(f)$ Can only interseCT the Component $C(v)$ of $\mathcal{Y}_v$. now (\[eq:FiBErinT\]) ShOws $$\laNgle \matHCaL{d}_1(F),n_{C(V)} C(V)\rAngLe_V = \LaNgle \mAthCal{D}_1(f),\maThcal{Y}_v\ranGle_V = \mAthrm{deg}(D_1(f)) = m.$$ THIs Shows (\[eq:craNk1\]) aNd completes tHe pRoof of (Iv).
finALly, thE inEqualIties iN (\[Eq:cranK\]) of part (v) aRe A ConseqUEnce of (\[eQ:URp\]), (\[Eq:boUnds\]) and (\[eq:crank1\]).
ConTRoLLing vertical diVisors {#S:RUMcAPDiv}
=============================
\[Lem:Fixit\] Let $\maThCal{D}$, $M$ ANd $S_v$ be aS In pROPosITion \[prop:divprOp\]. Suppose $M' \sUBseT M$ and tHaT $C_0(v) \IN S_v$ for $V \in M'$. THeN TheRe is a functiOn $h \iN K(\mathcal{y})$ such tHAt $$\label{EQ:divisoRnow}
\maThrM{diV}_{\matHCaL{Y}}(H) = \maThCAl{D}_1 - \MAtHcaL{d}_2 + \suM_{v \in M'} E_v$$ wHeRe $\MathcAl{D}_1$ aND $\MAThcaL{D}_2$ aRe hoRizonTal effective dIviSors WHicH do noT inteRsecT, $\mAthcaL{D}_2$ has tHe samE sUpport as $\mathcal{d}$, $E_v$ iS supporteD on $\MaThcAl{y}_v$, and FOr $v \in M'$ And $c' \in s_v$ we havE $$\label{eQ:CraNk2}
\LANGlE E_v, C' \rangle_v > 0 \quad \matHrM{IF} \qUad C_0(v) \ne C' \In S_v \quAD \mAtHRm{and}\quaD
\lAngLe E_v, c_0(V)\RanglE_v
< 0.$$
We USe InductioN on the NUmBeR of elemEnTs of $M - M'$. if $m = M'$, tHe LEmma iS ShowN by ProPosition \[Prop:dIVprop\]. Suppose noW That | ii) in the definition of $ t$ followi ng eq uat ion ( \[eq :Vde f\]). Thus $\m a thca l{D}_1(f)$ can only in terse ct thec om ponen t $C(v) $ o f $\m at hc al{ Y} _ v$ . Now (\ [eq:fib erint\]) s how s$$\langle \m a th cal{D}_1(f ),n _{C(v)} C(v) \ra ngle_v = \l a ngle\ma thcal {D}_1( f ),\mat hcal{Y}_v \r a ngle_v = \math r m {d eg}( D_1(f)) = m.$$ Th i ss hows (\[eq:cra nk1\]) a n dc o mpl ete s the proo fof (i v ).
Fin a ll y , the inequalitiesin (\[eq:cr a nk\ ]) ofpa rt( v) are a co ns e que nce of (\[e q:ur p\]), (\[ eq:bou n ds\]) a n d (\[eq :crank 1\] ).
Con t ro ll ing v e rti c al di v iso rs {#s:R um ca pdiv}
=== = = = = ==== === ==== ===== ======
\[lem :fi xit\ ] Le t $\m athca l{D} $, $M$and $S _v$ b eas in Propositi on \ [prop:div pro p\ ].Su ppose $M' \s ubs etM$ andthat $C _ 0(v )\ i n S _v$ for $v \in M'$ .T h en there i s a fu n ct io n $h \inK( \ma thca l { Y})$such th at $$\la bel{eq : di vi sornow}
\mathr m{ div }_{ \math c al{Y }}(h)= \mathc al{D} _ 1 - \mathcal{D } _2 + \sum_{v\ in M '} E_v$ $ w here $\math cal{ D }_1$ and $\ mat h cal{D }_2$ar e h o rizontal effectivedi visors whic h do not inte rsect, $\m a t h cal{D}_2 $ ha s t h e same support as $ \mathcal{D } $, $E_v$ is s upported on $\mat h c al{Y}_v$ , a ndfor $v \ in M'$ and $C'\ i n S_ v$ we hav e $ $\label {eq :cr ank 2} \ langle E_ v, C' \r an gl e_ v> 0 \qua d \mathrm {i f}\q uad C_0( v ) \neC' \i n S_ v\q u ad \mathr m {a n d }\qu ad
\la ngl eE_v,C_0( v )\r angle_v
< 0.$$
We usein du ction o n the numberof elementsof $M - M'$ . If $M =M'$, the Lemma is shown by Prop osi tion\[pr op:divpro p\] . Supp ose now th at | ii) in_the definition_of $t$ following equation_(\[eq:Vdef\]). Thus_$\mathcal{D}_1(f)$_can only_intersect_the component $C(v)$_of $\mathcal{Y}_v$. Now_(\[eq:fiberint\]) shows $$\langle \mathcal{D}_1(f),n_{C(v)}_C(v)\rangle_v = \langle_\mathcal{D}_1(f),\mathcal{Y}_v\rangle_v_= \mathrm{deg}(D_1(f)) = m.$$ This shows (\[eq:crank1\]) and completes the proof of (iv).
Finally, the_inequalities_in (\[eq:crank\])_of_part_(v) are a consequence of_(\[eq:urp\]), (\[eq:bounds\]) and (\[eq:crank1\]).
Controlling vertical_divisors {#s:Rumcapdiv}
=============================
\[lem:fixit\]_Let $\mathcal{D}$, $M$ and $S_v$ be as in_Proposition_\[prop:divprop\]. Suppose $M'_\subset M$ and that $C_0(v) \in S_v$ for $v_\in M'$. Then there is a_function $h \in_K(\mathcal{Y})$_such_that $$\label{eq:divisornow}
\mathrm{div}_{\mathcal{Y}}(h) =_\mathcal{D}_1 - \mathcal{D}_2 + \sum_{v \in_M'} E_v$$ where $\mathcal{D}_1$ and $\mathcal{D}_2$_are horizontal effective divisors which do not_intersect, $\mathcal{D}_2$ has the same support_as $\mathcal{D}$, $E_v$ is supported_on $\mathcal{Y}_v$,_and for $v \in M'$_and $C' \in_S_v$ we_have $$\label{eq:crank2}
\langle_E_v, C' \rangle_v > 0 \quad_\mathrm{if} \quad C_0(v)_\ne C' \in S_v \quad _\mathrm{and}\quad
_ \langle E_v,_C_0(v)\rangle_v
_<_0.$$
We use_induction on the_number_of elements_of_$M - M'$. If $M =_M'$,_the Lemma is shown by Proposition \[prop:divprop\]._Suppose now that |
\operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I) \rightarrow \operatorname*{\mathbf{D}}(R)$ to be the left derived functor $\mathbf{L}\operatorname*{colim}_{i \in I}$ of the usual colimit functor $\operatorname*{colim}_{i \in I}: \operatorname*{Mod-R}^I \rightarrow \operatorname*{Mod-R}$. Dually, we define the [homotopy limit]{} functor as $\operatorname*{holim}_{i \in I} := \mathbf{R} \lim_{i \in I}: \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I) \rightarrow \operatorname*{\mathbf{D}}(R)$.
The objects of $\operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$, that is, the coherent diagrams of shape $I$, are all represented by diagrams of chain complexes of $R$-modules. Let $\mathscr{X} \in \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$ be represented by a diagram $(X_i \mid i \in I)$ of chain complexes. Then clearly, $\mathscr{X}_i \simeq X_i$ as objects of $\operatorname*{\mathbf{D}}(R)$ for any $i \in I$.
We will be especially interested in the homotopy colimit construction in the case when the small category $I$ is directed. In this situation, we call $\operatorname*{hocolim}_{i \in I}$ a [directed homotopy colimit]{}. Because the direct limit functor $\varinjlim_{i \in I} = \operatorname*{colim}_{i \in I}$ on the category of chain complexes of $R$-modules is exact, we have for each object $\mathscr{X} \in \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$, represented by a diagram $(X_i \mid i \in I)$ of chain complexes, the isomorphism $\operatorname*{hocolim}_{i \in I} \mathscr{X} \simeq \varinjlim_{i \in I} X_i$ in $\operatorname*{\mathbf{D}}(R)$. In particular, we have for any $n \in \mathbb{Z}$ the following isomorphism on cohomologies $$H^n(\operatorname*{hocolim}_{i \in I} \mathscr{X}) \simeq \varinjlim_{ | \operatorname*{\mathbf{D}}(\operatorname*{Mod - R}^I) \rightarrow \operatorname*{\mathbf{D}}(R)$ to be the left derived functor $ \mathbf{L}\operatorname*{colim}_{i \in I}$ of the usual colimit functor $ \operatorname*{colim}_{i \in I }: \operatorname*{Mod - R}^I \rightarrow \operatorname*{Mod - R}$. Dually, we specify the [ homotopy terminus ad quem ] { } functor as $ \operatorname*{holim}_{i \in I }: = \mathbf{R } \lim_{i \in I }: \operatorname*{\mathbf{D}}(\operatorname*{Mod - R}^I) \rightarrow \operatorname*{\mathbf{D}}(R)$.
The objects of $ \operatorname*{\mathbf{D}}(\operatorname*{Mod - R}^I)$, that is, the coherent diagram of shape $ I$, are all defend by diagrams of chain complex of $ R$-modules. Let $ \mathscr{X } \in \operatorname*{\mathbf{D}}(\operatorname*{Mod - R}^I)$ be represented by a diagram $ (X_i \mid i \in I)$ of range complex. Then clearly, $ \mathscr{X}_i \simeq X_i$ as object of $ \operatorname*{\mathbf{D}}(R)$ for any $ i \in I$.
We will be especially interested in the homotopy colimit structure in the case when the small category $ I$ is directed. In this situation, we call $ \operatorname*{hocolim}_{i \in I}$ a [ directed homotopy colimit ] { }. Because the lineal limit functor $ \varinjlim_{i \in I } = \operatorname*{colim}_{i \in I}$ on the category of chain complex of $ R$-modules is exact, we have for each aim $ \mathscr{X } \in \operatorname*{\mathbf{D}}(\operatorname*{Mod - R}^I)$, represented by a diagram $ (X_i \mid i \in I)$ of range complexes, the isomorphism $ \operatorname*{hocolim}_{i \in I } \mathscr{X } \simeq \varinjlim_{i \in I } X_i$ in $ \operatorname*{\mathbf{D}}(R)$. In particular, we have for any $ n \in \mathbb{Z}$ the following isomorphism on cohomologies $ $ H^n(\operatorname*{hocolim}_{i \in I } \mathscr{X }) \simeq \varinjlim _ { | \opfratorname*{\mathbf{D}}(\operatovname*{Mod-R}^I) \rightcerow \o'eratorhame*{\mathcf{D}}(R)$ to be the left derived huncror $\mqthbf{L}\operatorname*{colio}_{i \in I}$ ov the usyal rolimit functor $\operatorkcme*{comlm}_{i \iu M}: \operatorname*{Mpd-R}^I \rightdrrow \operatortaoe*{Lod-R}$. Dually, we define the [homotopy jimit]{} finftor as $\operatjrnakq*{holjm}_{i \in I} := \mathbf{R} \lim_{i \in I}: \operatkrname*{\methbf{D}}(\operatornsme*{Mod-R}^I) \rightarrow \operatlrnale*{\mathbf{D}}(R)$.
The objefts of $\operqtorgqme*{\mathbf{D}}(\opdratorname*{Mod-R}^I)$, that ia, the coherent diagrams of shapd $I$, axe all reprgssnhgd by diagrans of chain complcqes of $S$-modulex. Let $\mathscr{X} \in \opwratorname*{\mathbf{D}}(\operetorname*{Mod-R}^I)$ be reptesented bf c diagram $(X_i \mid i \in I)$ of cvain comooexds. Uhei cmearly, $\mavhscr{X}_i \simsq X_i$ as obhects of $\operatornake*{\iqthbf{D}}(R)$ for ahy $i \ig Y$.
We will be especially interested in tht homktopy colimit constructuon in the case when jhe small sategory $I$ is directed. In this situation, we call $\mperavofnane*{mocouum}_{l \in I}$ a [directed homotopy colimit]{}. Because thq dorvct limit functor $\varinjlim_{i \ij O} = \operatornamg*{colim}_{i \in I}$ on the category of chayn conplexes os $R$-mpdules is exact, we have for each object $\nathscr{X} \in \operatlrname*{\mathby{D}}(\operstornsme*{Mod-R}^I)$, represented by a djagram $(X_i \mld i \in I)$ uf chain complexds, nhe hsomorphism $\operatorname*{hosolim}_{i \in I} \mcthscr{X} \rimea \varigjlim_{i \in L} X_i$ lt $\operatorname*{\mathhf{D}}(R)$. Nn pasticular, wf have for any $n \in \mathbb{Z}$ the following isokosphpsm on cokomolonies $$H^n(\operatorgame*{hocolim}_{i \iu I} \mathfcr{X}) \rimeq \varihjlim_{ | \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I) \rightarrow \operatorname*{\mathbf{D}}(R)$ to be the left $\mathbf{L}\operatorname*{colim}_{i I}$ of usual colimit functor \operatorname*{Mod-R}$. we define the limit]{} functor as \in I} := \mathbf{R} \lim_{i \in \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I) \rightarrow \operatorname*{\mathbf{D}}(R)$. The objects of $\operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$, that is, the coherent diagrams of $I$, are all represented by diagrams of chain complexes of $R$-modules. Let $\mathscr{X} \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$ represented a $(X_i \mid i \in I)$ of chain complexes. Then clearly, $\mathscr{X}_i \simeq X_i$ as objects of for any $i \in I$. We will be interested in the homotopy construction in the case when small $I$ is In situation, call $\operatorname*{hocolim}_{i \in a [directed homotopy colimit]{}. Because the direct limit functor $\varinjlim_{i \in I} = \operatorname*{colim}_{i \in I}$ on category of of $R$-modules exact, have each object $\mathscr{X} represented by a diagram $(X_i \mid of chain complexes, the isomorphism $\operatorname*{hocolim}_{i \in I} \simeq \varinjlim_{i I} X_i$ in $\operatorname*{\mathbf{D}}(R)$. In particular, have for any $n \in \mathbb{Z}$ the following on cohomologies $$H^n(\operatorname*{hocolim}_{i \in I} \mathscr{X}) \simeq \varinjlim_{ | \operatorname*{\mathbf{D}}(\operatOrname*{Mod-R}^i) \righTarRow \OpEratOrnaMe*{\mathbf{D}}(R)$ to be THe leFt derived functor $\mathbf{l}\operAtORnamE*{CoLim}_{i \iN I}$ of the USuAL ColImIt FunCtOR $\oPeratOrnAme*{coliM}_{i \in I}: \operaTorNaMe*{Mod-R}^I \rightARrOw \operatorNamE*{Mod-R}$. Dually, wE deFine thE [hOmoTOpy liMit]{} FunctOr as $\opERatornAme*{holim}_{i \In i} := \Mathbf{r} \Lim_{i \in I}: \OPErAtorName*{\mathbf{D}}(\operatORnAMe*{Mod-R}^I) \rightarRow \opeRaTOrNAMe*{\mAthBf{D}}(R)$.
The objEcTs of $\oPEratornAMe*{\MATHbf{d}}(\Operatorname*{MOd-R}^I)$, that is, tHE coHerent DiAgrAMs of shApe $I$, aRe ALl rEpresented bY diaGrams of chAin comPLexes of $r$-Modules. let $\matHscR{X} \iN \opeRAtOrNamE*{\mAThbF{d}}(\oPerATorName*{Mod-R}^i)$ bE rEpresEnteD BY A DiagRam $(x_i \miD i \in I)$ Of chain compleXes. then CLeaRly, $\maThscr{x}_i \siMeQ X_i$ as ObjectS of $\opErAtorname*{\mathbf{D}}(r)$ for Any $i \in I$.
We WilL bE esPeCiallY IntereSteD in The homoTopy colIMit CoNSTRuCtion in the case when ThE SMaLl categoRy $I$ is dIReCtED. In this sItUatIon, wE CAll $\opEratORnAme*{hocolIm}_{i \in I}$ A [DiReCted homOtOpy colImIt]{}. BEcaUse thE DireCt limiT functor $\VarinJLim_{i \in I} = \operatoRName*{colim}_{i \in I}$ ON tHE CaTEgorY of Chain compleXes oF $r$-modUles IS eXacT, We havE for eAcH ObJEct $\mathscr{X} \in \operatOrName*{\maThbf{D}}(\Operatorname*{MOd-R}^I)$, represENTEd by a diaGram $(x_I \mID i \in I)$ of chain coMplexEs, the isomoRPhism $\opeRatorName*{hocoLim}_{i \in I} \maTHScr{X} \simeQ \vaRinJliM_{i \iN i} x_i$ In $\operatornamE*{\MAthbF{D}}(r)$. In partIcuLar, we haVe fOr aNy $n \In \mAtHbb{Z}$ the foLlowing iSoMoRpHiSm oN cohoMOlogies $$H^N(\oPerAtOrnAme*{hoCOlim}_{i \iN I} \matHscr{x}) \sImEQ \vaRinjlim_{ | \operatorname*{\mathbf{D} }(\operato rname *{M od- R} ^I)\rig htarrow \opera t orna me*{\mathbf{D}}(R)$ to be t he left de rived functo r $ \ m ath bf {L }\o pe r at ornam e*{ colim}_ {i \in I}$ of t he usual col i mi t functor$\o peratorname* {co lim}_{ i\in I}: \ ope rator name*{ M od-R}^ I \righta rr o w \ope r atornam e * {M od-R }$. Dually, we de f in e the [homotopy limit ]{ } f u n cto r a s $\operat or name* { holim}_ { i\ i n I} := \mathbf{R} \lim_{i \i n I} : \ope ra tor n ame*{\ mathb f{ D }}( \operatorna me*{ Mod-R}^I) \righ t arrow \ o perator name*{ \ma thb f{D} } (R )$ .
Th e ob j ec tso f $ \operato rn am e*{\m athb f { D } }(\o per ator name* {Mod-R}^I)$,tha t is , th e coh erent dia gr ams o f shap e $I$ ,are all represe nted by diagr ams o f c ha in co m plexes of $R $-modul es. Let $\m at h s c r{ X} \in \operatorna me * { \m athbf{D} }(\ope r at or n ame*{Mod -R }^I )$ b e repre sent e dby a dia gram $ ( X_ i\mid i\i n I)$of ch ain comp l exes . Then clearly , $\m a thscr{X}_i \si m eq X_i$ as ob j ec t s o f $\o per atorname*{\ math b f{D} }(R) $ f ora ny $i \inI$ .
W e will be especiall yintere stedin the homoto py colimit c o nstructi on i n t h e case when th e sma ll categor y $I$ isdirec ted. Inthis situ a t ion, wecal l $ \op era t o rn ame*{hocolim} _ { i \i nI}$ a [ dir ected h omo top y c oli mi t]{}. Bec ause the d ir ec tlim it fu n ctor $\v ar inj li m_{ i \in I} = \ opera torn am e* { col im}_{i\ in I }$ o nth e ca teg or y ofchai n co mplexes of $R$-m odu l es i sex act, we have for eac hobject $\m at hsc r{X} \ i n \operat orname*{\mathbf{D}}(\op e ratorna me* {Mod- R}^I )$, repre sen ted by ad iagram $(X_i \mid i \i n I)$ o f ch ain c omplexes,t h e i somor ph ism$\opera torname*{hocolim}_ { i \ in I} \mathsc r{X } \s i m eq \v a ri n jli m_ { i \ i n I} X_i$ in $\o peratornam e* { \m athbf{D}}( R )$. I n parti cular,we ha v e for a ny $n \in \mathbb{ Z} $ th e fol lowing iso morphism on cohom o logie s $ $H^n( \op erator na me* {hoco lim}_{ i \i n I}\maths cr {X}) \ simeq \ varinjli m_{ | \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)_\rightarrow \operatorname*{\mathbf{D}}(R)$_to be the left_derived functor_$\mathbf{L}\operatorname*{colim}_{i_\in I}$_of_the usual colimit_functor $\operatorname*{colim}_{i \in_I}: \operatorname*{Mod-R}^I \rightarrow \operatorname*{Mod-R}$._Dually, we define_the_[homotopy limit]{} functor as $\operatorname*{holim}_{i \in I} := \mathbf{R} \lim_{i \in I}: \operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I) \rightarrow_\operatorname*{\mathbf{D}}(R)$.
The_objects of_$\operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$,_that_is, the coherent diagrams of_shape $I$, are all represented_by diagrams_of chain complexes of $R$-modules. Let $\mathscr{X} \in_\operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$_be represented by_a diagram $(X_i \mid i \in I)$ of chain_complexes. Then clearly, $\mathscr{X}_i \simeq X_i$_as objects of_$\operatorname*{\mathbf{D}}(R)$_for_any $i \in I$.
We_will be especially interested in the_homotopy colimit construction in the case_when the small category $I$ is directed._In this situation, we call $\operatorname*{hocolim}_{i_\in I}$ a [directed homotopy_colimit]{}. Because_the direct limit functor $\varinjlim_{i_\in I} =_\operatorname*{colim}_{i \in_I}$ on the_category of chain complexes of $R$-modules_is exact, we_have for each object $\mathscr{X} \in_\operatorname*{\mathbf{D}}(\operatorname*{Mod-R}^I)$,_represented by a_diagram_$(X_i_\mid i_\in I)$ of_chain_complexes, the_isomorphism_$\operatorname*{hocolim}_{i \in I} \mathscr{X} \simeq \varinjlim_{i_\in_I} X_i$ in $\operatorname*{\mathbf{D}}(R)$. In particular, we_have for any $n_\in_\mathbb{Z}$ the following isomorphism_on cohomologies $$H^n(\operatorname*{hocolim}_{i \in I}_\mathscr{X}) \simeq \varinjlim_{ |
mathrm{v},\sharp}(\alpha)+\alpha,\gamma^{\mathrm{v},\sharp}(\beta)+\beta]_D\in L_{\gamma}, \ \ \forall \alpha,\beta\in \Gamma(H_{\gamma}^{\circ}).$$ Explicitly, this expression is given by: $$[\gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]+\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta-\iota_{\gamma^{\mathrm{v},\sharp}(\beta)} d\alpha+d\gamma^{\mathrm{v}}(\alpha,\beta),$$ and denote the form part by $[\alpha,\beta]_{\gamma^{\mathrm{v}}}$. Since the vector part is vertical, (\[EQ\_condition1\]) is equivalent to the following two conditions $$\label{EQ_condition2}
[\gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]=\gamma^{\mathrm{v},\sharp}([\alpha,\beta]_{\gamma^{\mathrm{v}}}),$$ $$\label{EQ_condition3}
[\alpha,\beta]_{\gamma^{\mathrm{v}}}\in \Gamma(H_{\gamma}^{\circ}),$$ for all $\alpha,\beta\in \Gamma(H_{\gamma}^{\circ})$. Now for $\beta\in \Gamma(V^{\circ})$, and any 1-form $\alpha$, we have that $$[\alpha,\beta]_{\gamma^{\mathrm{v}}}=\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta\in \Gamma(H_{\gamma}^{\circ}).$$ Therefore, if (\[EQ\_condition2\]) holds, then it holds for all 1-forms $\alpha,\beta\in\Omega^1(E)$. On the other hand, one easily sees (e.g. by applying (\[EQ\_condition2\]) to $\alpha=df,\beta=dg$) that this is equivalent to $\gamma^{\mathrm{v}}$ being Poisson i.e. $[\gamma^{\mathrm{v}},\gamma^{\mathrm{v}}]=0$. For $\alpha,\beta\in\Gamma(H_{\gamma}^{\circ})$, contracting $[\alpha,\beta]_{\gamma^{\mathrm{v}}}$ with $\mathrm{hor}(X)$, for $X\in\mathfrak{X}(S)$, one obtains $$\begin{aligned}
-\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}&L_{\mathrm{hor}(X)}(\beta)+\iota_{\gamma | mathrm{v},\sharp}(\alpha)+\alpha,\gamma^{\mathrm{v},\sharp}(\beta)+\beta]_D\in L_{\gamma }, \ \ \forall \alpha,\beta\in \Gamma(H_{\gamma}^{\circ}).$$ Explicitly, this expression is given by: $ $ [ \gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]+\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta-\iota_{\gamma^{\mathrm{v},\sharp}(\beta) } d\alpha+d\gamma^{\mathrm{v}}(\alpha,\beta),$$ and denote the form character by $ [ \alpha,\beta]_{\gamma^{\mathrm{v}}}$. Since the vector character is vertical, (\[EQ\_condition1\ ]) is equivalent to the following two conditions $ $ \label{EQ_condition2 }
[ \gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]=\gamma^{\mathrm{v},\sharp}([\alpha,\beta]_{\gamma^{\mathrm{v}}}),$$ $ $ \label{EQ_condition3 }
[ \alpha,\beta]_{\gamma^{\mathrm{v}}}\in \Gamma(H_{\gamma}^{\circ}),$$ for all $ \alpha,\beta\in \Gamma(H_{\gamma}^{\circ})$. immediately for $ \beta\in \Gamma(V^{\circ})$, and any 1 - form $ \alpha$, we have that $ $ [ \alpha,\beta]_{\gamma^{\mathrm{v}}}=\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta\in \Gamma(H_{\gamma}^{\circ}).$$ consequently, if (\[EQ\_condition2\ ]) holds, then it holds for all 1 - forms $ \alpha,\beta\in\Omega^1(E)$. On the early bridge player, one easily go steady (e.g. by apply (\[EQ\_condition2\ ]) to $ \alpha = df,\beta = dg$) that this is equivalent to $ \gamma^{\mathrm{v}}$ being Poisson i.e. $ [ \gamma^{\mathrm{v}},\gamma^{\mathrm{v}}]=0$. For $ \alpha,\beta\in\Gamma(H_{\gamma}^{\circ})$, contracting $ [ \alpha,\beta]_{\gamma^{\mathrm{v}}}$ with $ \mathrm{hor}(X)$, for $ X\in\mathfrak{X}(S)$, one obtains $ $ \begin{aligned }
-\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}&L_{\mathrm{hor}(X)}(\beta)+\iota_{\gamma | matjrm{v},\sharp}(\alpha)+\alpha,\gamma^{\oathrm{v},\sharp}(\betc)+\veta]_D\ii L_{\gammz}, \ \ \foraul \alpha,\beta\in \Gamma(H_{\gamma}^{\cicc}).$$ Ezplicutly, this expression ir given bj: $$[\gamma^{\marhrm{t},\sharp}(\alpha),\gamma^{\mathrm{v},\smcrp}(\befw)]+\iotc_{\gemma^{\mathrm{v},\shark}(\alpha)}d\beta-\imta_{\gamma^{\mathrm{e},\sfaxp}(\beta)} d\alpha+d\gamma^{\mathrm{v}}(\alpha,\beta),$$ wnd denptf the form parj by $[\sjpha,\gvtc]_{\gamma^{\mathrm{v}}}$. Since the vector lart is vertical, (\[EQ\_cpndition1\]) is equivalent to hhe vollowing two condltions $$\labeo{EQ_cjbdition2}
[\gamma^{\oathrm{v},\shag'}(\alpha),\gamma^{\jathrm{v},\sharp}(\beta)]=\gamma^{\mathrm{v},\shafp}([\alpka,\beta]_{\gamma^{\narhrl{e}}}),$$ $$\label{EQ_coiditiog3}
[\alpha,\beta]_{\gamma^{\mathrk{v}}}\in \Gakma(H_{\gamma}^{\circ}),$$ nor anl $\qlpha,\beta\in \Gamma(H_{\gamka}^{\circ})$. Now for $\betw\in \Gamma(E^{\cnrc})$, and any 1-form $\alphq$, qe haee tvat $$[\xophx,\beua]_{\gemmz^{\mathrl{v}}}=\ikta_{\gamma^{\mafhrm{v},\sharp}(\aopha)}d\beta\in \Gamma(H_{\gsmiq}^{\circ}).$$ Therefode, if (\[QQ\_sondition2\]) holds, then it holds for all 1-fmrma $\alpha,\beta\in\Omega^1(E)$. On rhe other hand, one eadily sees (e.g. by applying (\[EQ\_condition2\]) to $\alpha=df,\beta=dg$) that dhis ms eqblyalevr ho $\gamma^{\mathrm{v}}$ being Poisson i.e. $[\gamma^{\mathrm{v}},\gwjms^{\msthrm{v}}]=0$. For $\alpma,\beta\in\Gamma(H_{\gamms}^{\clrv})$, contracting $[\xlpha,\bzfa]_{\famma^{\mathrm{v}}}$ with $\lathrm{hjr}(X)$, fir $X\in\matrfral{X}(S)$, one obtains $$\begin{alignee}
-\iota_{\gamma^{\manhrm{c},\sharp}(\alpha)}&L_{\mathrm{kor}(X)}(\beta)+\iota_{\yamma | mathrm{v},\sharp}(\alpha)+\alpha,\gamma^{\mathrm{v},\sharp}(\beta)+\beta]_D\in L_{\gamma}, \ \ \forall \alpha,\beta\in \Gamma(H_{\gamma}^{\circ}).$$ expression given by: d\alpha+d\gamma^{\mathrm{v}}(\alpha,\beta),$$ and denote Since vector part is (\[EQ\_condition1\]) is equivalent the following two conditions $$\label{EQ_condition2} [\gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]=\gamma^{\mathrm{v},\sharp}([\alpha,\beta]_{\gamma^{\mathrm{v}}}),$$ [\alpha,\beta]_{\gamma^{\mathrm{v}}}\in \Gamma(H_{\gamma}^{\circ}),$$ for all $\alpha,\beta\in \Gamma(H_{\gamma}^{\circ})$. Now for $\beta\in \Gamma(V^{\circ})$, and any 1-form we have that $$[\alpha,\beta]_{\gamma^{\mathrm{v}}}=\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta\in \Gamma(H_{\gamma}^{\circ}).$$ Therefore, if (\[EQ\_condition2\]) holds, then it holds for 1-forms On other one easily sees (e.g. by applying (\[EQ\_condition2\]) to $\alpha=df,\beta=dg$) that this is equivalent to $\gamma^{\mathrm{v}}$ being i.e. $[\gamma^{\mathrm{v}},\gamma^{\mathrm{v}}]=0$. For $\alpha,\beta\in\Gamma(H_{\gamma}^{\circ})$, contracting $[\alpha,\beta]_{\gamma^{\mathrm{v}}}$ with $\mathrm{hor}(X)$, $X\in\mathfrak{X}(S)$, one obtains $$\begin{aligned} | mathrm{v},\sharp}(\alpha)+\alpha,\gamMa^{\mathrm{v},\sHarp}(\bEta)+\BetA]_D\In L_{\gAmma}, \ \ \Forall \alpha,\betA\In \GaMma(H_{\gamma}^{\circ}).$$ ExplicitlY, this ExPRessIOn Is givEn by: $$[\gamMA^{\mATHrm{V},\sHaRp}(\aLpHA),\gAmma^{\mAthRm{v},\sharP}(\beta)]+\iota_{\gAmmA^{\mAthrm{v},\sharp}(\aLPhA)}d\beta-\iota_{\GamMa^{\mathrm{v},\shaRp}(\bEta)} d\alPhA+d\gAMma^{\maThrM{v}}(\alpHa,\beta),$$ ANd denoTe the form PaRT by $[\alpHA,\beta]_{\gaMMA^{\mAthrM{v}}}$. Since the vector pARt IS vertical, (\[EQ\_conDition1\]) Is EQuIVAleNt tO the followInG two cONditionS $$\LaBEL{eQ_cONdition2}
[\gamma^{\mAthrm{v},\sharp}(\ALphA),\gamma^{\MaThrM{V},\sharp}(\Beta)]=\gAmMA^{\maThrm{v},\sharp}([\aLpha,\Beta]_{\gamma^{\Mathrm{V}}}),$$ $$\Label{EQ_COnditioN3}
[\alpha,\BetA]_{\gaMma^{\mAThRm{V}}}\in \gaMMa(H_{\GAmMa}^{\cIRc}),$$ fOr all $\alpHa,\BeTa\in \GAmma(h_{\GAMMa}^{\ciRc})$. NOw foR $\beta\In \Gamma(V^{\circ})$, aNd aNy 1-foRM $\alPha$, we Have tHat $$[\aLpHa,\betA]_{\gamma^{\MathrM{v}}}=\Iota_{\gamma^{\mathrm{V},\shaRp}(\alpha)}d\bEta\In \gamMa(h_{\gammA}^{\Circ}).$$ ThEreForE, if (\[EQ\_coNdition2\]) HOldS, tHEN It Holds for all 1-forms $\alPhA,\BEtA\in\Omega^1(e)$. On the OThEr HAnd, one eaSiLy sEes (e.G. BY applYing (\[eq\_cOndition2\]) To $\alphA=Df,\BeTa=dg$) thaT tHis is eQuIvaLenT to $\gaMMa^{\maThrm{v}}$ bEing PoisSon i.e. $[\GAmma^{\mathrm{v}},\gamMA^{\mathrm{v}}]=0$. For $\alPHa,\BETa\IN\GamMa(H_{\Gamma}^{\circ})$, coNtraCTing $[\AlphA,\BeTa]_{\gAMma^{\maThrm{v}}}$ WiTH $\mAThrm{hor}(X)$, for $X\in\mathfRaK{X}(S)$, one ObtaiNs $$\begin{aligneD}
-\iota_{\gamma^{\MATHrm{v},\sharP}(\alpHA)}&L_{\MAthrm{hor}(X)}(\beta)+\iOta_{\gaMma | mathrm{v},\sharp}(\alpha)+ \alpha,\ga mma^{ \ma thr m{ v},\ shar p}(\beta)+\bet a ]_D\ in L_{\gamma}, \ \ \fo rall\a l pha, \ be ta\in \Gamma ( H_ { \ gam ma }^ {\c ir c }) .$$ E xpl icitly, this expr ess io n is given b y :$$[\gamma^ {\m athrm{v},\sh arp }(\alp ha ),\ g amma^ {\m athrm {v},\s h arp}(\ beta)]+\i ot a _{\gam m a^{\mat h r m{ v},\ sharp}(\alpha)}d\ b et a -\iota_{\gamma ^{\mat hr m {v } , \sh arp }(\beta)}d\ alpha + d\gamma ^ {\ m a t hrm { v}}(\alpha,\b eta),$$ and den ote th efor m partby $[ \a l pha ,\beta]_{\g amma ^{\mathrm {v}}}$ . Sincet he vect or par t i s v erti c al ,(\[ EQ \ _co n di tio n 1\] ) is equ iv al ent t o th e f o llow ing two cond itions $$\lab el{ EQ_c o ndi tion2 }
[\g amma ^{ \math rm{v}, \shar p} (\alpha),\gamma ^{\m athrm{v}, \sh ar p}( \b eta)] = \gamma ^{\ mat hrm{v}, \sharp} ( [\a lp h a , \b eta]_{\gamma^{\mat hr m { v} }}),$$ $ $\labe l {E Q_ c ondition 3}
[\ alph a , \beta ]_{\ g am ma^{\mat hrm{v} } }\ in \Gamma (H _{\gam ma }^{ \ci rc}), $ $ fo r all$\alpha, \beta \ in \Gamma(H_{\ g amma}^{\circ} ) $. N ow for$\b eta\in \Gam ma(V ^ {\ci rc}) $ ,and any 1 -form $ \ al p ha$, we have that $ $[ \alpha ,\bet a]_{\gamma^{\ mathrm{v}} } = \ iota_{\g amma ^ {\ m athrm{v},\shar p}(\a lpha)}d\be t a\in \Ga mma(H _{\gamma }^{\circ} ) . $$ There for e,if(\[ E Q \_ condition2\]) h olds ,then it ho lds for al l 1 -fo rms $ \alpha,\b eta\in\O me ga ^1 (E )$. On t h e otherha nd, o neeasil y sees(e.g. byap pl y ing (\[EQ\ _ co n d itio n2 \] ) to $\ al pha=d f,\b e ta= dg$) th at this i s e q uiva le nt to $\g amma^{\mathrm {v }}$ beingPo iss on i.e . $[\gamma ^{\mathrm{v}},\gamma^{\ m athrm{v }}] =0$.For$\alpha,\ bet a\in\G amm a (H_{\g amma}^ {\cir c} )$, c ontra c t in g $ [\ alpha,\bet a ] _{\ gamma ^{ \mat hrm{v}} }$ with $\mathrm{h o r}( X)$, for $X\i n\m athf r a k{ X}( S )$ , on eo bta i n s $$\begin{alig ned}
-\iot a_ { \g amma^{\mat h rm{ v} ,\sharp }(\alph a)}&L _ {\mathr m{hor}(X) }(\beta)+ \i ota_ { \ gam ma | mathrm{v},\sharp}(\alpha)+\alpha,\gamma^{\mathrm{v},\sharp}(\beta)+\beta]_D\in L_{\gamma},_\ \_\forall \alpha,\beta\in \Gamma(H_{\gamma}^{\circ}).$$ Explicitly,_this expression_is_given by:_$$[\gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]+\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta-\iota_{\gamma^{\mathrm{v},\sharp}(\beta)}_d\alpha+d\gamma^{\mathrm{v}}(\alpha,\beta),$$ and denote_the form part_by $[\alpha,\beta]_{\gamma^{\mathrm{v}}}$. Since the_vector part is_vertical,_(\[EQ\_condition1\]) is equivalent to the following two conditions $$\label{EQ_condition2}
[\gamma^{\mathrm{v},\sharp}(\alpha),\gamma^{\mathrm{v},\sharp}(\beta)]=\gamma^{\mathrm{v},\sharp}([\alpha,\beta]_{\gamma^{\mathrm{v}}}),$$ $$\label{EQ_condition3}
[\alpha,\beta]_{\gamma^{\mathrm{v}}}\in \Gamma(H_{\gamma}^{\circ}),$$ for all $\alpha,\beta\in_\Gamma(H_{\gamma}^{\circ})$._Now for_$\beta\in_\Gamma(V^{\circ})$,_and any 1-form $\alpha$, we_have that $$[\alpha,\beta]_{\gamma^{\mathrm{v}}}=\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}d\beta\in \Gamma(H_{\gamma}^{\circ}).$$ Therefore,_if (\[EQ\_condition2\])_holds, then it holds for all 1-forms $\alpha,\beta\in\Omega^1(E)$._On_the other hand,_one easily sees (e.g. by applying (\[EQ\_condition2\]) to $\alpha=df,\beta=dg$) that_this is equivalent to $\gamma^{\mathrm{v}}$ being_Poisson i.e. $[\gamma^{\mathrm{v}},\gamma^{\mathrm{v}}]=0$. For_$\alpha,\beta\in\Gamma(H_{\gamma}^{\circ})$,_contracting_$[\alpha,\beta]_{\gamma^{\mathrm{v}}}$ with $\mathrm{hor}(X)$, for_$X\in\mathfrak{X}(S)$, one obtains $$\begin{aligned}
-\iota_{\gamma^{\mathrm{v},\sharp}(\alpha)}&L_{\mathrm{hor}(X)}(\beta)+\iota_{\gamma |
.
A main contribution is simple and general asymptotic theory for LR estimators that use cross-fitting in the construction of the average moments. This theory is based on the structure of LR moment conditions as an identifying moment condition depending on one first step plus an influence adjustment that can depend on an additional first step. We give a remainder decomposition that leads to mean square consistency conditions for first steps plus a few readily interpretable rate conditions. For DR estimators there is only one rate condition, on a product of sample remainders from two first step estimators, leading to particularly simple conditions. This simplicity motivates our inclusion of results for DR estimators. This asymptotic theory is also useful for existing moment conditions that are already known to be LR. Whenever the moment condition can be decomposed into an identifying moment condition depending on one first step and an influence function term that may depend on two first steps the simple and general regularity conditions developed here will apply.
LR moments reduce that smoothing bias that results from first step nonparametric estimation relative to original moment conditions. There are other sources of bias arising from nonlinearity of moment conditions in the first step and the empirical distribution. Cattaneo and Jansson (2017) and Cattaneo, Jansson, and Ma (2017) give useful bootstrap and jackknife methods that reduce nonlinearity bias. Newey and Robins (2017) show that one can also remove this bias by cross fitting in some settings. We allow for cross-fitting in this paper.
Section 2 describes the general construction of LR moment functions for semiparametric GMM. Section 3 gives LR moment conditions for dynamic discrete choice. Section 4 shows how to estimate the first step influence adjustment. Section 5 gives novel classes of DR moment functions and characterizes double robustness. Section 6 gives an orthogonal instrument construction of LR moments based on conditional moment restrictions. Section 7 provides simple and general asymptotic theory for LR estimators.
Locally Robust Moment Functions
===============================
The subject of this paper is GMM estimators of parameters where the sample moment functions depend on a first step nonparametric or large dimensional estimator. We refer to these estimators as semiparametric. We could also refer to them as GMM where first step estimators are plugged in the moments. This terminology seems awkward though, so we simply refer to them as semiparametric GMM estimators. We denote such an estimator by $\hat{\beta}$, which is a function of the data $z_{1},...,z_{n}$ where $n$ | .
A main contribution is simple and general asymptotic theory for LR calculator that practice crabbed - fitting in the construction of the average moments. This hypothesis is based on the structure of LR moment circumstance as an identifying moment circumstance depend on one first footfall plus an influence adjustment that can depend on an extra first step. We give a remainder decomposition that leads to mean straight consistency conditions for first steps plus a few promptly interpretable rate condition. For DR calculator there is only one rate condition, on a product of sample remainder from two first step estimators, leading to particularly simple conditions. This simplicity motivates our inclusion of results for DR estimators. This asymptotic theory is also useful for existing moment condition that are already known to be LR. Whenever the moment circumstance can be decompose into an identify moment condition depending on one first step and an influence function term that may count on two first steps the simple and general regularity conditions developed here will apply.
LR moments reduce that smoothing bias that results from inaugural step nonparametric appraisal relative to original moment conditions. There are other source of bias arising from nonlinearity of moment condition in the beginning step and the empirical distribution. Cattaneo and Jansson (2017) and Cattaneo, Jansson, and Ma (2017) collapse useful bootstrap and clasp knife methods that reduce nonlinearity bias. Newey and Robins (2017) show that one can also get rid of this bias by crisscross fitting in some settings. We leave for cross - fitting in this paper.
incision 2 identify the general construction of LR moment functions for semiparametric GMM. Section 3 gives LR moment conditions for dynamic discrete option. Section 4 shows how to calculate the first step influence adjustment. department 5 gives novel classes of DR moment functions and characterizes double robustness. Section 6 gives an orthogonal legal document structure of LR moments based on conditional moment restrictions. Section 7 provides simple and general asymptotic hypothesis for LR estimators.
Locally Robust Moment Functions
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
The subject of this paper is GMM estimators of parameters where the sample moment functions depend on a first measure nonparametric or large dimensional estimator. We refer to these estimators as semiparametric. We could also refer to them as GMM where inaugural step estimators are plugged in the moments. This terminology seems awkward though, so we just refer to them as semiparametric GMM estimators. We denote such an estimator by $ \hat{\beta}$, which is a function of the data $ z_{1},... ,z_{n}$ where $ n$ | .
A mwin contribution is simpue and general cwymptovic thekry for UR estimators that use cross-hittung ib the construction of ghe averahe momenrs. Tiis theory is based on tmz strhgture if LR moment cpnditions ds an identifyhne loment condition depending on one fyrst strp plus an influgnce sqjusfment that can depend on an additiknal figst step. We give s remainder decomposition hhat leads to mean squwre consistghcy xonditions fur first sue's plus a fgw readily interpretable rate covditipns. For DR ewtildtors there is ogly one rate conditimn, on a product of sakplx renainders from two firvt step estimators, leading do particularly simplw congitimns. Gyis sijpkidity mltitates our ihclusion of results for DR estomwnprs. This asyjptotis eheory is also useful for existing momett donditions that are alrwady known to be LR. Wjenever tre moment condition can be decomposed into an idettifymne mimcnt zinfition depending on one first step and an infjhemcv function term tmat may depend on ywl gyrst steps thg simplz ahd general regularlty conqitiobs develoked hrre will apply.
LR moments reeuce that smjithing bias that rzsults from yirst xtep monparametric estimatiou relafive to orihinal momsvt conditions. Thdre ase other sources of bias awising frim npnlineafity of moient condihions in the first step wnd tke em[irical didtribution. Cattaneo and Jansson (2017) and Cattaneo, Jdnsvon, and Ia (2017) glve useful bootftrap and jacklnife mzthods that redube nonlinxarity bias. Gewey and Robhjs (2017) show thav one can alsi renove thkr bias by crosx fitting in some swttings. We allow fpr zdoss-fitting in ckiw paper.
Section 2 dercrybvs vhe gqteral constrgctiun ug LR ooment funcbiovs fpr semiparametric GMK. Sedtion 3 gives LR mokekt condituons for dynamic disctete choice. Sectioj 4 shmws how yo gstimate the first step influende adjustlenb. Section 5 gides kovej classes pf DR moment functions and characterizew double robustnesw. Section 6 gives an prthogonal mnstruient consdruction of LR momenrs based on condiuional moment restrictikns. Sewtion 7 provides simple and general asymptotic theory for LR estimators.
Locally Eobust Moment Funcfionx
===============================
The vuyjeet of tris 'a'er is GMM estimanors of parameters where the sam'le moment fbnctions depend on a first strp nonparametriz or large dimensional eatimator. We refer to these estimators ax semiparametric. We coulc also rwfer fo them as GMM where first syep escimatoes are plugned in the momevts. Tmis terminllogb szems awkward though, so we simply eefer ro them as semiparemetric GMM csuimatora. We dgnote such an estjmatkg by $\fat{\beta}$, which is a function of the data $z_{1},...,z_{n}$ whege $g$ | . A main contribution is simple and theory LR estimators use cross-fitting in moments. theory is based the structure of moment conditions as an identifying moment depending on one first step plus an influence adjustment that can depend on additional first step. We give a remainder decomposition that leads to mean square conditions first plus few readily interpretable rate conditions. For DR estimators there is only one rate condition, on a of sample remainders from two first step estimators, to particularly simple conditions. simplicity motivates our inclusion of for estimators. This theory also for existing moment that are already known to be LR. Whenever the moment condition can be decomposed into an identifying condition depending first step an function that may depend first steps the simple and general here will apply. LR moments reduce that smoothing that results first step nonparametric estimation relative to moment conditions. There are other sources of bias from nonlinearity of moment conditions in the first step and the empirical distribution. Cattaneo and and Cattaneo, Jansson, and (2017) give useful and methods reduce bias. Newey Robins (2017) show that one can also remove this bias by fitting in some settings. We allow for cross-fitting in this 2 the general construction LR moment functions for GMM. 3 gives LR moment dynamic choice. how estimate first step influence adjustment. 5 gives novel classes of moment functions and characterizes an orthogonal instrument construction of LR moments based conditional moment restrictions. Section 7 provides simple general asymptotic theory for LR estimators. Locally Robust Moment Functions =============================== The of this GMM estimators of parameters where the sample moment depend on a first nonparametric or large dimensional estimator. We refer to these as We could refer to them GMM where first estimators are plugged moments. This seems though, refer to them as semiparametric GMM We such an estimator by $\hat{\beta}$, is of the data $z_{1},...,z_{n}$ where $n$ | .
A main contribution is simple And general AsympTotIc tHeOry fOr LR Estimators that USe crOss-fitting in the construCtion Of THe avERaGe momEnts. ThiS ThEORy iS bAsEd oN tHE sTructUre Of LR momEnt conditiOns As An identifyinG MoMent conditIon Depending on oNe fIrst stEp PluS An infLueNce adJustmeNT that cAn depend oN aN AdditiONal firsT STeP. We gIve a remainder decoMPoSItion that leads To mean SqUArE COnsIstEncy conditIoNs for FIrst stePS pLUS A feW Readily interpRetable rate COndItions. foR DR EStimatOrs thErE Is oNly one rate cOndiTion, on a prOduct oF Sample rEMainderS from tWo fIrsT steP EsTiMatOrS, LeaDInG to PArtIcularly SiMpLe conDitiONS. tHis sImpLiciTy motIvates our inclUsiOn of REsuLts foR DR esTimaToRs. ThiS asympTotic ThEory is also usefuL for Existing mOmeNt ConDiTions THat are AlrEadY known tO be LR. WhENevEr THE MoMent condition can be DeCOMpOsed into An idenTIfYiNG moment cOnDitIon dEPEndinG on oNE fIrst step And an iNFlUeNce funcTiOn term ThAt mAy dEpend ON two First sTeps the sImple ANd general regulARity conditionS DeVELoPEd heRe wIll apply.
LR mOmenTS redUce tHAt SmoOThing Bias tHaT ReSUlts from first step noNpArametRic esTimation relatIve to origiNAL Moment coNditIOnS. there are other sOurceS of bias ariSIng from nOnlinEarity of Moment conDITions in tHe fIrsT stEp aND ThE empirical disTRIbutIoN. CattanEo aNd JanssOn (2017) aNd CAttAneO, JAnsson, and ma (2017) give usEfUl BoOtStrAp and JAckknife MeThoDs ThaT reduCE nonliNeariTy biAs. neWEy aNd RobinS (2017) ShOW That OnE cAn alSo rEmOve thIs biAS by Cross fiTting in soMe sETtinGs. we Allow foR cross-fitting In This paper.
SEcTioN 2 descrIBEs the genEral construction of LR momENt functIonS for sEmipArametric gMM. sectioN 3 giVEs LR moMent coNditiOnS foR DYnamiC DIsCreTe Choice. SectION 4 shOws hoW tO estImate thE first step influencE AdjUstment. SectioN 5 giVes nOVEl ClaSSeS Of Dr mOMenT FUnctions and charActerizes dOuBLe Robustness. sEctIoN 6 gives aN orthogOnal iNStrumenT construcTion of LR mOmEnts BASed On conditioNal momenT restrictIOns. SeCTiOn 7 proVidEs simpLe And GenerAl asymPTotIc theOry for lR EstimaTors.
LOcAlly RobuSt Moment Functions
===============================
The subJect of This pApeR is GMM estImaTOrs Of parametErs wHere the samPle MomEnt fuNctIOns dePend ON a FirST step NonpARametric oR LaRge DIMeNsional estiMATOr. WE refeR to THese esTimaTors as semiparametRIc. We could also rEfer TO TheM as gmM whErE first step estiMatOrS ARe pluggeD iN the moments. this termInOLogy sEems awKward tHough, so WE SiMPly refEr to TheM as semipaRamEtRIc GMM esTiMaTOrs. We dEnotE sUch an eStimatOR by $\hAT{\Beta}$, which is a funcTion oF THe datA $Z_{1},...,z_{n}$ Where $N$ | .
A main contribution issimple and gene ral as ym ptot ic t heory for LR e s tima tors that use cross-fi tting i n the co nstru ction o f t h e av er ag e m om e nt s. Th istheoryis based o n t he structure o f L R moment c ond itions as an id entify in g m o mentcon ditio n depe n ding o n one fir st step p l us an i n f lu ence adjustment thatc an depend on an a dditio na l f i r stste p. We give a rema i nder de c om p o s iti o n that leadsto mean squ a reconsis te ncy condit ionsfo r fi rst steps p lusa few rea dily i n terpret a ble rat e cond iti ons . Fo r D Rest im a tor s t her e is only on era te co ndit i o n , ona p rodu ct of sample remai nde rs f r omtwo f irststep e stima tors,leadi ng to particularl y si mple cond iti on s.Th is si m plicit y m oti vates o ur incl u sio no f re sults for DR estim at o r s. This as ymptot i cth e ory is a ls o u sefu l for e xist i ng momentcondit i on sthat ar ealread ykno wnto be LR.Whenev er the m oment condition canb e decomposedi nt o an iden tif ying moment con d itio n de p en din g on o ne fi rs t s t ep and an influence f unctio n ter m that may de pend on tw o f irst ste ps t h es imple and gene ral r egularityc ondition s dev eloped h ere willa p ply.
LR mo men tsred u c ethat smoothin g bias t hat res ult s fromfir stste p n on parametri c estima ti on r el ati ve to original m ome nt co nditi o ns. Th ere a re o th er sou rces of bi a s ari si ng fro m n on linea rity ofmomentcondition s i n the f ir st step and the empi ri cal distri bu tio n. Cat t a neo andJansson (2017) and Catt a neo, Ja nss on, a nd M a (2017)giv e usef ulb ootstr ap and jack kn ife m ethod s th atre duce nonli n e ari ty bi as . Ne wey and Robins (2017) sho w th at one can al soremo v e t his bi a s b yc ros s fitting in some settings. W e a llow for c r oss -f ittingin this pape r .
Sect ion 2 des cribes th egene r a l c onstructio n of LRmoment fu n ction s f or se mip aramet ri c G MM. S ection 3 g ivesLR mom en t cond ition sfor dyna mic discrete choice. Se ction4 sho wshow to es tim a tethe first ste p influenc e a dju stmen t.S ectio n 5g iv esn ovelclas s es of DRm om ent f un ctions andc h a rac teriz esd oublerobu stness. Section 6 gives an ortho gona l ins tru m entco nstruction ofLRmo m e nts base don conditio nal mome nt restr iction s. Sec tion 7p r ov i des si mple an d general as ym p totic t he or y for L R es ti mators .
Loc a llyR o bust Moment Func tions = ===== = === ===== == ======= = ==== ==
The su bject of th is pap er i s GMM estima to rs ofpar am eters wher e the samp le mo ment fu nc tion s d ependon a f irststep n onp arametric o rl ar ge dim ensi onales tima tor. We r e fer to t hes e estima to rsa s semip a ra m e tric. We c oul d als o refer to t h em a s G M M whe re fir st ste p estim a tor sare plu gge d in the mo ments. Th i s t er mino logy see ms awk wa rdth ough, so we sim ply refe r to the m as sem ipa r amet r ic GMM e stimators. W e den ot e such a n est imatorb y $ \ha t{\ beta }$, whi c h is a function of the d ata$z_{1 } ,.. .,z_{ n}$ w he re $n$ | .
A main_contribution is_simple and general asymptotic_theory for_LR_estimators that_use_cross-fitting in the_construction of the_average moments. This theory_is based on_the_structure of LR moment conditions as an identifying moment condition depending on one first_step_plus an_influence_adjustment_that can depend on an_additional first step. We give_a remainder_decomposition that leads to mean square consistency conditions_for_first steps plus_a few readily interpretable rate conditions. For DR estimators_there is only one rate condition,_on a product_of_sample_remainders from two first_step estimators, leading to particularly simple_conditions. This simplicity motivates our inclusion_of results for DR estimators. This asymptotic_theory is also useful for existing_moment conditions that are already_known to_be LR. Whenever the moment_condition can be_decomposed into_an identifying moment_condition depending on one first step_and an influence_function term that may depend on_two_first steps the_simple_and_general regularity_conditions developed here_will_apply.
LR moments_reduce_that smoothing bias that results from_first_step nonparametric estimation relative to original moment_conditions. There are other_sources_of bias arising from_nonlinearity of moment conditions in_the first step and the empirical_distribution. Cattaneo_and Jansson_(2017) and Cattaneo, Jansson, and Ma (2017) give useful bootstrap and_jackknife methods that reduce nonlinearity bias._Newey and Robins (2017)_show that_one_can also remove_this_bias by_cross fitting in some settings. We allow_for cross-fitting_in this paper.
Section 2 describes the_general construction of LR_moment_functions for semiparametric GMM. Section 3_gives LR moment conditions for dynamic_discrete choice. Section 4 shows_how_to_estimate the first step influence_adjustment. Section 5 gives novel classes_of DR moment_functions and characterizes double robustness. Section 6_gives_an orthogonal instrument construction of LR_moments_based on conditional moment restrictions. Section_7_provides_simple and general asymptotic theory_for LR estimators.
Locally Robust Moment Functions
===============================
The_subject of this paper is GMM estimators of parameters_where the sample_moment functions depend on a_first_step_nonparametric or large dimensional estimator. We refer to these estimators_as semiparametric._We could also_refer to them as GMM where first step estimators are_plugged in the moments. This terminology seems awkward_though, so we simply refer to them as semiparametric GMM estimators._We denote such an estimator by $\hat{\beta}$, which_is a function of the data $z_{1},...,z_{n}$_where $n$ |
in Quintessence*, *Astropart. Phys.* [**54**]{} (2014) 125
M. Meehan and I. Whittingham, *Asymmetric dark matter in braneworld cosmology*, *JCAP* [**06**]{} (2014) 018
G. Gelmini, *Experimental signatures of non-standard pre-BBN cosmologies*, *Nucl. Phys. Proc. Suppl.* [**194**]{} (2009) 63
D. Langlois, *Brane Cosmology: An Introduction*, *Prog. Theor. Phys. Suppl.* [**148**]{} (2003) 181 R. Maartens and K. Koyama, *Brane-World Gravity*, *Living Rev. Rel.* [**13**]{} (2010) 5
L. Randall and R. Sundrum, *An Alternative to Compactification*, *Phys. Rev. Lett.* [**83**]{} (1999) 4690.
N. Okada and O. Seto, *Relic Density of Dark Matter in Brane World Cosmology*, *Phys. Rev. D* [**70**]{} (2004) 083531.
T. Nihei, N. Okada and O. Seto, *Neutralino dark matter in brane world cosmology*, *Phys. Rev. D* [**71**]{} (2005) 063535.
T. Nihei, N. Okada and O. Seto, *Light W-ino dark matter in brane world cosmology*, *Phys. Rev. D* [**73**]{} (2006) 063518.
E. Abou El Dahab and S. Khalil, *Cold dark matter in brane cosmology scenario*, *JHEP09* (2006) 042
W.-L. Guo and X. Zhang, *Constraints on dark matter annihilation cross section in the brane-world and quintessence scenarios*, *Phys. Rev. D* [**79**]{} (2009) 115023
J.-F. Dufaux, J. E. Lidsey, R. Maartens and M. Sami, *Cosmological perturbations from braneworld inflation with a Gauss-Bonnet term*, *Phys. Rev. D* [**70**]{} (2004) 083525
S. Tsujikawa, M. Sami and R. Maartens, *Observational constraints on braneworld inflation | in Quintessence *, * Astropart. Phys. * [ * * 54 * * ] { } (2014) 125
M. Meehan and I. Whittingham, * Asymmetric dark matter in braneworld cosmology *, * JCAP * [ * * 06 * * ] { } (2014) 018
G. Gelmini, * Experimental signatures of non - standard pre - BBN cosmology *, * Nucl. Phys. Proc. Suppl. * [ * * 194 * * ] { } (2009) 63
D. Langlois, * Brane Cosmology: An initiation *, * Prog. Theor. Phys. Suppl. * [ * * 148 * * ] { } (2003) 181 R. Maartens and K. Koyama, * Brane - World Gravity *, * Living Rev. Rel. * [ * * 13 * * ] { } (2010) 5
L. Randall and R. Sundrum, * An Alternative to Compactification *, * Phys. Rev. Lett. * [ * * 83 * * ] { } (1999) 4690.
N. Okada and O. Seto, * Relic Density of Dark Matter in Brane World Cosmology *, * Phys. Rev. D * [ * * 70 * * ] { } (2004) 083531.
T. Nihei, N. Okada and O. Seto, * Neutralino dark matter in brane world cosmology *, * Phys. Rev. D * [ * * 71 * * ] { } (2005) 063535.
T. Nihei, N. Okada and O. Seto, * Light W - ino benighted matter in brane world cosmology *, * Phys. Rev. D * [ * * 73 * * ] { } (2006) 063518.
E. Abou El Dahab and S. Khalil, * Cold dark topic in brane cosmology scenario *, * JHEP09 * (2006) 042
W.-L. Guo and X. Zhang, * Constraints on dark matter annihilation crisscross section in the brane - world and quintessence scenario *, * Phys. Rev. D * [ * * 79 * * ] { } (2009) 115023
J.-F. Dufaux, J. E. Lidsey, R. Maartens and M. Sami, * Cosmological perturbations from braneworld ostentation with a Gauss - Bonnet term *, * Phys. Rev. D * [ * * 70 * * ] { } (2004) 083525
S. Tsujikawa, M. Sami and R. Maartens, * Observational constraint on braneworld inflation | in Quintessence*, *Astropart. Khys.* [**54**]{} (2014) 125
M. Meehan aue I. Whmttinghzm, *Asymmdtric dark matter in branewocld xosmooogy*, *JCAP* [**06**]{} (2014) 018
G. Gelmini, *Dxperimennal signarurew of non-staisard prc-YBN ckdmolmjies*, *Nucl. Phys. Kroc. Suppl.* [**194**]{} (2009) 63
D. Langlois, *Brdnd Eosmology: An Introduction*, *Prog. Theor. Phys. Sipol.* [**148**]{} (2003) 181 R. Maartegs amq K. Ioyama, *Brane-World Gravity*, *Living Rsv. Rel.* [**13**]{} (2010) 5
L. Randall amd R. Sundrum, *An Alternativf to Compactification*, *Ohys. Rev. Lejf.* [**83**]{} (1999) 4690.
B. Okada and U. Seto, *Relpe Density or Dark Matter in Brane World Cormoloyy*, *Phys. Rev. D* [**70**]{} (2004) 083531.
H. Nihei, N. Okeda anq O. Seto, *Neubgalino gark mayter in brane eornd xosmology*, *Phys. Rev. D* [**71**]{} (2005) 063535.
T. Nihei, N. Okada wnd O. Setm, *Might W-ino dark mqtrer it brdne dirla ckskomogy*, *Pjys. Rev. D* [**73**]{} (2006) 063518.
E. Abou El Dayab and S. Khalil, *Cokd eark matter ih branq sosmology scenario*, *JHEP09* (2006) 042
W.-L. Guo and X. Zvanf, *Constraints on dark mqtter annihilation crlss sectijn in the brane-world and quintessence scenarios*, *Pvys. Rxv. D* [**79**]{} (2009) 115023
M.-F. Djdakx, J. E. Lidsey, R. Maartens and M. Sami, *Cosmologiszl pvrturbations from braneworld ijfkwtion with a Eauss-Bonnst term*, *Phys. Rev. D* [**70**]{} (2004) 083525
S. Tsojikawq, M. Sami wnd T. Maartens, *Observational cobstraints on vraneworld inflatiln | in Quintessence*, *Astropart. Phys.* [**54**]{} (2014) 125 and Whittingham, *Asymmetric matter in braneworld G. *Experimental signatures of pre-BBN cosmologies*, *Nucl. Proc. Suppl.* [**194**]{} (2009) 63 D. *Brane Cosmology: An Introduction*, *Prog. Theor. Phys. Suppl.* [**148**]{} (2003) 181 R. Maartens K. Koyama, *Brane-World Gravity*, *Living Rev. Rel.* [**13**]{} (2010) 5 L. Randall and Sundrum, Alternative Compactification*, Rev. Lett.* [**83**]{} (1999) 4690. N. Okada and O. Seto, *Relic Density of Dark Matter in World Cosmology*, *Phys. Rev. D* [**70**]{} (2004) 083531. Nihei, N. Okada and Seto, *Neutralino dark matter in world *Phys. Rev. [**71**]{} 063535. Nihei, N. Okada O. Seto, *Light W-ino dark matter in brane world cosmology*, *Phys. Rev. D* [**73**]{} (2006) 063518. E. El Dahab Khalil, *Cold matter brane scenario*, *JHEP09* (2006) Guo and X. Zhang, *Constraints on cross section in the brane-world and quintessence scenarios*, Rev. D* (2009) 115023 J.-F. Dufaux, J. E. R. Maartens and M. Sami, *Cosmological perturbations from inflation with a Gauss-Bonnet term*, *Phys. Rev. D* [**70**]{} (2004) 083525 S. Tsujikawa, M. Sami Maartens, *Observational constraints on inflation | in Quintessence*, *Astropart. PhYs.* [**54**]{} (2014) 125
M. Meehan aNd I. WhIttIngHaM, *AsyMmetRic dark matter iN BranEworld cosmology*, *JCAP* [**06**]{} (2014) 018
G. GeLmini, *exPErimENtAl sigNatures OF nON-StaNdArD prE-Bbn cOsmolOgiEs*, *Nucl. PHys. Proc. SupPl.* [**194**]{} (2009) 63
D. laNglois, *Brane COSmOlogy: An IntRodUction*, *Prog. ThEor. phys. SuPpL.* [**148**]{} (2003) 181 R. MAArtenS anD K. KoyAma, *BraNE-World gravity*, *LiViNG Rev. ReL.* [**13**]{} (2010) 5
l. RandalL ANd r. SunDrum, *An Alternative TO COMpactification*, *phys. ReV. LETt.* [**83**]{} (1999) 4690.
n. oKadA anD O. Seto, *ReliC DEnsitY Of Dark MATtER IN BrANe World CosmolOgy*, *Phys. Rev. D* [**70**]{} (2004) 083531.
t. nihEi, N. OkaDa And o. seto, *NeUtralInO DarK matter in brAne wOrld cosmoLogy*, *PhYS. Rev. D* [**71**]{} (2005) 063535.
T. NIHei, N. OkaDa and O. setO, *LiGht W-INo DaRk mAtTEr iN BrAne WOrlD cosmoloGy*, *phYs. Rev. d* [**73**]{} (2006) 063518.
E. AbOU eL dahaB anD S. KhAlil, *COld dark matter In bRane COsmOlogy ScenaRio*, *JhEp09* (2006) 042
W.-L. GuO and X. ZHang, *COnStraints on dark mAtteR annihilaTioN cRosS sEctioN In the bRanE-woRld and qUintessENce ScENARiOs*, *Phys. Rev. D* [**79**]{} (2009) 115023
J.-F. Dufaux, j. E. lIDsEy, R. MaartEns and m. saMi, *cOsmologiCaL peRturBATions From BRaNeworld iNflatiON wItH a Gauss-boNnet teRm*, *phyS. ReV. D* [**70**]{} (2004) 083525
S. TsUJikaWa, M. SamI and R. MaaRtens, *oBservational coNStraints on braNEwORLd INflaTioN | in Quintessence*, *Astrop art. Phys. * [** 54* *]{ }(201 4) 1 25
M. Meehana nd I . Whittingham, *Asymme tricda r k ma t te r inbranewo r ld c osm ol og y*, * J CA P* [* *06 **]{} ( 2014) 018
G. G elmini, *Exp e ri mental sig nat ures of non- sta ndardpr e-B B N cos mol ogies *, *Nu c l. Phy s. Proc.Su p pl.* [ * *194**] { } ( 2009 ) 63
D. Langlois , * B rane Cosmology : An I nt r od u c tio n*, *Prog. Th eo r. Ph y s. Supp l .* [ * *14 8 **]{} (2003)181 R. Maar t ens and K .Koy a ma, *B rane- Wo r ldGravity*, * Livi ng Rev. R el.* [ * *13**]{ } (2010) 5
L. Ra nda ll a n dR. Su nd r um, *A n A l ter native t oCo mpact ific a t i o n*,*Ph ys.Rev.Lett.* [**83* *]{ } (1 9 99) 4690 .
N. Oka da andO. Set o, *R el ic Density of D arkMatter in Br an e W or ld Co s mology *,*Ph ys. Rev . D* [* * 70* *] { } (2 004) 083531.
T. N ih e i ,N. Okada and O . S et o , *Neutr al ino dar k matte r in br ane worl d cosm o lo gy *, *Phy s. Rev.D* [* *71 **]{} (200 5) 063 535.
T. Nihe i , N. Okada and O. Seto, *Lig h tW - in o dar k m atter in br anew orld cos m ol ogy * , *Ph ys. R ev . D * [**73**]{} (2006)06 3518.
E. A bou El Dahaband S. Kha l i l , *Colddark ma t ter in brane c osmol ogy scenar i o*, *JHE P09*(2006) 0 42
W.-L. G uo and X . Z han g,*Co n s tr aints on dark m atte rannihil ati on cros s s ect ion in t he brane- world an dqu in te sse nce s c enarios* ,*Ph ys . R ev. D * [**79 **]{} (20 09 )1 150 23
J.- F .D u faux ,J. E.Lid se y, R. Maa r ten s and M . Sami, * Cos m olog ic al pertur bations frombr aneworld i nf lat ion wi t h a Gauss -Bonnet term*, *Phys. R e v. D* [ **7 0**]{ } (2 004) 0835 25
S. Ts uji k awa, M . Sami andR. Ma a r tens, * Ob ser va tional con s t rai nts o nbran eworldinflation | in_Quintessence*, *Astropart._Phys.* [**54**]{} (2014) 125
M._Meehan and_I._Whittingham, *Asymmetric_dark_matter in braneworld_cosmology*, *JCAP* [**06**]{}_(2014) 018
G. Gelmini, *Experimental_signatures of non-standard_pre-BBN_cosmologies*, *Nucl. Phys. Proc. Suppl.* [**194**]{} (2009) 63
D. Langlois, *Brane Cosmology: An Introduction*, *Prog._Theor._Phys. Suppl.*_[**148**]{}_(2003)_181 R. Maartens and K._Koyama, *Brane-World Gravity*, *Living Rev._Rel.* [**13**]{}_(2010) 5
L. Randall and R. Sundrum, *An Alternative_to_Compactification*, *Phys. Rev._Lett.* [**83**]{} (1999) 4690.
N. Okada and O. Seto, *Relic_Density of Dark Matter in Brane_World Cosmology*, *Phys._Rev._D*_[**70**]{} (2004) 083531.
T. Nihei,_N. Okada and O. Seto, *Neutralino_dark matter in brane world cosmology*,_*Phys. Rev. D* [**71**]{} (2005) 063535.
T. Nihei,_N. Okada and O. Seto, *Light_W-ino dark matter in brane_world cosmology*,_*Phys. Rev. D* [**73**]{} (2006)_063518.
E. Abou El_Dahab and_S. Khalil, *Cold_dark matter in brane cosmology scenario*,_*JHEP09* (2006) 042
W.-L._Guo and X. Zhang, *Constraints on_dark_matter annihilation cross_section_in_the brane-world_and quintessence scenarios*,_*Phys._Rev. D*_[**79**]{}_(2009) 115023
J.-F. Dufaux, J. E. Lidsey,_R._Maartens and M. Sami, *Cosmological perturbations from_braneworld inflation with a_Gauss-Bonnet_term*, *Phys. Rev. D*_[**70**]{} (2004) 083525
S. Tsujikawa, M._Sami and R. Maartens, *Observational constraints_on braneworld_inflation |
&0.02 &41.69\
G24.94+0.07 & 18 36 31.333 & 0.006 & -07 04 19.968 & 0.007 & 0.60 &0.02 &41.69\
G24.94+0.07 & 18 36 31.812 & 0.014 & -07 04 23.244 & 0.016 & 0.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.705 & 0.008 & -07 04 10.801 & 0.010 & 0.43 &0.02 &41.85\
G24.94+0.07 & 18 36 31.672 & 0.006 & -07 04 09.024 & 0.007 & 0.61 &0.02 &41.85\
G24.94+0.07 & 18 36 31.591 & 0.003 & -07 04 07.731 & 0.003 & 1.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.523 & 0.031 & -07 04 07.774 & 0.036 & 0.12 &0.02 &41.85\
G24.94+0.07 & 18 36 31.401 & 0.029 & -07 04 16.537 & 0.035 & 0.12 &0.02 &41.85\
G24.94+0.07 & 18 36 31.378 & 0.011 & -07 04 15.677 & 0.014 & 0.30 &0.02 &41.85\
G24.94+0.07 & 18 36 31.334 & 0.007 & -07 04 19.962 & 0.008 & 0.53 &0.02 &41.85\
G24.94+0.07 & 18 36 31.706 & 0.008 & -07 04 10.805 & 0.009 & 0.47 &0.02 &42.02\
G24.94+0.07 & 18 36 31.672 & 0.008 & -07 04 09.028 & 0.009 & 0.46 &0.02 &42.02\
G24.94+0.07 & 18 36 31.591 & 0.003 & -07 04 07.730 & | & 0.02 & 41.69\
G24.94 + 0.07 & 18 36 31.333 & 0.006 & -07 04 19.968 & 0.007 & 0.60 & 0.02 & 41.69\
G24.94 + 0.07 & 18 36 31.812 & 0.014 & -07 04 23.244 & 0.016 & 0.27 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.705 & 0.008 & -07 04 10.801 & 0.010 & 0.43 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.672 & 0.006 & -07 04 09.024 & 0.007 & 0.61 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.591 & 0.003 & -07 04 07.731 & 0.003 & 1.27 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.523 & 0.031 & -07 04 07.774 & 0.036 & 0.12 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.401 & 0.029 & -07 04 16.537 & 0.035 & 0.12 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.378 & 0.011 & -07 04 15.677 & 0.014 & 0.30 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.334 & 0.007 & -07 04 19.962 & 0.008 & 0.53 & 0.02 & 41.85\
G24.94 + 0.07 & 18 36 31.706 & 0.008 & -07 04 10.805 & 0.009 & 0.47 & 0.02 & 42.02\
G24.94 + 0.07 & 18 36 31.672 & 0.008 & -07 04 09.028 & 0.009 & 0.46 & 0.02 & 42.02\
G24.94 + 0.07 & 18 36 31.591 & 0.003 & -07 04 07.730 & | &0.02 &41.69\
G24.94+0.07 & 18 36 31.333 & 0.006 & -07 04 19.968 & 0.007 & 0.60 &0.02 &41.69\
G24.94+0.07 & 18 36 31.812 & 0.014 & -07 04 23.244 & 0.016 & 0.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.705 & 0.008 & -07 04 10.801 & 0.010 & 0.43 &0.02 &41.85\
G24.94+0.07 & 18 36 31.672 & 0.006 & -07 04 09.024 & 0.007 & 0.61 &0.02 &41.85\
G24.94+0.07 & 18 36 31.591 & 0.003 & -07 04 07.731 & 0.003 & 1.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.523 & 0.031 & -07 04 07.774 & 0.036 & 0.12 &0.02 &41.85\
G24.94+0.07 & 18 36 31.401 & 0.029 & -07 04 16.537 & 0.035 & 0.12 &0.02 &41.85\
G24.94+0.07 & 18 36 31.378 & 0.011 & -07 04 15.677 & 0.014 & 0.30 &0.02 &41.85\
G24.94+0.07 & 18 36 31.334 & 0.007 & -07 04 19.962 & 0.008 & 0.53 &0.02 &41.85\
G24.94+0.07 & 18 36 31.706 & 0.008 & -07 04 10.805 & 0.009 & 0.47 &0.02 &42.02\
J24.94+0.07 & 18 36 31.672 & 0.008 & -07 04 09.028 & 0.009 & 0.46 &0.02 &42.02\
G24.94+0.07 & 18 36 31.591 & 0.003 & -07 04 07.730 & | &0.02 &41.69\ G24.94+0.07 & 18 36 31.333 & 04 19.968 0.007 & 0.60 36 & 0.014 & 04 23.244 & & 0.27 &0.02 &41.85\ G24.94+0.07 & 36 31.705 & 0.008 & -07 04 10.801 & 0.010 & 0.43 &0.02 G24.94+0.07 & 18 36 31.672 & 0.006 & -07 04 09.024 & 0.007 0.61 &41.85\ & 36 31.591 & 0.003 & -07 04 07.731 & 0.003 & 1.27 &0.02 &41.85\ G24.94+0.07 & 36 31.523 & 0.031 & -07 04 07.774 0.036 & 0.12 &0.02 G24.94+0.07 & 18 36 31.401 0.029 -07 04 & & &0.02 &41.85\ G24.94+0.07 18 36 31.378 & 0.011 & -07 04 15.677 & 0.014 & 0.30 &0.02 &41.85\ G24.94+0.07 & 36 31.334 & -07 19.962 0.008 0.53 &0.02 &41.85\ 18 36 31.706 & 0.008 & & 0.009 & 0.47 &0.02 &42.02\ G24.94+0.07 & 36 31.672 0.008 & -07 04 09.028 & & 0.46 &0.02 &42.02\ G24.94+0.07 & 18 36 & 0.003 & -07 04 07.730 & | &0.02 &41.69\
G24.94+0.07 & 18 36 31.333 & 0.006 & -07 04 19.968 & 0.007 & 0.60 &0.02 &41.69\
G24.94+0.07 & 18 36 31.812 & 0.014 & -07 04 23.244 & 0.016 & 0.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.705 & 0.008 & -07 04 10.801 & 0.010 & 0.43 &0.02 &41.85\
G24.94+0.07 & 18 36 31.672 & 0.006 & -07 04 09.024 & 0.007 & 0.61 &0.02 &41.85\
G24.94+0.07 & 18 36 31.591 & 0.003 & -07 04 07.731 & 0.003 & 1.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.523 & 0.031 & -07 04 07.774 & 0.036 & 0.12 &0.02 &41.85\
G24.94+0.07 & 18 36 31.401 & 0.029 & -07 04 16.537 & 0.035 & 0.12 &0.02 &41.85\
G24.94+0.07 & 18 36 31.378 & 0.011 & -07 04 15.677 & 0.014 & 0.30 &0.02 &41.85\
G24.94+0.07 & 18 36 31.334 & 0.007 & -07 04 19.962 & 0.008 & 0.53 &0.02 &41.85\
G24.94+0.07 & 18 36 31.706 & 0.008 & -07 04 10.805 & 0.009 & 0.47 &0.02 &42.02\
G24.94+0.07 & 18 36 31.672 & 0.008 & -07 04 09.028 & 0.009 & 0.46 &0.02 &42.02\
G24.94+0.07 & 18 36 31.591 & 0.003 & -07 04 07.730 & | &0.02 &41.69\
G24.94+0.07 & 18 36 3 1.333 &0.0 06 & - 07 0 4 19.968 & 0.0 0 7 &0.60 &0.02 &41.69\
G24 .94+0 .0 7 & 1 8 3 6 31. 812 & 0 . 01 4 & - 07 0 4 2 3. 2 44 & 0. 016 & 0.27 &0.02 &41 .85 \G24.94+0.07& 1 8 36 31.70 5 & 0.008 & -07 04 10.80 1& 0 . 010 & 0. 43 &0 .02 &4 1 .85\
G 24.94+0.0 7& 18 36 31.672& 0. 006& -07 04 09.024 & 0. 0 07 & 0.61 &0.0 2 &41. 85 \
G 2 4 .94 +0. 07 & 18 36 3 1.591 & 0.003 &- 0 7 04 07.731 & 0.00 3 & 1.27 &0 . 02&41.85 \G24 . 94+0.0 7 & 1 83 6 3 1.523 & 0.0 31 & -07 04 0 7.774& 0.036& 0.12 & 0.02 & 41. 85\
G24 . 94 +0 .07 & 183 631. 4 01& 0.029&-0 7 0416.5 3 7 & 0.0 35& 0. 12 &0 .02 &41.85\
G 24. 94+0 . 07& 1836 31 .378 & 0.01 1 & -0 7 0415 .677 & 0.014 &0.30 &0.02 &4 1.8 5\
G2 4. 94+0. 0 7 & 18 36 31 .334 &0.007 & -07 0 4 1 9. 962 & 0.008 & 0.53 & 0 . 02 &41.85\
G24.9 4 +0 .0 7 & 18 36 3 1.7 06 & 0 .008& -0 7 0 4 10.805 & 0.0 0 9&0.47 &0 .0 2 &42. 02 \
G 24. 94+0. 0 7 &18 3631.672 & 0.00 8 & -07 04 09.0 2 8 & 0.009 & 0 . 46 & 0. 0 2 &4 2.0 2\
G24.94+0 .07& 1836 3 1 .5 91& 0.00 3 & - 07 04 07.730 & | &0.02_&41.69\
G24.94+0.07 &_18 36 31.333 &_0.006 &_-07_04 19.968_&_0.007 & 0.60_&0.02 &41.69\
G24.94+0.07 &_18 36 31.812 &_0.014 & -07_04_23.244 & 0.016 & 0.27 &0.02 &41.85\
G24.94+0.07 & 18 36 31.705 & 0.008 &_-07_04 10.801_&_0.010_& 0.43 &0.02 &41.85\
G24.94+0.07 &_18 36 31.672 & 0.006_& -07_04 09.024 & 0.007 & 0.61 &0.02 &41.85\
G24.94+0.07_&_18 36 31.591_& 0.003 & -07 04 07.731 & 0.003 &_1.27 &0.02 &41.85\
G24.94+0.07 & 18 36_31.523 & 0.031_&_-07_04 07.774 & 0.036_& 0.12 &0.02 &41.85\
G24.94+0.07 & 18_36 31.401 & 0.029 & -07_04 16.537 & 0.035 & 0.12 &0.02_&41.85\
G24.94+0.07 & 18 36 31.378 &_0.011 & -07 04 15.677_& 0.014_& 0.30 &0.02 &41.85\
G24.94+0.07 &_18 36 31.334_& 0.007_& -07 04_19.962 & 0.008 & 0.53 &0.02_&41.85\
G24.94+0.07 & 18_36 31.706 & 0.008 & -07_04_10.805 & 0.009_&_0.47_&0.02 &42.02\
G24.94+0.07_& 18 36_31.672_& 0.008_&_-07 04 09.028 & 0.009 &_0.46_&0.02 &42.02\
G24.94+0.07 & 18 36 31.591 &_0.003 & -07 04_07.730_& |
1}{\Gamma(\frac{4-n}{2})} \int_0^{{\infty}} k(v) v^{\frac{4-n}{2}-1}
dv \\
&= f(0) \Lambda^0 a_4(D^2) + 2 f_2 \Lambda^2 a_2(D^2) + 2\Lambda^4 f_4 a_0(D^2).
\label{eq:Diracexpansion}\end{aligned}$$ where the $f_k$ are moments of the function $f$: $$\begin{aligned}
f_{k} := \int_{0}^{\infty} f(w)w^{k-1}dw; \qquad (k>0) \nonumber. \end{aligned}$$
\[lem:ED2\] For the spectral triple ${(\Gamma(M,B), L^2(M,B\otimes S),D_B,J,\gamma_B)}$, the square of the fluctuated Dirac operator $D_A^2$ is locally of the form $-g_{\mu \nu} \partial_{\mu} \partial_{\nu} + K_{\mu} \partial_{\mu} + L$ and we have the following expressions for $\Omega_{\mu\nu}$ and $E$: $$\begin{aligned}
E = -\frac{1}{4} R \otimes 1_{N^2} - \sum_{\mu < \nu} \gamma^{\mu} \gamma^{\nu} \otimes F_{\mu \nu} \nonumber \\
\Omega_{\mu \nu} = \frac{1}{4} R^{ab}_{\mu \nu} \gamma_{ab} \otimes 1_{n^2} + id_4 \otimes F_{\mu \nu} \nonumber,\end{aligned}$$ where $F_{\mu \nu}$ is the curvature of the connection $\nabla^B + \bA^{pert}$.
This result allows us to compute the bosonic spectral action for the fluctuated Dirac operator $D_A$, essentially reducing the computation in terms of a local trivialization to the trivial case of [@ChamseddineConnes] with the following result.
\[thm:actionnontriv\] For the spectral triple ${(\Gamma(M,B), L^2(M,B\otimes S),D_B,J,\gamma_B)}$, the spectral action equals the | 1}{\Gamma(\frac{4 - n}{2 }) } \int_0^{{\infty } } k(v) v^{\frac{4 - n}{2}-1 }
dv \\
& = f(0) \Lambda^0 a_4(D^2) + 2 f_2 \Lambda^2 a_2(D^2) + 2\Lambda^4 f_4 a_0(D^2).
\label{eq: Diracexpansion}\end{aligned}$$ where the $ f_k$ are moments of the function $ f$: $ $ \begin{aligned }
f_{k }: = \int_{0}^{\infty } f(w)w^{k-1}dw; \qquad (k>0) \nonumber. \end{aligned}$$
\[lem: ED2\ ] For the spectral triple $ { (\Gamma(M, B), L^2(M, B\otimes S),D_B, J,\gamma_B)}$, the public square of the fluctuated Dirac hustler $ D_A^2 $ is locally of the phase $ -g_{\mu \nu } \partial_{\mu } \partial_{\nu } + K_{\mu } \partial_{\mu } + L$ and we have the following expression for $ \Omega_{\mu\nu}$ and $ E$: $ $ \begin{aligned }
E = -\frac{1}{4 } R \otimes 1_{N^2 } - \sum_{\mu < \nu } \gamma^{\mu } \gamma^{\nu } \otimes F_{\mu \nu } \nonumber \\
\Omega_{\mu \nu } = \frac{1}{4 } R^{ab}_{\mu \nu } \gamma_{ab } \otimes 1_{n^2 } + id_4 \otimes F_{\mu \nu } \nonumber,\end{aligned}$$ where $ F_{\mu \nu}$ is the curvature of the connection $ \nabla^B + \bA^{pert}$.
This consequence allows us to compute the bosonic apparitional legal action for the fluctuated Dirac operator $ D_A$, basically reducing the computation in terms of a local trivialization to the trivial shell of [ @ChamseddineConnes ] with the following result.
\[thm: actionnontriv\ ] For the spectral triple $ { (\Gamma(M, B), L^2(M, B\otimes S),D_B, J,\gamma_B)}$, the spectral action equal the | 1}{\Gamla(\frac{4-n}{2})} \int_0^{{\infty}} k(v) v^{\frag{4-n}{2}-1}
dv \\
&= f(0) \Lambda^0 a_4(B^2) + 2 f_2 \Lembda^2 a_2(S^2) + 2\Lambdx^4 f_4 a_0(D^2).
\label{eq:Diracexpansion}\eid{alugned}$$ where the $f_k$ are momevts of thv functiob $f$: $$\uegin{aligned}
h_{i} := \int_{0}^{\ikyty} f(s)a^{k-1}dw; \wquad (k>0) \nonumbgr. \end{aligneg}$$
\[lem:ED2\] For the soeetral triple ${(\Gamma(M,B), L^2(M,B\otimes S),D_B,J,\gwmma_B)}$, tne square of the flubttates Dirac operator $D_A^2$ is locally of fhe fork $-g_{\mu \nu} \partoal_{\mu} \partial_{\nu} + K_{\mu} \partiwl_{\mu} + L$ and we have thf following expwwssions for $\Umega_{\mu\nu}$ and $E$: $$\begin{amigned}
E = -\frac{1}{4} R \otimes 1_{N^2} - \sum_{\mu < \nu} \yamma^{\mu} \gamna^{\bu} \ljimes F_{\mu \nu} \nonuiber \\
\Omega_{\mu \nu} = \fraw{1}{4} R^{ab}_{\mu \nu} \gamma_{ab} \otlmes 1_{i^2} + ie_4 \otimes F_{\mu \nu} \nonumuer,\end{aligned}$$ where $S_{\mu \nu}$ is tke curvature of the cinbectimn $\ndbla^C + \bX^{pedt}$.
Vhia resupt ellows us tk compute tye bosonic spectral asnoon for the rluctuwtqd Dirac operator $D_A$, essentially reducitg fhe computation in termw of a local trivialieation to ehe trivial case of [@ChamseddineConnes] with the fonlowiig rewujg.
\[rhl:actionnontriv\] For the spectral triple ${(\Gamma(M,F), L^2(K,B\ptimes S),D_B,J,\gamia_B)}$, the spevtgak action equalr the | 1}{\Gamma(\frac{4-n}{2})} \int_0^{{\infty}} k(v) v^{\frac{4-n}{2}-1} dv \\ &= a_4(D^2) 2 f_2 a_2(D^2) + 2\Lambda^4 $f_k$ moments of the $f$: $$\begin{aligned} f_{k} \int_{0}^{\infty} f(w)w^{k-1}dw; \qquad (k>0) \nonumber. \end{aligned}$$ For the spectral triple ${(\Gamma(M,B), L^2(M,B\otimes S),D_B,J,\gamma_B)}$, the square of the fluctuated Dirac $D_A^2$ is locally of the form $-g_{\mu \nu} \partial_{\mu} \partial_{\nu} + K_{\mu} \partial_{\mu} L$ we the expressions for $\Omega_{\mu\nu}$ and $E$: $$\begin{aligned} E = -\frac{1}{4} R \otimes 1_{N^2} - \sum_{\mu < \nu} \gamma^{\nu} \otimes F_{\mu \nu} \nonumber \\ \Omega_{\mu \nu} \frac{1}{4} R^{ab}_{\mu \nu} \gamma_{ab} 1_{n^2} + id_4 \otimes F_{\mu \nonumber,\end{aligned}$$ $F_{\mu \nu}$ the of connection $\nabla^B + This result allows us to compute the bosonic spectral action for the fluctuated Dirac operator $D_A$, essentially the computation of a trivialization the case of [@ChamseddineConnes] following result. \[thm:actionnontriv\] For the spectral S),D_B,J,\gamma_B)}$, the spectral action equals the | 1}{\Gamma(\frac{4-n}{2})} \int_0^{{\infty}} k(v) v^{\frac{4-N}{2}-1}
dv \\
&= f(0) \Lambda^0 A_4(D^2) + 2 f_2 \LaMbdA^2 a_2(D^2) + 2\laMbda^4 F_4 a_0(D^2).
\lAbel{eq:DiracexpANsioN}\end{aligned}$$ where the $f_k$ aRe momEnTS of tHE fUnctiOn $f$: $$\begiN{AlIGNed}
F_{k} := \InT_{0}^{\inFtY} F(w)W^{k-1}dw; \qQuaD (k>0) \nonumBer. \end{aligNed}$$
\[LeM:ED2\] For the speCTrAl triple ${(\GaMma(m,B), L^2(M,B\otimes S),d_B,J,\Gamma_B)}$, ThE sqUAre of The FluctUated DIRac opeRator $D_A^2$ is LoCAlly of THe form $-g_{\MU \Nu} \PartIal_{\mu} \partial_{\nu} + K_{\mu} \PArTIal_{\mu} + L$ and we havE the foLlOWiNG ExpResSions for $\OmEgA_{\mu\nu}$ ANd $E$: $$\begiN{AlIGNEd}
E = -\FRac{1}{4} R \otimes 1_{N^2} - \suM_{\mu < \nu} \gamma^{\mU} \GamMa^{\nu} \otImEs F_{\MU \nu} \nonUmber \\
\omEGa_{\mU \nu} = \frac{1}{4} R^{ab}_{\mU \nu} \gAmma_{ab} \otiMes 1_{n^2} + id_4 \OTimes F_{\mU \Nu} \nonumBer,\end{AliGneD}$$ wheRE $F_{\Mu \Nu}$ iS tHE cuRVaTurE Of tHe connecTiOn $\Nabla^b + \bA^{pERT}$.
tHis rEsuLt alLows uS to compute the BosOnic SPecTral aCtion For tHe FluctUated DIrac oPeRator $D_A$, essentiaLly rEducing thE coMpUtaTiOn in tERms of a LocAl tRivialiZation tO The TrIVIAl Case of [@ChamseddineCOnNES] wIth the foLlowinG ReSuLT.
\[thm:actiOnNonTriv\] fOR the sPectRAl Triple ${(\GaMma(M,B), L^2(m,b\oTiMes S),D_B,J,\GaMma_B)}$, thE sPecTraL actiON equAls the | 1}{\Gamma(\frac{4-n}{2})}\int_0^{{\ infty }}k(v )v^{\ frac {4-n}{2}-1}
dv \\
& = f(0) \Lambda^0 a_4(D ^2) + 2 f_2\ La mbda^ 2 a_2(D ^ 2) + 2\ La mb da^ 4f _4 a_0( D^2 ).
\lab el{eq:Dira cex pa nsion}\end{a l ig ned}$$ whe rethe $f_k$ ar e m oments o f t h e fun cti on $f $: $$\ b egin{a ligned}
f_{k}: = \int_ { 0 }^ {\in fty} f(w)w^{k-1}d w ;\ qquad (k>0) \n onumbe r. \e n d {al ign ed}$$
\[l em :ED2\ ] For th e s p e c tra l triple ${(\G amma(M,B),L ^2( M,B\ot im esS ),D_B, J,\ga mm a _B) }$, the squ areof the fl uctuat e d Dirac operato r $D_A ^2$ is loc a ll yofth e fo r m$-g _ {\m u \nu} \ pa rt ial_{ \mu} \ p a rtia l_{ \nu} + K_ {\mu} \partia l_{ \mu} + L $ and we h aveth e fol lowing expr es sions for $\Ome ga_{ \mu\nu}$and $ E$: $ $\beg i n{alig ned }
E = -\fr ac{1}{4 } R\o t i m es 1_{N^2} - \sum_{\ mu < \ nu} \gam ma^{\m u }\g a mma^{\nu }\ot imes F _{\mu \nu } \ nonumber \\
\O m eg a_ {\mu \n u} = \fr ac {1} {4} R^{a b }_{\ mu \nu } \gamma _{ab} \otimes 1_{n^2 } + id_4 \otim e sF _ {\ m u \n u}\nonumber,\ end{ a lign ed}$ $ w her e $F_{ \mu \ nu } $i s the curvature ofth e conn ectio n $\nabla^B + \bA^{pert } $ .
This r esul t a l lows us to com putethe bosoni c spectra l act ion forthe fluct u a ted Dira c o per ato r $ D _ A$ , essentially r educ in g the c omp utation in te rms of a local tr ivializa ti on t othe triv i al caseof [@ Ch ams eddin e Connes ] wit h th efo l low ing res u lt .
\[t hm :a ctio nno nt riv\] For the spectr al triple ${ ( \Gam ma (M ,B), L^ 2(M,B\otimesS) ,D_B,J,\ga mm a_B )}$, t h e spectra l action equals the | 1}{\Gamma(\frac{4-n}{2})} \int_0^{{\infty}}_k(v) v^{\frac{4-n}{2}-1}
dv_\\
&= f(0) \Lambda^0 a_4(D^2)_+ 2_f_2_\Lambda^2 a_2(D^2)_+_2\Lambda^4 f_4 a_0(D^2).
\label{eq:Diracexpansion}\end{aligned}$$_where the $f_k$_are moments of the_function $f$: $$\begin{aligned}
__ f_{k} := \int_{0}^{\infty} f(w)w^{k-1}dw; \qquad (k>0) \nonumber. \end{aligned}$$
\[lem:ED2\] For the spectral triple_${(\Gamma(M,B),_L^2(M,B\otimes S),D_B,J,\gamma_B)}$,_the_square_of the fluctuated Dirac operator_$D_A^2$ is locally of the_form $-g_{\mu_\nu} \partial_{\mu} \partial_{\nu} + K_{\mu} \partial_{\mu} + L$_and_we have the_following expressions for $\Omega_{\mu\nu}$ and $E$: $$\begin{aligned}
E = -\frac{1}{4}_R \otimes 1_{N^2} - \sum_{\mu <_\nu} \gamma^{\mu} \gamma^{\nu}_\otimes_F_{\mu_\nu} \nonumber \\
\Omega_{\mu \nu}_= \frac{1}{4} R^{ab}_{\mu \nu} \gamma_{ab} \otimes_1_{n^2} + id_4 \otimes F_{\mu \nu}_\nonumber,\end{aligned}$$ where $F_{\mu \nu}$ is the curvature_of the connection $\nabla^B + \bA^{pert}$.
This_result allows us to compute_the bosonic_spectral action for the fluctuated_Dirac operator $D_A$,_essentially reducing_the computation in_terms of a local trivialization to_the trivial case_of [@ChamseddineConnes] with the following result.
\[thm:actionnontriv\]_For_the spectral triple_${(\Gamma(M,B),_L^2(M,B\otimes_S),D_B,J,\gamma_B)}$, the_spectral action equals_the |
spins leaves the energy unchanged[@Liu2017]. Therefore, our experiment implies that this single-ion anisotropy is vanishingly small and what lifts the U(1) degeneracy is the in-plane magnetic field. Therefore, instead of having *six* domains (which would have been the case if the degeneracy was lifted by the single-ion anisotropy), we have only *two* domains set by the orientation of the magnetic field. This is the first outcome of our study. A field as small as 0.5 T easily sets the orientation of the spins by coupling to the in-plane magnetization, which is 1.5$\times 10^{-2} \mu_{B}$/f.u. at 0.5 T. This corresponds to an energy as small as 0.5 $\mu$eV/f.u. Such an upper boundary is in agreement with the angle-dependent torque magnetometry data reported by Duan and co-workers who quantified the magnetic crystalline anisotropy of the system[@Duan2015](see section H in Appendix).
Momentum-space Berry curvature
------------------------------
A second consequence is about the in-plane anisotropy of the momentum-space Berry curvature. Previous studies [@Nakatsuji2015; @Nayak2016; @Kiyohara2016; @Li2017] detected a finite AHC for two perpendicular orientations of magnetic field. The magnitude of $\sigma^{A}_{yz}$($B\|x$) and $\sigma^{A}_{xz}$($B\|y$) was found to be close to each other in both Mn$_{3}$Ge [@Kiyohara2016] and in Mn$_{3}$Sn [@Li2017]. Measuring numerous samples (see Appendix), we also found that the anisotropy is small and below 0.15 (see Fig. 3a). An even smaller upper bound ($\sim$0.05) on any in-plane anisotropy is implied by our angle-dependent experiment. Our density functional theory calculations find a band structure (see the Appendix) and a Fermi surface unchanged as the spins rotate (Figs. 3b-e), in agreement with the experimental observation that finds unchanged $\sigma^{A}_{ij}$ for any arbitrary orientation of magnetic field in the xy plane when the current is along the $z$ axis and the electric field is measured perpendicular to the magnetic field.
A finite $\sigma^{A}_{ij}$ arises when the overall integration of Berry curvature, $\Omega^{k}$ in the entire Brillouin | spins leaves the energy unchanged[@Liu2017 ]. Therefore, our experiment entail that this individual - ion anisotropy is vanishingly small and what lifts the U(1) degeneracy is the in - airplane magnetic field. Therefore, alternatively of having * six * domains (which would have been the event if the degeneracy was lifted by the individual - ion anisotropy), we have only * two * world set by the predilection of the magnetic field. This is the inaugural outcome of our study. A field equally small as 0.5 T well sets the orientation course of the spins by copulate to the in - plane magnetization, which is 1.5$\times 10^{-2 } \mu_{B}$/f.u. at 0.5 T. This corresponds to an energy equally small as 0.5 $ \mu$eV / f.u. Such an upper boundary is in agreement with the angle - dependent torque magnetometry data report by Duan and colorado - workers who quantified the magnetic crystalline anisotropy of the system[@Duan2015](see section H in Appendix).
Momentum - space Berry curvature
------------------------------
A second consequence is about the in - plane anisotropy of the momentum - space Berry curvature. Previous study [ @Nakatsuji2015; @Nayak2016; @Kiyohara2016; @Li2017 ] detected a finite AHC for two perpendicular orientation of charismatic field. The magnitude of $ \sigma^{A}_{yz}$($B\|x$) and $ \sigma^{A}_{xz}$($B\|y$) was found to be close to each other in both Mn$_{3}$Ge [ @Kiyohara2016 ] and in Mn$_{3}$Sn [ @Li2017 ]. Measuring numerous samples (see Appendix), we besides found that the anisotropy is small and below 0.15 (see Fig. 3a). An even smaller upper bound ($ \sim$0.05) on any in - plane anisotropy is implied by our angle - dependent experiment. Our density functional theory calculation find a set structure (see the Appendix) and a Fermi surface unchanged as the spins revolve (Figs. 3b - e), in agreement with the experimental observation that finds unaltered $ \sigma^{A}_{ij}$ for any arbitrary orientation of charismatic field in the xy plane when the current is along the $ z$ axis and the electric field is quantify perpendicular to the magnetic battlefield.
A finite $ \sigma^{A}_{ij}$ arises when the overall consolidation of Berry curvature, $ \Omega^{k}$ in the entire Brillouin | splns leaves the energy unghanged[@Liu2017]. Thereyire, ouc experjment imolies that this single-ion anmsoteopy us vanishingly small avd what lpfts the Y(1) dejeneracy is the mh-plane magnetjg fienv. Therefore, insjead of havitg *six* domains (wfieh would have been the case if the dqgeneravy was lifted by the fingmv-iin anisotropy), we have only *twk* domaiis set by the otientation of the magnetic fiepd. This is the firdt outcome if ote study. A fidld as small as 0.5 T easjly sets the orientation of the spinx by couplunt tl the in-planx magnvtization, whign is 1.5$\thmes 10^{-2} \mi_{B}$/f.u. at 0.5 T. This cocresponds to an energy as small as 0.5 $\mu$eV/f.u. Fuch an u[pzr boundary is in agrwenent fith the qngue-dtpeideht toreue magnetomefry data reported by Duan and vo-rirkers who quzntifiqd the magnetic crystalline anisotropy of ths system[@Duan2015](see section H in Appendix).
Momentul-space Bewry curvature
------------------------------
A second consequence is about the in-[lane xniwobroph ov the momentum-space Berry curvature. Previous ffucivs [@Nakatsuji2015; @Nayah2016; @Kiyohara2016; @Ki2017] drjected a finitg AHC for fwo perpendicular lrientajions if magnetyc foeld. The magnitude of $\sigma^{Q}_{yz}$($B\|x$) and $\siyma^{Q}_{xz}$($B\|y$) was found to be close tu eavh otner in both Mn$_{3}$Ge [@Kiyohaxa2016] and in Mn$_{3}$Sn [@Li2017]. Measurinf numerous sampler (sve A[pendix), we also found that the anisitro'y is smxll snd bejow 0.15 (see Flg. 3a). Ak even smaller uppeg bouud ($\sik$0.05) on any ij-plane anisotropy is implied by our angle-depemdant experimznt. Ouv density functyonal theory cclculatijns fknd a band structnre (see the Wppendix) and d Fermi surfare unchanded qs tye spinr rotate (Figs. 3b-e), in agreement with rhe experimental onservxfion that finds unxhanged $\sigma^{A}_{ik}$ fur wnj acbitrwsy orientatimn ow mxbnetiz field in bhe xy llane when the currett ia along the $z$ axis akd the elgctric fiqld is measurrd perpendicular tl the majnetic figld.
A finite $\sigma^{A}_{ij}$ arises wheh the ovegalj integratiog of Berry curvacure, $\Omega^{k}$ in the entire Brillouin | spins leaves the energy unchanged[@Liu2017]. Therefore, our that single-ion anisotropy vanishingly small and is in-plane magnetic field. instead of having domains (which would have been the if the degeneracy was lifted by the single-ion anisotropy), we have only *two* set by the orientation of the magnetic field. This is the first outcome our A as as 0.5 T easily sets the orientation of the spins by coupling to the in-plane magnetization, is 1.5$\times 10^{-2} \mu_{B}$/f.u. at 0.5 T. This to an energy as as 0.5 $\mu$eV/f.u. Such an boundary in agreement the torque data reported by and co-workers who quantified the magnetic crystalline anisotropy of the system[@Duan2015](see section H in Appendix). Momentum-space Berry ------------------------------ A is about in-plane of momentum-space Berry curvature. [@Nakatsuji2015; @Nayak2016; @Kiyohara2016; @Li2017] detected a two perpendicular orientations of magnetic field. The magnitude $\sigma^{A}_{yz}$($B\|x$) and was found to be close to other in both Mn$_{3}$Ge [@Kiyohara2016] and in Mn$_{3}$Sn Measuring numerous samples (see Appendix), we also found that the anisotropy is small and below Fig. 3a). An even upper bound ($\sim$0.05) any anisotropy implied our angle-dependent Our density functional theory calculations find a band structure (see the and a Fermi surface unchanged as the spins rotate (Figs. agreement the experimental observation finds unchanged $\sigma^{A}_{ij}$ for arbitrary of magnetic field in plane the the axis the electric field is perpendicular to the magnetic field. finite $\sigma^{A}_{ij}$ arises when curvature, $\Omega^{k}$ in the entire Brillouin | spins leaves the energy unchaNged[@Liu2017]. TheReforE, ouR exPeRimeNt imPlies that this sINgle-Ion anisotropy is vanishiNgly sMaLL and WHaT liftS the U(1) deGEnERAcy Is ThE in-PlANe MagneTic Field. ThErefore, insTeaD oF having *six* doMAiNs (which wouLd hAve been the caSe iF the deGeNerACy was LifTed by The sinGLe-ion aNisotropy), We HAve onlY *Two* domaINS sEt by The orientation of tHE mAGnetic field. ThiS is the FiRSt OUTcoMe oF our study. A FiEld as SMall as 0.5 T EAsILY SetS The orientatioN of the spins BY coUpling To The IN-plane MagneTiZAtiOn, which is 1.5$\tiMes 10^{-2} \mU_{B}$/f.u. at 0.5 T. ThIs corrESponds tO An energY as smaLl aS 0.5 $\mu$EV/f.u. sUcH aN upPeR BouNDaRy iS In aGreement WiTh The anGle-dEPENDent TorQue mAgnetOmetry data repOrtEd by dUan And co-WorkeRs whO qUantiFied thE magnEtIc crystalline anIsotRopy of the SysTeM[@DuAn2015](See seCTion H iN ApPenDix).
MomeNtum-spaCE BeRrY CURvAture
------------------------------
A second conseqUeNCE iS about thE in-plaNE aNiSOtropy of ThE moMentUM-Space berrY CuRvature. PReviouS StUdIes [@NakaTsUji2015; @NayAk2016; @kiyOhaRa2016; @Li2017] dETectEd a finIte AHC foR two pERpendicular oriENtations of magNEtIC FiELd. ThE maGnitude of $\siGma^{A}_{YZ}$($B\|x$) aNd $\siGMa^{a}_{xz}$($b\|Y$) was fOund tO bE ClOSe to each other in both mn$_{3}$ge [@KiyoHara2016] aNd in Mn$_{3}$Sn [@Li2017]. MeaSuring numeROUS samples (See APPeNDix), we also found That tHe anisotroPY is small And beLow 0.15 (see FiG. 3a). An even sMALler uppeR boUnd ($\Sim$0.05) On aNY In-Plane anisotroPY Is imPlIed by ouR anGle-depeNdeNt eXpeRimEnT. Our densiTy functiOnAl ThEoRy cAlculATions finD a BanD sTruCture (SEe the APpendIx) anD a feRMi sUrface uNChANGed aS tHe SpinS roTaTe (FigS. 3b-e), iN AgrEement wIth the expEriMEntaL oBsErvatioN that finds uncHaNged $\sigma^{A}_{Ij}$ For Any arbITRary orieNtation of magnetic field iN The xy plAne When tHe cuRrent is alOng The $z$ axIs aND the elEctric Field Is MeaSURed peRPEnDicUlAr to the magNETic Field.
a fInitE $\sigma^{A}_{Ij}$ arises when the oveRAll Integration of berRy cuRVAtUre, $\oMeGA^{k}$ iN tHE enTIRe Brillouin | spins leaves the energy u nchanged[@ Liu20 17] . T he refo re,our experiment impl ies that this single-i on an is o trop y i s van ishingl y s m a llan dwha tl if ts th e U (1) deg eneracy is th ein-plane mag n et ic field.The refore, inst ead of ha vi ng* six*dom ains(which wouldhave been t h e case if thed e ge nera cy was lifted byt he single-ion ani sotrop y) , w e hav e o nly *two*do mains set byt he o r ien t ation of themagnetic fi e ld. Thisis th e first outc om e of our study. A f ield as s mall a s 0.5 Te asily s ets th e o rie ntat i on o f t he spi n sbyc oup ling toth ein-pl anem a g n etiz ati on,which is 1.5$\time s 1 0^{- 2 } \ mu_{B }$/f. u. a t0.5 T . This corr es ponds to an ene rgyas smallas0. 5 $ \m u$eV/ f .u. Su chanupper b oundary isin a g re ement with the ang le - d ep endent t orquem ag ne t ometry d at a r epor t e d byDuan an d co-wor kers w h oqu antifie dthe ma gn eti c c rysta l line aniso tropy of thes ystem[@Duan201 5 ](see section Hi n A p pend ix) .
Momentum -spa c e Be rryc ur vat u re
-- ----- -- - -- - -----------------
Asecond cons equence is ab out the in - p l ane anis otro p yo f the momentum -spac e Berry cu r vature.Previ ous stud ies [@Nak a t suji2015 ; @ Nay ak2 016 ; @K iyohara2016;@ L i201 7] detect eda finit e A HCfor tw operpendic ular ori en ta ti on s o f mag n etic fie ld . T he ma gnitu d e of $ \sigm a^{A }_ {y z }$( $B\|x$) an d $\si gm a^ {A}_ {xz }$ ($B\| y$)w asfound t o be clos e t o eac hot her inboth Mn$_{3}$ Ge [@Kiyohar a2 016 ] andi n Mn$_{3} $Sn [@Li2017]. Measurin g numero ussampl es ( see Appen dix ), weals o found thatthe a ni sot r o py is s ma llan d below 0. 1 5 (s ee Fi g. 3a) . An ev en smaller upper b o und ($\sim$0.05) on any i n- pla n ea nis ot r opy i s implied by ou r angle-de pe n de nt experim e nt. O ur dens ity fun ction a l theor y calcula tions fin da ba n d st ructure (s ee the A ppendix)a nd aF er mi su rfa ce unc ha nge d asthe sp i nsrotat e (Fig s. 3b-e) , inag reementwith the experimental o bserva tiontha t finds u nch a nge d $\sigma ^{A} _{ij}$ for an y a rbitr ary orien tati o nofm agnet ic f i eld in th e x y p l a ne when the c u r r ent is a lon g the $ z$ a xis and the elect r ic field is me asur e d pe rpe n dicu la r to the magne tic f i e ld.
A f in ite $\sigma ^{A}_{ij }$ arise s when the o veralli n te g ration ofBer ry curvat ure ,$ \Omega^ {k }$ in the ent ir e Bril louin | spins_leaves the_energy unchanged[@Liu2017]. Therefore, our_experiment implies_that_this single-ion_anisotropy_is vanishingly small_and what lifts_the U(1) degeneracy is_the in-plane magnetic_field._Therefore, instead of having *six* domains (which would have been the case if the_degeneracy_was lifted_by_the_single-ion anisotropy), we have only_*two* domains set by the_orientation of_the magnetic field. This is the first outcome_of_our study. A_field as small as 0.5 T easily sets the_orientation of the spins by coupling_to the in-plane_magnetization,_which_is 1.5$\times 10^{-2} \mu_{B}$/f.u. at_0.5 T. This corresponds to an_energy as small as 0.5 $\mu$eV/f.u._Such an upper boundary is in agreement_with the angle-dependent torque magnetometry data_reported by Duan and co-workers_who quantified_the magnetic crystalline anisotropy of_the system[@Duan2015](see section_H in_Appendix).
Momentum-space Berry curvature
------------------------------
A_second consequence is about the in-plane_anisotropy of the_momentum-space Berry curvature. Previous studies [@Nakatsuji2015;_@Nayak2016;_@Kiyohara2016; @Li2017] detected_a_finite_AHC for_two perpendicular orientations_of_magnetic field._The_magnitude of $\sigma^{A}_{yz}$($B\|x$) and $\sigma^{A}_{xz}$($B\|y$) was_found_to be close to each other in_both Mn$_{3}$Ge [@Kiyohara2016] and_in_Mn$_{3}$Sn [@Li2017]. Measuring numerous_samples (see Appendix), we also_found that the anisotropy is small_and below_0.15 (see_Fig. 3a). An even smaller upper bound ($\sim$0.05) on any in-plane anisotropy_is implied by our angle-dependent experiment._Our density functional theory_calculations find_a_band structure (see_the_Appendix) and_a Fermi surface unchanged as the spins_rotate (Figs. 3b-e),_in agreement with the experimental observation_that finds unchanged $\sigma^{A}_{ij}$_for_any arbitrary orientation of magnetic field_in the xy plane when the_current is along the $z$_axis_and_the electric field is measured_perpendicular to the magnetic field.
A finite_$\sigma^{A}_{ij}$ arises when_the overall integration of Berry curvature, $\Omega^{k}$_in_the entire Brillouin |
.\end{gathered}$$
The $S$ orbit function $S_{\lambda}(x)$, $\lambda\in P^{++}$ is defined as $$\begin{gathered}
\label{def_s-function1}
S_\lambda(x) := \sum_{\mu\in W_\lambda(G)} (-1)^{p(\mu)}e^{2\pi i \l\mu, x\r},\qquad
x\in\R^n,\end{gathered}$$ where $p(\mu)$ is the number of reflections necessary to obtain $\mu$ from $\lambda$. Of course the same $\mu$ can be obtained by different successions of reflections, but all routes from $\lambda$ to $\mu$ will have a length of the same parity, and thus the salient detail given by $p(\mu)$, in the context of an $S$-function, is meaningful and unchanging.
We define $E$ orbit function $E_{\lambda}(x)$, $\lambda\in P^e$ as $$\begin{gathered}
\label{def_e-function1}
E_\lambda(x) := \sum_{\mu\in W^e_{\lambda}(G)} e^{2\pi i \l\mu, x\r},
\qquad
x\in\R^n,\end{gathered}$$ where $P^e:=P^+\cup r_i P^+$ and $r_i$ is a reflection from $W$.
If we always suppose that $\lambda, \mu\in P$ are given in the $\w$-basis, and $x\in\R^n$ is given in the $\alpha$ basis, namely $\lambda=\sum\limits^n_{j=1}\lambda_j\w_j$, $\mu=\sum\limits^n_{j=1}\mu_j\w_j$, $\lambda_j, \mu_j\in\Z$ and $x=\sum\limits^n_{j=1}x_j\alpha_j$, $x_j\in \R$, then the orbit functions of $A_n$ have the following forms $$\begin{gathered}
C_\lambda(x)
= \sum_{\mu\in W_\lambda} e^{2\pi i \sum\limits^n_{j=1}\mu_jx_j}
= \sum_{\mu\ | .\end{gathered}$$
The $ S$ orbit function $ S_{\lambda}(x)$, $ \lambda\in P^{++}$ is defined as $ $ \begin{gathered }
\label{def_s - function1 }
S_\lambda(x): = \sum_{\mu\in W_\lambda(G) } (-1)^{p(\mu)}e^{2\pi i \l\mu, x\r},\qquad
x\in\R^n,\end{gathered}$$ where $ p(\mu)$ is the number of expression necessary to receive $ \mu$ from $ \lambda$. Of course the same $ \mu$ can be obtained by unlike successions of reflections, but all routes from $ \lambda$ to $ \mu$ will hold a length of the same parity bit, and therefore the salient detail given by $ p(\mu)$, in the context of an $ S$-function, is meaningful and static.
We define $ E$ orbit routine $ E_{\lambda}(x)$, $ \lambda\in P^e$ as $ $ \begin{gathered }
\label{def_e - function1 }
E_\lambda(x): = \sum_{\mu\in W^e_{\lambda}(G) } e^{2\pi i \l\mu, x\r },
\qquad
x\in\R^n,\end{gathered}$$ where $ P^e:=P^+\cup r_i P^+$ and $ r_i$ is a mirror image from $ W$.
If we always suppose that $ \lambda, \mu\in P$ are given in the $ \w$-basis, and $ x\in\R^n$ is given in the $ \alpha$ basis, namely $ \lambda=\sum\limits^n_{j=1}\lambda_j\w_j$, $ \mu=\sum\limits^n_{j=1}\mu_j\w_j$, $ \lambda_j, \mu_j\in\Z$ and $ x=\sum\limits^n_{j=1}x_j\alpha_j$, $ x_j\in \R$, then the sphere functions of $ A_n$ have the following shape $ $ \begin{gathered }
C_\lambda(x)
= \sum_{\mu\in W_\lambda } e^{2\pi i \sum\limits^n_{j=1}\mu_jx_j }
= \sum_{\mu\ | .\end{hathered}$$
The $S$ orbit funcuion $S_{\lambda}(x)$, $\lamyea\in P^{++}$ is derined as $$\begin{gathered}
\label{def_s-functmon1}
S_\oambdq(x) := \sum_{\mu\in W_\lambda(G)} (-1)^{p(\ou)}e^{2\pi i \l\lu, x\r},\qquqd
x\ii\R^n,\end{gathered}$$ wisre $p(\mu)$ is ths numyec of reflectionx necessarf to obtain $\mu$ ffol $\lambda$. Of course the same $\mu$ can fe obtaonfd by differenj sucbefsiohs of reflections, but all routes fdom $\lamuda$ to $\mu$ will nave a length of the same oarihy, and thus the sapient detaio gidwn by $p(\mu)$, in the conteqc of an $S$-fuhction, is meaningful and unchaneing.
Wz define $E$ irvit xunction $E_{\lembda}(x)$, $\lambda\in P^e$ as $$\begit{gatherrd}
\label{def_e-fungtion1}
X_\lamvda(x) := \sum_{\mu\in W^e_{\lambde}(G)} e^{2\pi i \l\mu, x\r},
\qquad
v\in\R^n,\end{gdtkered}$$ where $P^e:=P^+\cup r_i P^+$ and $t_i$ is a rddleztikn fdom $W$.
Iv wx always sulpose that $\oambda, \mu\in P$ are govqb in the $\w$-basjs, and $x\yn\R^n$ is given in the $\alpha$ basis, namely $\lajbda=\sum\limits^n_{j=1}\lambda_j\w_h$, $\mu=\sum\limits^n_{j=1}\mu_j\w_j$, $\pambda_j, \mt_j\in\Z$ and $x=\sum\limits^n_{j=1}x_j\alpha_j$, $x_j\in \R$, then the ortit fnnztiinf if $W_n$ have the following forms $$\begin{gathered}
C_\lamfsa(c)
= \xum_{\mu\in W_\lambdc} e^{2\pi i \sum\limitx^n_{u=1}\mi_tx_j}
= \sum_{\mu\ | .\end{gathered}$$ The $S$ orbit function $S_{\lambda}(x)$, $\lambda\in defined $$\begin{gathered} \label{def_s-function1} := \sum_{\mu\in W_\lambda(G)} where is the number reflections necessary to $\mu$ from $\lambda$. Of course the $\mu$ can be obtained by different successions of reflections, but all routes from to $\mu$ will have a length of the same parity, and thus the detail by in context of an $S$-function, is meaningful and unchanging. We define $E$ orbit function $E_{\lambda}(x)$, $\lambda\in P^e$ $$\begin{gathered} \label{def_e-function1} E_\lambda(x) := \sum_{\mu\in W^e_{\lambda}(G)} e^{2\pi i x\r}, \qquad x\in\R^n,\end{gathered}$$ where r_i P^+$ and $r_i$ is reflection $W$. If always that \mu\in P$ are in the $\w$-basis, and $x\in\R^n$ is given in the $\alpha$ basis, namely $\lambda=\sum\limits^n_{j=1}\lambda_j\w_j$, $\mu=\sum\limits^n_{j=1}\mu_j\w_j$, $\lambda_j, \mu_j\in\Z$ and $x_j\in \R$, orbit functions $A_n$ the forms $$\begin{gathered} C_\lambda(x) W_\lambda} e^{2\pi i \sum\limits^n_{j=1}\mu_jx_j} = \sum_{\mu\ | .\end{gathered}$$
The $S$ orbit functIon $S_{\lambda}(X)$, $\lambDa\iN P^{++}$ iS dEfinEd as $$\Begin{gathered}
\lABel{dEf_s-function1}
S_\lambda(x) := \sum_{\Mu\in W_\LaMBda(G)} (-1)^{P(\Mu)}E^{2\pi i \l\Mu, x\r},\qquAD
x\IN\r^n,\eNd{GaTheReD}$$ WhEre $p(\mU)$ is The numbEr of reflecTioNs Necessary to oBTaIn $\mu$ from $\laMbdA$. Of course the SamE $\mu$ can Be ObtAIned bY diFfereNt succESsions Of reflectIoNS, but alL Routes fROM $\lAmbdA$ to $\mu$ will have a lenGTh OF the same parity, And thuS tHE sALIenT deTail given bY $p(\Mu)$, in tHE contexT Of AN $s$-FunCTion, is meaningFul and unchaNGinG.
We defInE $E$ oRBit funCtion $e_{\lAMbdA}(x)$, $\lambda\in P^E$ as $$\bEgin{gatheRed}
\labEL{def_e-fuNCtion1}
E_\lAmbda(x) := \Sum_{\Mu\iN W^e_{\lAMbDa}(g)} e^{2\pI i \L\Mu, x\R},
\QqUad
X\In\R^N,\end{gathErEd}$$ Where $p^e:=P^+\cUP R_I p^+$ and $R_i$ iS a reFlectIon from $W$.
If we aLwaYs suPPosE that $\LambdA, \mu\iN P$ Are giVen in tHe $\w$-baSiS, and $x\in\R^n$ is giveN in tHe $\alpha$ baSis, NaMelY $\lAmbda=\SUm\limiTs^n_{J=1}\laMbda_j\w_j$, $\Mu=\sum\liMIts^N_{j=1}\MU_J\W_j$, $\Lambda_j, \mu_j\in\Z$ and $x=\sUm\LIMiTs^n_{j=1}x_j\alPha_j$, $x_j\IN \R$, ThEN the orbiT fUncTionS OF $A_n$ haVe thE FoLlowing fOrms $$\beGIn{GaThered}
C_\LaMbda(x)
= \sUm_{\Mu\iN W_\lAmbda} E^{2\Pi i \sUm\limiTs^n_{j=1}\mu_jx_J}
= \sum_{\mU\ | .\end{gathered}$$
The $S$ orbit fun ction $S _{\ la mbda }(x) $, $\lambda\in P^{+ +}$ is defined as $$\b egin{ ga t here d }\labe l{def_s - fu n c tio n1 }S_\ la m bd a(x):=\sum_{\ mu\in W_\l amb da (G)} (-1)^{p ( \m u)}e^{2\pi i\l\mu, x\r}, \qq uad
x\ in \R^ n ,\end {ga there d}$$ w h ere $p (\mu)$ is t h e numb e r of re f l ec tion s necessary to ob t ai n $\mu$ from $\ lambda $. Of c our sethe same $ \m u$ ca n be obt a in e d byd ifferent succ essions ofr efl ection s, bu t all r outes f r om$\lambda$ t o $\ mu$ willhave a lengtho f the s ame pa rit y,andt hu sthe s a lie n tdet a ilgiven by $ p( \mu)$ , in t h e con tex t of an $ S$-function,ismean i ngf ul an d unc hang in g.
W e defi ne $E $orbit function$E_{ \lambda}( x)$ ,$\l am bda\i n P^e$as$$\ begin{g athered }
\l ab e l { de f_e-function1}
E_\ la m b da (x) := \ sum_{\ m u\ in W^e_{\la mb da} (G)} e ^{2\p i i\ l\ mu, x\r} ,
\qqu a dx\ in\R^n, \e nd{gat he red }$$ wher e $P^ e:=P^+ \cup r_i P^+$ and $r_i$ is a reflection fr o m$ W $.
Ifwealways supp oset hat$\la m bd a,\ mu\in P$ a re gi v en in the $\w$-basi s, and $ x\in\ R^n$ is given in the $\ a l p ha$ basi s, n a me l y $\lambda=\su m\lim its^n_{j=1 } \lambda_ j\w_j $, $\mu= \sum\limi t s ^n_{j=1} \mu _j\ w_j $,$ \ la mbda_j, \mu_j \ i n\Z$ a nd $x=\ sum \limits ^n_ {j= 1}x _j\ al pha_j$, $ x_j\in \ R$ ,th en th e orb i t functi on s o f$A_ n$ ha v e thefollo wing f or m s $ $\begin { ga t h ered }C_ \lam bda (x )
= \ sum_ { \mu \in W_\ lambda} e ^{2 \ pi i \ su m\limit s^n_{j=1}\mu_ jx _j}
= \sum _{ \mu \ | .\end{gathered}$$
The $S$_orbit function_$S_{\lambda}(x)$, $\lambda\in P^{++}$ is_defined as_$$\begin{gathered}
\label{def_s-function1}
S_\lambda(x)_:= \sum_{\mu\in_W_\lambda(G)}_(-1)^{p(\mu)}e^{2\pi i \l\mu,_x\r},\qquad
x\in\R^n,\end{gathered}$$ where $p(\mu)$_is the number of_reflections necessary to_obtain_$\mu$ from $\lambda$. Of course the same $\mu$ can be obtained by different successions_of_reflections, but_all_routes_from $\lambda$ to $\mu$ will_have a length of the_same parity,_and thus the salient detail given by $p(\mu)$,_in_the context of_an $S$-function, is meaningful and unchanging.
We define $E$ orbit_function $E_{\lambda}(x)$, $\lambda\in P^e$ as $$\begin{gathered}
\label{def_e-function1}
E_\lambda(x)_:= \sum_{\mu\in W^e_{\lambda}(G)}_e^{2\pi_i_\l\mu, x\r},
\qquad
x\in\R^n,\end{gathered}$$ where $P^e:=P^+\cup_r_i P^+$ and $r_i$ is a_reflection from $W$.
If we always suppose_that $\lambda, \mu\in P$ are given in_the $\w$-basis, and $x\in\R^n$ is given_in the $\alpha$ basis, namely_$\lambda=\sum\limits^n_{j=1}\lambda_j\w_j$, $\mu=\sum\limits^n_{j=1}\mu_j\w_j$,_$\lambda_j, \mu_j\in\Z$ and $x=\sum\limits^n_{j=1}x_j\alpha_j$, $x_j\in_\R$, then the_orbit functions_of $A_n$ have the_following forms $$\begin{gathered}
C_\lambda(x)
= \sum_{\mu\in W_\lambda} e^{2\pi_i \sum\limits^n_{j=1}\mu_jx_j}
= \sum_{\mu\ |
part of the proof of Theorem \[NonObtuse\] is dealing with the $\sqrt{k}$ internal corners that propagate through the spiral; since we have no bound for the number $N$ of windings of the spiral in terms of $n$, this could produce arbitrarily many new vertices. Thus propagation paths of the internal corners must be bent to terminate earlier. Consider the case of paths that start near the inner end of the spiral (the outer part is handled in the same way, but is easier, since the windings of the spiral are longer). We consider what happens for very large spirals (where the number of windings $N$ is bigger than the number $k$ of isosceles pieces in the spiral; for smaller values of $N$ we truncate the construction at the appropriate stage.)
We first bend the propagation paths so that adjacent paths merge, and then merge adjacent merged paths, and continue until all the propagation paths generated by the $O(\sqrt{k})$ internal corners have merged into a single path. This occurs around winding $k^{1/3}$. This path is then propagated as a $P$-path out to winding $k^{1/2}$. See Figure \[BigPicture\].
![\[BigPicture\] This illustrates the stages in the spiral construction. First (light gray), we divide the tube into thinner tubes and entering paths are bent to hit the sides of these. Next (white), the thin tubes are bent and collapsed in pairs; in this figure four tubes are merged into one after two windings. In the third stage (gray), the single tube is propagated until we can bend it to intersect itself. Next (white) comes a sequence of closed loops that gradually grow further apart. Finally we reach the empty region (gray), where no paths propagate. This figure gives a rough idea of the construction, but scales have been drastically compressed to make all the stages visible in the same picture. ](BigPicture.eps){height="4in"}
At this stage we have enough freedom to bend the curve to hit itself, forming a closed loop that wraps once around the spiral. This is similar to what we did in the proof of Theorem \[Triangles\], but in this case, in order for this closed loop to be Gabriel, there must be another (larger) closed loop parallel to it. This did not occur in Theorem \[Triangles\]. This constraint requires us to construct a sequence of parallel closed loops in the spiral between windings $k^{1 | part of the proof of Theorem \[NonObtuse\ ] is dealing with the $ \sqrt{k}$ internal corner that spread through the spiral; since we have no bound for the phone number $ N$ of windings of the spiral in condition of $ n$, this could produce arbitrarily many fresh vertices. Thus generation paths of the internal corners must be bent to end earlier. Consider the case of paths that start near the inside end of the spiral (the outer function is handled in the same manner, but is easier, since the wind of the spiral are retentive). We consider what happens for very large helix (where the number of windings $ N$ is bigger than the number $ k$ of isosceles pieces in the coiling; for modest values of $ N$ we truncate the construction at the appropriate stage .)
We first bend the propagation paths so that adjacent paths merge, and then unify adjacent merged paths, and cover until all the generation path generated by the $ O(\sqrt{k})$ internal corner have merged into a single path. This occurs round winding $ k^{1/3}$. This path is then propagated as a $ P$-path out to winding $ k^{1/2}$. See Figure \[BigPicture\ ].
! [ \[BigPicture\ ] This illustrates the stage in the spiral construction. First (light grey), we divide the pipe into thinner tubes and entering paths are bent to reach the sides of these. Next (white), the flimsy tubes are bent and collapsed in pairs; in this number four tubes are blend into one after two windings. In the third stage (grey), the single pipe is propagated until we can flex it to intersect itself. Next (ashen) comes a sequence of closed loops that gradually grow farther apart. last we reach the empty region (gray), where no paths propagate. This human body gives a rough mind of the construction, but scales have been drastically compressed to make all the stages visible in the same movie. ] (BigPicture.eps){height="4 in " }
At this stage we have enough freedom to bend the curve to hit itself, form a shut closed circuit that wraps once around the spiral. This is similar to what we did in the proof of Theorem \[Triangles\ ], but in this case, in ordering for this closed loop to be Gabriel, there must be another (larger) close loop parallel to it. This did not occur in Theorem \[Triangles\ ]. This constraint requires us to construct a sequence of parallel shut loops in the spiral between windings $ k^{1 | pagt of the proof of Theortm \[NonObtuse\] is dgaoing wmth the $\sqrt{k}$ ivternal corners that propagave tyrougy the spiral; since we fave no blund for the bumber $N$ oh windinnf of bhe s'ical in terms of $n$, this cogld produce artigrcrily many new vertices. Thus propagaeion payhd of the intergal bownera must be bent to terminate earlied. Consiver the case of paths that start near the innfr end of the spirwl (the outet pawr is handled in the same way, but ia easier, since the windings of ghe s'iral are linter). Fe consider what happens for very lasge spitals (where the nukbee of windings $N$ is bijger than the number $k$ of isovczles pieces in the spurql; fot smanler caljes oh $N$ we trkncete the conatruction ar the appropriate suagq.)
Qe first bend the pwo[agation paths so that adjacent paths mtrge, znd then merge adjacent merged paths, and conjinue untij all the propagation paths generated by the $O(\sqrd{k})$ invefnao gorndes have merged into a single path. This occurs awkumd winding $k^{1/3}$. Thif path is tnej ltopagated as a $P$-path ouf to winding $k^{1/2}$. See Figure \[BigPucture\].
![\[BigKictute\] This illustrates the states in the siirao construction. Firdt (light grcy), we cividr the tube into thinner tubss and enteging patha are bent to hit thv siges of these. Next (white), thq thin tuues axe bent xnd vollapfed in paigs; in this figure four tkbes cre marged into one after two windings. In the viird stage (grsy), thv single cube ix propagated tntil we can bgnd it to intefsect itsemf. Next (white) comef a sequence mv closed loo's that gwaduqlly grow fjfther apart. Fimally we gecch the enpty region (gray), wmere vk paths propagact. Tyis figure givex a rotgj mdea jx the constrgctiun, cit scxles have bcen draxtically compressed do mzke all the stages vlsible in the samq picture. ](BigLicture.eps){height="4in"}
Wt thms staje we navg enough freedom to bend the cudve to hih ibself, forming a coosed loop tkat wraps once around the spiral. This is similar to what ww did in the proof pn Theorem \[Trmanglef\], but in dhis case, in order fir this closed lopp to be Gabriel, thers must be ajother (larger) closed loop parallel to it. This did not occur in Theorem \[Truangles\]. This constdainy reqlirzs bs to cjnstcurt a sequence of iarallel closed loops in the spical betweet cindings $k^{1 | part of the proof of Theorem \[NonObtuse\] with $\sqrt{k}$ internal that propagate through no for the number of windings of spiral in terms of $n$, this produce arbitrarily many new vertices. Thus propagation paths of the internal corners must bent to terminate earlier. Consider the case of paths that start near the end the (the part is handled in the same way, but is easier, since the windings of the spiral longer). We consider what happens for very large (where the number of $N$ is bigger than the $k$ isosceles pieces the for values of $N$ truncate the construction at the appropriate stage.) We first bend the propagation paths so that adjacent paths and then merged paths, continue all propagation paths generated $O(\sqrt{k})$ internal corners have merged into This occurs around winding $k^{1/3}$. This path is propagated as $P$-path out to winding $k^{1/2}$. See \[BigPicture\]. ![\[BigPicture\] This illustrates the stages in the construction. First (light gray), we divide the tube into thinner tubes and entering paths are hit the sides of Next (white), the tubes bent collapsed pairs; in figure four tubes are merged into one after two windings. In third stage (gray), the single tube is propagated until we it intersect itself. Next comes a sequence of loops gradually grow further apart. reach empty no propagate. figure gives a rough of the construction, but scales been drastically compressed to in the same picture. ](BigPicture.eps){height="4in"} At this stage have enough freedom to bend the curve hit itself, forming a closed loop that wraps once around the spiral. is similar we did in the proof of Theorem \[Triangles\], in this case, in for this closed loop to be Gabriel, there must another closed loop to it. This not occur in \[Triangles\]. This constraint to construct sequence parallel the spiral between windings $k^{1 | part of the proof of Theorem \[NoNObtuse\] is dEalinG wiTh tHe $\Sqrt{K}$ intErnal corners thAT proPagate through the spiral; Since We HAve nO BoUnd foR the numBEr $n$ OF wiNdInGs oF tHE sPiral In tErms of $n$, This could pRodUcE arbitrarily MAnY new verticEs. THus propagatiOn pAths of ThE inTErnal CorNers mUst be bENt to teRminate eaRlIEr. ConsIDer the cASE oF patHs that start near thE InNEr end of the spirAl (the oUtER pART is HanDled in the sAmE way, bUT is easiER, sINCE thE Windings of the Spiral are loNGer). we consIdEr wHAt happEns foR vERy lArge spirals (WherE the numbeR of winDIngs $N$ is BIgger thAn the nUmbEr $k$ Of isOScElEs pIeCEs iN ThE spIRal; For smallEr VaLues oF $N$ we TRUNCate The ConsTructIon at the approPriAte sTAge.)
we firSt benD the PrOpagaTion paThs so ThAt adjacent paths MergE, and then mErgE aDjaCeNt merGEd pathS, anD coNtinue uNtil all THe pRoPAGAtIon paths generated bY tHE $o(\sQrt{k})$ inteRnal coRNeRs HAve mergeD iNto A sinGLE path. this OCcUrs arounD windiNG $k^{1/3}$. thIs path iS tHen proPaGatEd aS a $P$-paTH out To windIng $k^{1/2}$. See FIgure \[bIgPicture\].
![\[BigPiCTure\] This illusTRaTES tHE staGes In the spiral ConsTRuctIon. FIRsT (liGHt graY), we diViDE tHE tube into thinner tubEs And entEring Paths are bent tO hit the sidES OF these. NeXt (whITe), THe thin tubes are Bent aNd collapseD In pairs; iN this Figure foUr tubes arE MErged intO onE afTer Two WINdIngs. In the thirD STage (GrAy), the siNglE tube is ProPagAteD unTiL we can benD it to intErSeCt ItSelF. Next (WHite) comeS a SeqUeNce Of cloSEd loopS that GradUaLlY GroW furtheR ApART. FinAlLy We reAch ThE emptY regIOn (gRay), wherE no paths pRopAGate. thIs Figure gIves a rough ideA oF the constrUcTioN, but scALEs have beEn drastically compressed TO make alL thE stagEs viSible in thE saMe pictUre. ](bIgPictUre.eps){HeighT="4iN"}
At THIs staGE We HavE eNough freedOM To bEnd thE cUrve To hit itSelf, forming a closed LOop That wraps once AroUnd tHE SpIraL. thIS is SiMIlaR TO what we did in the Proof of TheOrEM \[TRiangles\], buT In tHiS case, in Order foR this CLosed loOp to be GabRiel, there MuSt be ANOthEr (larger) clOsed loop Parallel tO It. ThiS DiD not oCcuR in TheOrEm \[TRiangLes\]. ThiS ConStraiNt requIrEs us to ConstRuCt a sequeNce of parallel closed loopS in the SpiraL beTween windIngS $K^{1 | part of the proof of Theo rem \[NonO btuse \]isde alin g wi th the $\sqrt{ k }$ i nternal corners that p ropag at e thr o ug h the spiral ; s i n cewe h ave n o b oundfor the nu mber $N$ o f w in dings of the sp iral in te rms of $n$, thi s c ould p ro duc e arbi tra rilymany n e w vert ices. Thu sp ropaga t ion pat h s o f th e internal corner s m u st be bent totermin at e e a r lie r.Consider t he case of path s t h a t st a rt near the i nner end of the spira l(th e outer part i s ha ndled in th e sa me way, b ut ise asier,s ince th e wind ing s o f th e s pi ral a r e l o ng er) . We conside rwh at ha ppen s f o r ve rylarg e spi rals (where t henumb e r o f win dings $N$ i s big ger th an th enumber $k$ of i sosc eles piec esin th espira l ; forsma lle r value s of $N $ we t r u n ca te the constructio na t t he appro priate st ag e .)
We f ir stbend t he pr opag a ti on paths so th a tad jacentpa ths me rg e,and then merg e adja cent mer ged p a ths, and conti n ue until allt he p ro p agat ion paths gene rate d bythe$ O( \sq r t{k}) $ int er n al corners have merged i nto asingl e path. Thisoccurs aro u n d winding $k^ { 1/ 3 }$. This pathis th en propaga t ed as a$P$-p ath outto windin g $k^{1/2} $.See Fi gur e \[ BigPicture\]. ![\[ Bi gPictur e\] This i llu str ate s t he stages i n the sp ir al c on str uctio n . First(l igh tgra y), w e divid e the tub ein t o t hinnert ub e s and e nt erin g p at hs ar e be n t t o hit t he sidesoft hese .Ne xt (whi te), the thin t ubes are b en t a nd col l a psed inpairs; in this figure f o ur tube s a re me rged into one af ter tw o w i ndings . In t he th ir d s t a ge (g r a y) , t he single tu b e is prop ag ated untilwe can bend it toi nte rsect itself. Ne xt ( w h it e)c om e s a s e que n c e of closed loo ps that gr ad u al ly grow fu r the rapart.Finally we r e ach the empty re gion (gra y) , wh e r e n o paths pr opagate. This fig u re gi v es a ro ugh ideaof th e con struct i on, butscales h ave be en dr as ticallycompressed to make allthe st agesvis ible in t hes ame picture. ](B igPicture. eps ){h eight ="4 i n"}
At t h is st a ge we hav e enough f r ee dom t obend the cu r v e to hitits e lf, fo rmin g a closed loop t h at wraps oncearou n d th e s p iral .This is simila r t ow h at we di din the proo f of The or e m \[T riangl es\],but int h is case,in o rde r for thi s c lo s ed loop t ob e Gabr iel, t here m ust be anot h e r (larger) close d loo p paral l elto it .This di d not occur inTheorem \[T riangl es\] . Thi s const ra int re qui re s us to co n struct aseque nce ofpa rall elclosed loo p s in t he s pi ral betweenw i nd i ng s$ k^{ 1 | part_of the_proof of Theorem \[NonObtuse\]_is dealing_with_the $\sqrt{k}$_internal_corners that propagate_through the spiral;_since we have no_bound for the_number_$N$ of windings of the spiral in terms of $n$, this could produce arbitrarily_many_new vertices._Thus_propagation_paths of the internal corners_must be bent to terminate_earlier. Consider_the case of paths that start near the_inner_end of the_spiral (the outer part is handled in the same_way, but is easier, since the_windings of the_spiral_are_longer). We consider what_happens for very large spirals (where_the number of windings $N$ is_bigger than the number $k$ of isosceles_pieces in the spiral; for smaller_values of $N$ we truncate_the construction_at the appropriate stage.)
We first_bend the propagation_paths so_that adjacent paths_merge, and then merge adjacent merged_paths, and continue_until all the propagation paths generated_by_the $O(\sqrt{k})$ internal_corners_have_merged into_a single path._This_occurs around_winding_$k^{1/3}$. This path is then propagated_as_a $P$-path out to winding $k^{1/2}$. See_Figure \[BigPicture\].
![\[BigPicture\] This illustrates_the_stages in the spiral_construction. First (light gray), we_divide the tube into thinner tubes_and entering_paths are_bent to hit the sides of these. Next (white), the thin_tubes are bent and collapsed in_pairs; in this figure_four tubes_are_merged into one_after_two windings._In the third stage (gray), the single_tube is_propagated until we can bend it_to intersect itself. Next_(white)_comes a sequence of closed loops_that gradually grow further apart. Finally_we reach the empty region_(gray),_where_no paths propagate. This figure_gives a rough idea of the_construction, but scales_have been drastically compressed to make all_the_stages visible in the same picture._](BigPicture.eps){height="4in"}
At_this stage we have enough freedom_to_bend_the curve to hit itself,_forming a closed loop that wraps_once around the spiral. This is similar to what_we did in_the proof of Theorem \[Triangles\],_but_in_this case, in order for this closed loop to be_Gabriel, there_must be another_(larger) closed loop parallel to it. This did not occur_in Theorem \[Triangles\]. This constraint requires us_to construct a sequence of parallel closed loops in the spiral_between windings $k^{1 |
described below. The transmission not only depends on the phase, frequency, and amplitude of the oscillation, but also on the coefficients $a_n$ and $b_n$ determined by the state selectors, and the horizontal velocity distribution $\eta_v$. Here, we assume $a_n\approx b_n$, which is valid in our experiment within $\sim 10\%$. Then, from $\sum_n a_n=1$ and $\sum_n |C_{n}(\tau_1)|^2=|C_{2}(\tau_1)|^2=1$ after the first state selector, it follows that the relative amplitude $\gamma$ of the $|2\rangle\leftrightarrow|4\rangle$ Ramsey fringes is mainly determined by the fraction $(a_2+a_4)/\sum_j a_j$ of states at $t=\tau_1$ that participate in the excitation process. As noted in [Section \[sec:theory\]]{} already, we know that $a_4\approx0$, leaving $a_2$ as single parameter determining the amplitude of the Ramsey fringe. As is illustrated in [Fig. \[fig:setup\_schematic\]]{}, other states (mainly $|1\rangle$ and $|3\rangle$) will pass through the setup unaffected by the oscillation. These states create a background (offset) in the relative transmission. As argued in [Section \[sec:theory\]]{}, we can restrict ourselves to a two-state approximation [@Abele:2012]. Hence, only $C_{2}({\tau}_4)$ and $C_{4}({\tau}_4)$ are determined by [Eqn. (\[eq:transmission\_coefficients\])]{}, describing the Ramsey fringe dependence on ${\alpha}$, while all other states are ignored. Under these assumptions, $P$ would result in a Ramsey fringe of amplitude $\sim 1$. In order to account for the contributions of other states (again, mainly $|1\rangle$ and $|3\rangle$) as well as losses, we introduce an amplitude parameter $\gamma$, and an offset $r_{off}$ to scale and shift the theoretical curve [^7]. Eventually, this results in the fit function $$\begin{aligned}
&P_{\hspace{-0.5pt}fit}({\alpha},\hspace{-0.5pt}\gamma,\hspace{-0.5pt}r_{off})\hspace{-1pt}=\hspace{-1pt}r_{off}\ | described below. The transmission not only depend on the phase, frequency, and amplitude of the cycle, but also on the coefficients $ a_n$ and $ b_n$ determine by the country selectors, and the horizontal velocity distribution $ \eta_v$. Here, we bear $ a_n\approx b_n$, which is valid in our experiment within $ \sim 10\%$. Then, from $ \sum_n a_n=1 $ and $ \sum_n |C_{n}(\tau_1)|^2=|C_{2}(\tau_1)|^2=1 $ after the first state picker, it follow that the proportional amplitude $ \gamma$ of the $ |2\rangle\leftrightarrow|4\rangle$ Ramsey fringes is mainly determine by the fraction $ (a_2+a_4)/\sum_j a_j$ of country at $ t=\tau_1 $ that participate in the excitation process. As noted in [ Section \[sec: theory\ ] ] { } already, we acknowledge that $ a_4\approx0 $, leaving $ a_2 $ as single parameter specify the amplitude of the Ramsey fringe. As is illustrated in [ Fig. \[fig: setup\_schematic\ ] ] { }, other states (chiefly $ |1\rangle$ and $ |3\rangle$) will pass through the frame-up unaffected by the oscillation. These states make a background (offset) in the relative transmission. As argued in [ Section \[sec: theory\ ] ] { }, we can restrict ourselves to a two - state approximation [ @Abele:2012 ]. Hence, only $ C_{2}({\tau}_4)$ and $ C_{4}({\tau}_4)$ are determined by [ Eqn. (\[eq: transmission\_coefficients\ ]) ] { }, describing the Ramsey fringe dependence on $ { \alpha}$, while all other states are ignore. Under these assumptions, $ P$ would result in a Ramsey bang of amplitude $ \sim 1$. In order to account for the contributions of other country (again, mainly $ |1\rangle$ and $ |3\rangle$) as well as losses, we introduce an amplitude argument $ \gamma$, and an offset $ r_{off}$ to scale and shift the theoretical curve [ ^7 ]. Eventually, this results in the fit function $ $ \begin{aligned }
& P_{\hspace{-0.5pt}fit}({\alpha},\hspace{-0.5pt}\gamma,\hspace{-0.5pt}r_{off})\hspace{-1pt}=\hspace{-1pt}r_{off}\ | dedcribed below. The transmlssion not only bwpends on ths phase, wrequency, and amplitude of tie owcillqtion, but also on the zoefficiejts $a_n$ abd $b_i$ determined by vge statc selsgtors, end the horizonjal velocity distribution $\atx_v$. Here, we assume $a_n\approx b_n$, which if valid ij our experimegt wptrin $\apm 10\%$. Then, from $\sum_n a_n=1$ and $\sum_n |D_{n}(\tau_1)|^2=|C_{2}(\teu_1)|^2=1$ after the fitst state selector, it folllws hhat the relative wmplitude $\gqmma$ if the $|2\rangld\leftrightarrow|4\rangle$ Tamsey fringes is mainly determived bv the fractuob $(a_2+w_4)/\vum_j a_j$ of wtatef at $t=\tau_1$ that partiwipate on the excitatlon pcocews. As noted in [Sectioi \[sec:theory\]]{} already, wg know thad $c_4\approx0$, leaving $a_2$ as wibgle karamater eetdrmjnmng the alplmtude of ths Ramsey frunge. As is illustraued pm [Fig. \[fig:setul\_schemwtyc\]]{}, other states (mainly $|1\rangle$ and $|3\ranglt$) wilm pass through the setup unaffected by the odcillatiog. These states create a background (offset) in the selatmvd txqnsmirwiln. As argued in [Section \[sec:theory\]]{}, we can restrydt olrselves to a two-ftate approcilayyon [@Abele:2012]. Hencg, only $E_{2}({\fah}_4)$ and $C_{4}({\tau}_4)$ are detfrmined by [Ewn. (\[eq:transiissoon\_coefficients\])]{}, describing rhe Ramsey fgingw dependence on ${\al'ha}$, while alu otner syates are ignored. Under theae assumptilns, $P$ wouma result in a Raosej frhnge of amplitude $\sim 1$. In jrder to eccouut for tfe cpntribttions of lther states (again, mainlj $|1\ranyle$ atd $|3\rangle$) ws well as losses, we introduce ei amplitude psrdmener $\gamma$, and sn offset $r_{ofs}$ to scale and shift che thdoretical burve [^7]. Eveitually, this results in tvg fit functioi $$\begin{alygnee}
&P_{\hwpace{-0.5pt}wkt}({\alpha},\hspace{-0.5py}\gamma,\hspcee{-0.5pt}r_{off})\hwpace{-1pt}=\hspace{-1pt}r_{ofn}\ | described below. The transmission not only depends phase, and amplitude the oscillation, but and determined by the selectors, and the velocity distribution $\eta_v$. Here, we assume b_n$, which is valid in our experiment within $\sim 10\%$. Then, from $\sum_n and $\sum_n |C_{n}(\tau_1)|^2=|C_{2}(\tau_1)|^2=1$ after the first state selector, it follows that the relative $\gamma$ the Ramsey is mainly determined by the fraction $(a_2+a_4)/\sum_j a_j$ of states at $t=\tau_1$ that participate in the process. As noted in [Section \[sec:theory\]]{} already, we that $a_4\approx0$, leaving $a_2$ single parameter determining the amplitude the fringe. As illustrated [Fig. other states (mainly and $|3\rangle$) will pass through the setup unaffected by the oscillation. These states create a background (offset) the relative argued in \[sec:theory\]]{}, can ourselves to a [@Abele:2012]. Hence, only $C_{2}({\tau}_4)$ and $C_{4}({\tau}_4)$ [Eqn. (\[eq:transmission\_coefficients\])]{}, describing the Ramsey fringe dependence on while all states are ignored. Under these assumptions, would result in a Ramsey fringe of amplitude 1$. In order to account for the contributions of other states (again, mainly $|1\rangle$ and well as losses, we an amplitude parameter and offset to and shift theoretical curve [^7]. Eventually, this results in the fit function $$\begin{aligned} | described below. The transmisSion not onlY depeNds On tHe PhasE, freQuency, and ampliTUde oF the oscillation, but also On the CoEFficIEnTs $a_n$ aNd $b_n$ detERmINEd bY tHe StaTe SElEctorS, anD the horIzontal velOciTy Distribution $\ETa_V$. Here, we assUme $A_n\approx b_n$, whIch Is valiD iN ouR ExperImeNt witHin $\sim 10\%$. tHen, froM $\sum_n a_n=1$ anD $\sUM_n |C_{n}(\taU_1)|^2=|c_{2}(\tau_1)|^2=1$ aftER ThE firSt state selector, it FOlLOws that the relaTive amPlITuDE $\GamMa$ oF the $|2\rangle\LeFtrigHTarrow|4\rANgLE$ rAmsEY fringes is maiNly determinED by The fraCtIon $(A_2+A_4)/\sum_j a_J$ of stAtES at $T=\tau_1$ that parTiciPate in the ExcitaTIon procESs. As notEd in [SeCtiOn \[sEc:thEOrY\]]{} aLreAdY, We kNOw ThaT $A_4\apProx0$, leavInG $a_2$ As sinGle pARAMEter DetErmiNing tHe amplitude of The ramsEY frInge. AS is ilLustRaTed in [fig. \[fig:Setup\_ScHematic\]]{}, other staTes (mAinly $|1\rangLe$ aNd $|3\RanGlE$) will PAss thrOugH thE setup uNaffectED by ThE OSCiLlation. These states CrEATe A backgroUnd (offSEt) In THe relatiVe TraNsmiSSIon. As ArguED iN [Section \[Sec:theORy\]]{}, We Can restRiCt oursElVes To a Two-stATe apProximAtion [@AbeLe:2012]. HenCE, only $C_{2}({\tau}_4)$ and $C_{4}({\tAU}_4)$ are determineD By [eQN. (\[eQ:TranSmiSsion\_coeffiCienTS\])]{}, desCribINg The rAmsey FringE dEPeNDence on ${\alpha}$, while alL oTher stAtes aRe ignored. UndeR these assuMPTIons, $P$ wouLd reSUlT In a Ramsey fringE of amPlitude $\sim 1$. iN order to AccouNt for the ContributIONs of otheR stAteS (agAin, MAInLy $|1\rangle$ and $|3\raNGLe$) as WeLl as losSes, We introDucE an AmpLitUdE parameteR $\gamma$, anD aN oFfSeT $r_{oFf}$ to sCAle and shIfT thE tHeoReticAL curve [^7]. eventUallY, tHiS ResUlts in tHE fIT FuncTiOn $$\BegiN{alIgNed}
&P_{\hSpacE{-0.5Pt}fIt}({\alpha},\Hspace{-0.5pt}\gAmmA,\HspaCe{-0.5Pt}R_{off})\hspAce{-1pt}=\hspace{-1pt}R_{oFf}\ | described below. The tran smission n ot on lydep en ds o n th e phase, frequ e ncy, and amplitude of theoscil la t ion, bu t als o on th e c o e ffi ci en ts$a _ n$ and$b_ n$ dete rmined bythe s tate selecto r s, and the h ori zontal veloc ity distr ib uti o n $\e ta_ v$. H ere, w e assum e $a_n\ap pr o x b_n$ , whichi s v alid in our experimen t w i thin $\sim 10\ %$. Th en , f r o m $ \su m_n a_n=1$ a nd $\ s um_n |C _ {n } ( \ tau _ 1)|^2=|C_{2}( \tau_1)|^2= 1 $ a fter t he fi r st sta te se le c tor , it follow s th at the re lative amplitu d e $\gam ma$ of th e $ |2\r a ng le \le ft r igh t ar row | 4\r angle$ R am se y fri nges i s main lydete rmine d by the frac tio n $( a _2+ a_4)/ \sum_ j a_ j$ of s tatesat $t =\ tau_1$ that par tici pate in t heex cit at ion p r ocess. As no ted in[Sectio n \[ se c : t he ory\]]{} already,we k no w that $ a_4\ap p ro x0 $ , leavin g$a_ 2$ a s singl e pa r am eter det ermini n gth e ampli tu de ofth e R ams ey fr i nge. As is illustr atedi n [Fig. \[fig: s etup\_schemat i c\ ] ] {} , oth erstates (mai nly$ |1\r angl e $and $|3\r angle $) wi l l pass through these tup un affec ted by the os cillation. T h ese stat es c r ea t e a background (off set) in th e relativ e tra nsmissio n. As arg u e d in [Se cti on\[s ec: t h eo ry\]]{}, we c a n res tr ict our sel ves toa t wo- sta teap proximati on [@Abe le :2 01 2] . H ence, only $C_ {2 }({ \t au} _4)$a nd $C_ {4}({ \tau }_ 4) $ ar e deter m in e d by[E qn . (\ [eq :t ransm issi o n\_ coeffic ients\])] {}, desc ri bi ng theRamsey fringe d ependenceon ${ \alpha } $ , whileall other states are ig n ored. U nde r the se a ssumption s,$P$ wo uld result in aRamse yfri n g e ofa m pl itu de $\sim 1$. I n o rderto acc ount fo r the contribution s of other states (a gain , ma inl y $ | 1\r an g le$ a nd $|3\rangle$) as well a sl os ses, we in t rod uc e an am plitude para m eter $\ gamma$, a nd an off se t $r _ { off }$ to scal e and sh ift the t h eoret i ca l cur ve[^7].Ev ent ually , this res ultsin the f it fun ction $ $\begin{ aligned}
&P_{\hspace{- 0.5pt} fit}( {\a lpha},\hs pac e {-0 .5pt}\gam ma,\ hspace{-0. 5pt }r_ {off} )\h s pace{ -1pt } =\ hsp a ce{-1 pt}r _ {off}\ | described_below. The_transmission not only depends_on the_phase,_frequency, and_amplitude_of the oscillation,_but also on_the coefficients $a_n$ and_$b_n$ determined by_the_state selectors, and the horizontal velocity distribution $\eta_v$. Here, we assume $a_n\approx b_n$, which_is_valid in_our_experiment_within $\sim 10\%$. Then, from_$\sum_n a_n=1$ and $\sum_n |C_{n}(\tau_1)|^2=|C_{2}(\tau_1)|^2=1$_after the_first state selector, it follows that the relative_amplitude_$\gamma$ of the_$|2\rangle\leftrightarrow|4\rangle$ Ramsey fringes is mainly determined by the fraction_$(a_2+a_4)/\sum_j a_j$ of states at $t=\tau_1$_that participate in_the_excitation_process. As noted in_[Section \[sec:theory\]]{} already, we know that $a_4\approx0$,_leaving $a_2$ as single parameter determining_the amplitude of the Ramsey fringe. As_is illustrated in [Fig. \[fig:setup\_schematic\]]{}, other states_(mainly $|1\rangle$ and $|3\rangle$) will_pass through_the setup unaffected by the_oscillation. These states_create a_background (offset) in_the relative transmission. As argued in_[Section \[sec:theory\]]{}, we can_restrict ourselves to a two-state approximation [@Abele:2012]._Hence,_only $C_{2}({\tau}_4)$ and_$C_{4}({\tau}_4)$_are_determined by_[Eqn. (\[eq:transmission\_coefficients\])]{}, describing the_Ramsey_fringe dependence_on_${\alpha}$, while all other states are_ignored._Under these assumptions, $P$ would result in_a Ramsey fringe of_amplitude_$\sim 1$. In order_to account for the contributions_of other states (again, mainly $|1\rangle$_and $|3\rangle$)_as well_as losses, we introduce an amplitude parameter $\gamma$, and an offset_$r_{off}$ to scale and shift the_theoretical curve [^7]. Eventually, this_results in_the_fit function $$\begin{aligned}
_&P_{\hspace{-0.5pt}fit}({\alpha},\hspace{-0.5pt}\gamma,\hspace{-0.5pt}r_{off})\hspace{-1pt}=\hspace{-1pt}r_{off}\ |
, or the modes $n_*^{(1)}<n_c$ and $n_*^{(2)}>n_c$ in a, of $10^5$ simulation trajectories. For finite-size correction we use $n_c = 10\tau_d$ in a and $n_c = 22\tau_d^{4/5}$ in b. Time is in units of $1/k_1^-$.[]{data-label="fig:kz"}](fig4){width="\linewidth"}
When testing these predictions using simulations of a spatially extended physical system, the finite size of the system causes a truncation of the autocorrelation time. This truncation is usually accounted for using a finite-size correction [@chandran2012kibble]. In our system, a similar truncation of the autocorrelation time is caused by the finite number of molecules. Specifically, the inset of Fig. \[fig:ss\](b) shows that at criticality we have $\tau_c \sim n_c^{1/2}$ for large $n_c$, where $n_c$ sets the typical number of molecules in the system. Therefore, we interpret $n_c$ as a “system size,” and we correct for finite-size effects in the following way. Combining the relation $\tau_c \sim n_c^{1/2}$ with Eqs. \[eq:tauc1\] and \[eq:tauc2\], and Eqs. \[eq:kz1\] and \[eq:kz2\], we obtain $$\begin{aligned}
\label{eq:finite1}
n_c &\sim \tau_d^{2\nu z/(\nu z+1)},\\
\label{eq:finite2}
n_c &\sim \tau_d^{2\nu z/(\nu z + \beta\delta)},\end{aligned}$$ for the driving of $\theta$ or $h$, respectively. We choose $n_c$ arbitrarily for a particular driving time $\tau_d$, and when we choose a new $\tau_d$, we scale $n_c$ appropriately according to Eqs. \[eq:finite1\] and \[eq:finite2\].
This procedure allows us to test the predictions of the Kibble-Zurek mechanism using simulations of the Schlögl model. The results are shown in Fig. \[fig:kz\]. We | , or the modes $ n_*^{(1)}<n_c$ and $ n_*^{(2)}>n_c$ in a, of $ 10 ^ 5 $ simulation trajectories. For finite - size correction we practice $ n_c = 10\tau_d$ in a and $ n_c = 22\tau_d^{4/5}$ in b. Time is in unit of $ 1 / k_1 ^ -$.[]{data - label="fig: kz"}](fig4){width="\linewidth " }
When testing these predictions use simulation of a spatially extended physical organization, the finite size of the organization causes a shortness of the autocorrelation meter. This truncation is normally accounted for using a finite - size correction [ @chandran2012kibble ]. In our system, a exchangeable truncation of the autocorrelation time is caused by the finite number of molecules. Specifically, the insert of Fig. \[fig: ss\](b) shows that at criticality we have $ \tau_c \sim n_c^{1/2}$ for large $ n_c$, where $ n_c$ place the typical number of molecules in the system. consequently, we interpret $ n_c$ as a “ organization size, ” and we correct for finite - size effects in the following direction. Combining the relation $ \tau_c \sim n_c^{1/2}$ with Eqs. \[eq: tauc1\ ] and \[eq: tauc2\ ], and Eqs. \[eq: kz1\ ] and \[eq: kz2\ ], we obtain $ $ \begin{aligned }
\label{eq: finite1 }
n_c & \sim \tau_d^{2\nu z/(\nu z+1)},\\
\label{eq: finite2 }
n_c & \sim \tau_d^{2\nu z/(\nu z + \beta\delta)},\end{aligned}$$ for the driving of $ \theta$ or $ h$, respectively. We choose $ n_c$ arbitrarily for a particular driving time $ \tau_d$, and when we choose a new $ \tau_d$, we scale $ n_c$ appropriately according to Eqs. \[eq: finite1\ ] and \[eq: finite2\ ].
This procedure allows us to test the predictions of the Kibble - Zurek mechanism using simulations of the Schlögl model. The consequence are shown in Fig. \[fig: kz\ ]. We | , or the modes $n_*^{(1)}<n_c$ and $n_*^{(2)}>n_c$ ik a, of $10^5$ simulation trajxctoriea. For fivite-size correction we use $n_r = 10\tqu_d$ ib a and $n_c = 22\tau_d^{4/5}$ in b. Gime is ij units if $1/k_1^-$.[]{vata-label="fig:kz"}](fij4){sidth="\likzwidtg"}
Ahen vesting these ptedictions uving simulatiots oy a spatially extended physical systqm, the gijite size of tre sjseem dauses a truncation of the autocordelatioi time. This trumcation is usually accountfd flr using a finite-slze correctuon [@syandran2012kibbld]. In our sjvtem, a simjlar truncation of the autocorrdlatipn time is cqusfg by the fiiite nlmber of moleglles. Spacificakly, the inset pf Hig. \[fug:ss\](b) shows that at cciticality we have $\twu_c \sim n_w^{1/2}$ yor large $n_c$, where $n_c$ swts tve tfpicxo njmbtr pf molecklea in the sgstem. Theredore, we interpret $n_v$ ww a “system siae,” and wq correct for finite-size effects in the fomlowing way. Combining tye relation $\tau_c \sim j_c^{1/2}$ with Ezs. \[eq:tauc1\] and \[eq:tauc2\], and Eqs. \[eq:kz1\] and \[eq:kz2\], we obtait $$\begmn{xliykcd}
\lacwl{fq:finite1}
n_c &\sim \tau_d^{2\nu z/(\nu z+1)},\\
\label{eq:finite2}
n_c &\sii \tsu_c^{2\nu z/(\nu z + \betc\delta)},\end{aligned}$$ flr jhe driving of $\theta$ or $h$, respectively. We choose $n_c$ aebitrarilr fot a particular driving time $\tau_d$, and whvn ww choose a new $\tau_b$, we scale $n_e$ apprppriayely according to Eqs. \[eq:yinite1\] and \[eq:finihe2\].
This prkzedure allows us to tast the kfedictions of the Kibble-Zucek mzchanism usimg simtlations ov the Schlögl model. The rfsultd dre shown ln Fig. \[fig:kz\]. We | , or the modes $n_*^{(1)}<n_c$ and $n_*^{(2)}>n_c$ of simulation trajectories. finite-size correction we a $n_c = 22\tau_d^{4/5}$ b. Time is units of $1/k_1^-$.[]{data-label="fig:kz"}](fig4){width="\linewidth"} When testing these using simulations of a spatially extended physical system, the finite size of the causes a truncation of the autocorrelation time. This truncation is usually accounted for a correction In system, a similar truncation of the autocorrelation time is caused by the finite number of molecules. the inset of Fig. \[fig:ss\](b) shows that at we have $\tau_c \sim for large $n_c$, where $n_c$ the number of in system. we interpret $n_c$ a “system size,” and we correct for finite-size effects in the following way. Combining the relation $\tau_c n_c^{1/2}$ with and \[eq:tauc2\], Eqs. and we obtain $$\begin{aligned} &\sim \tau_d^{2\nu z/(\nu z+1)},\\ \label{eq:finite2} n_c z + \beta\delta)},\end{aligned}$$ for the driving of $\theta$ $h$, respectively. choose $n_c$ arbitrarily for a particular time $\tau_d$, and when we choose a new we scale $n_c$ appropriately according to Eqs. \[eq:finite1\] and \[eq:finite2\]. This procedure allows us to predictions of the Kibble-Zurek using simulations of Schlögl The are in Fig. We | , or the modes $n_*^{(1)}<n_c$ and $n_*^{(2)}>n_c$ in a, of $10^5$ sImulation tRajecTorIes. foR finIte-sIze correction wE Use $n_C = 10\tau_d$ in a and $n_c = 22\tau_d^{4/5}$ in b. TiMe is iN uNIts oF $1/K_1^-$.[]{dAta-laBel="fig:kZ"}](FiG4){WIdtH="\lInEwiDtH"}
whEn tesTinG these pRedictions UsiNg Simulations oF A sPatially exTenDed physical sYstEm, the fInIte SIze of The SysteM causeS A truncAtion of thE aUTocorrELation tIME. THis tRuncation is usuallY AcCOunted for using A finitE-sIZe CORreCtiOn [@chandran2012KiBble]. IN Our systEM, a SIMIlaR Truncation of tHe autocorreLAtiOn time Is CauSEd by thE finiTe NUmbEr of moleculEs. SpEcificallY, the inSEt of Fig. \[FIg:ss\](b) shOws thaT at CriTicaLItY wE haVe $\TAu_c \SIm N_c^{1/2}$ fOR laRge $n_c$, wheRe $N_c$ Sets tHe tyPICAL numBer Of moLeculEs in the system. theRefoRE, we InterPret $n_C$ as a “SyStem sIze,” and We corReCt for finite-size EffeCts in the fOllOwIng WaY. CombINing thE reLatIon $\tau_c \Sim n_c^{1/2}$ wiTH EqS. \[eQ:TAUc1\] And \[eq:tauc2\], and Eqs. \[eq:kZ1\] aND \[Eq:Kz2\], we obtaIn $$\begiN{AlIgNEd}
\label{eQ:fIniTe1}
n_c &\SIM \tau_d^{2\Nu z/(\nU Z+1)},\\
\lAbel{eq:fiNite2}
n_c &\SIm \TaU_d^{2\nu z/(\nu Z + \bEta\delTa)},\End{AliGned}$$ fOR the DrivinG of $\theta$ Or $h$, reSPectively. We choOSe $n_c$ arbitrariLY fOR A pARticUlaR driving timE $\tau_D$, And wHen wE ChOosE A new $\tAu_d$, we ScALe $N_C$ appropriately accorDiNg to EqS. \[eq:fiNite1\] and \[eq:finiTe2\].
This procEDURe allows Us to TEsT The predictions Of the kibble-ZureK MechanisM usinG simulatIons of the sCHlögl modEl. THe rEsuLts ARE sHown in Fig. \[fig:kZ\]. wE | , or the modes $n_*^{(1)}< n_c$ and $ n_*^{ (2) }>n _c $ in a,of $10^5$ simu l atio n trajectories. For fi nite- si z e co r re ction we use $n _ c =10 \t au_ d$ in a an d $ n_c = 2 2\tau_d^{4 /5} $in b. Time i s i n units of $1 /k_1^-$.[]{d ata -label =" fig : kz"}] (fi g4){w idth=" \ linewi dth"}
Wh en testin g thesep r ed icti ons using simulat i on s of a spatiall y exte nd e dp h ysi cal system, t he fini t e sizeo ft h e sy s tem causes atruncationo f t he aut oc orr e lation time .T his truncation isusually a ccount e d for u s ing a f inite- siz e c orre c ti on [@ ch a ndr a n2 012 k ibb le]. Inou rsyste m, a s i m ilar tr unca tionof the autoco rre lati o n t ime i s cau sedby thefinite numb er of molecules.Spec ifically, th eins et of F i g. \[f ig: ss\ ](b) sh ows tha t at c r i t ic ality we have $\ta u_ c \s im n_c^{ 1/2}$f or l a rge $n_c $, wh ere$ n _c$ s etst he typical numbe r o fmolecul es in th esys tem . The r efor e, weinterpre t $n_ c $ as a “system size,” and we co r r ec t for fi nite-size e ffec t s in the fo llo w ing w ay. C om b in i ng the relation $\t au _c \si m n_c ^{1/2}$ withEqs. \[eq: t a u c1\] and \[e q :t a uc2\], and Eqs . \[e q:kz1\] an d \[eq:kz 2\],we obtai n $$\begi n { aligned}
\l abe l{e q:f i n it e1}
n_c &\sim \ tau_ d^ {2\nu z /(\ nu z+1) },\ \
\ lab el{ eq :finite2}
n_c &\s im \ ta u_ d^{ 2\nuz /(\nu z+\be ta \de lta)} , \end{a ligne d}$$ f or the drivin g o f $\th et a$ or$h$ ,respe ctiv e ly. We cho ose $n_c$ ar b itra ri ly for aparticular dr iv ing time $ \t au_ d$, an d when wechoose a new $\tau_d$,w e scale $n _c$ a ppro priatelyacc ording to Eqs. \ [eq:fi nite1 \] an d \[eq: f i ni te2 \] .
This pr o c edu re al lo ws u s to te st the predictions ofthe Kibble-Zu rek mec h a ni smu si n g s im u lat i o ns of the Schlö gl model.Th e r esults are sho wn in Fig . \[fig :kz\] . We | , or_the modes_$n_*^{(1)}<n_c$ and $n_*^{(2)}>n_c$ in_a, of_$10^5$_simulation trajectories._For_finite-size correction we_use $n_c =_10\tau_d$ in a and_$n_c = 22\tau_d^{4/5}$_in_b. Time is in units of $1/k_1^-$.[]{data-label="fig:kz"}](fig4){width="\linewidth"}
When testing these predictions using simulations of a_spatially_extended physical_system,_the_finite size of the system_causes a truncation of the_autocorrelation time._This truncation is usually accounted for using a_finite-size_correction [@chandran2012kibble]. In_our system, a similar truncation of the autocorrelation time_is caused by the finite number_of molecules. Specifically,_the_inset_of Fig. \[fig:ss\](b) shows that_at criticality we have $\tau_c \sim_n_c^{1/2}$ for large $n_c$, where $n_c$_sets the typical number of molecules in_the system. Therefore, we interpret $n_c$_as a “system size,” and_we correct_for finite-size effects in the_following way. Combining_the relation_$\tau_c \sim n_c^{1/2}$_with Eqs. \[eq:tauc1\] and \[eq:tauc2\], and Eqs. \[eq:kz1\]_and \[eq:kz2\], we_obtain $$\begin{aligned}
\label{eq:finite1}
n_c &\sim \tau_d^{2\nu z/(\nu z+1)},\\
\label{eq:finite2}
n_c_&\sim_\tau_d^{2\nu z/(\nu z_+_\beta\delta)},\end{aligned}$$_for the_driving of $\theta$_or_$h$, respectively._We_choose $n_c$ arbitrarily for a particular_driving_time $\tau_d$, and when we choose a_new $\tau_d$, we scale_$n_c$_appropriately according to Eqs. \[eq:finite1\]_and \[eq:finite2\].
This procedure allows us_to test the predictions of the_Kibble-Zurek mechanism_using simulations_of the Schlögl model. The results are shown in Fig. \[fig:kz\]. We |
)
,\omega\right) =1$$ whenever $m$ is sufficiently large and $\varepsilon$ is sufficiently small.
The proof of next theorem, being a slight modification of that of Theorem 1, will be omitted.
Suppose that $\mathcal{M}$ is a finitely generated von Neumann algebra with a tracial state $\tau$. Suppose that $\{\mathcal{N}_{i}\}_{i=1}^{\infty}$ is an ascending sequence of von Neumann subalgebras of $\mathcal{M}$ such that $\mathfrak{K}_{2}(\mathcal{N}_{i})=0$ for all $i\geq1$ and $\mathcal{M}=\overline{\cup_{i}\mathcal{N}_{i}}^{SOT}$. Then $\mathfrak{K}_{2}(
\mathcal{M})=0$.
A unitary matrix $U$ in $\mathcal{M}_{k}(\mathbb{C})$ is a *Haar unitary matrix* if $\tau_{k}(U^{m})=0$ for all $1\leq m<k$ and $\tau_{k}(U^{k})=1$.
The proof of following lemma can be found in [@GS2] ( see also [@V4]). For the sake of completeness, we also sketch its proof here.
Let $V_{1},V_{2}$ be two Haar unitary matrices in $\mathcal{M}_{k}(\mathbb{C})$. For every $\delta>0$, let $$\Omega(V_{1},V_{2};\delta)=\{U\in\mathcal{U}(k)\ |\ \Vert UV_{1}-V_{2}U\Vert_{2}\leq\delta\}.$$ Then, for every $0< \delta<r$, there exists a set $\{Ball
(U_{\lambda}; \frac {4\delta} r)\}_{\lambda \in\Lambda}$ of $\frac{4\delta}{r}$-balls in $\mathcal{U}(k)$ that cover $\Omega(V_{1},V_{2};\delta)$ with the cardinality of $\Lambda$ satisfying $|\Lambda|\leq\left( \frac{3r}{2\delta}\right)
^{4rk^{2}}$.
** ** Let $D$ be a diagonal unitary matrix, $diag(\lambda
_{1},\ldots
,\lambda _{k})$, where $\lambda _{j}$ is the $j$-th root of unity $1$. Since $V_{1},V_{2}$ are | )
, \omega\right) = 1$$ whenever $ m$ is sufficiently large and $ \varepsilon$ is sufficiently small.
The proof of next theorem, being a slender alteration of that of Theorem 1, will be omitted.
Suppose that $ \mathcal{M}$ is a finitely generated von Neumann algebra with a tracial country $ \tau$. Suppose that $ \{\mathcal{N}_{i}\}_{i=1}^{\infty}$ is an ascending succession of von Neumann subalgebras of $ \mathcal{M}$ such that $ \mathfrak{K}_{2}(\mathcal{N}_{i})=0 $ for all $ i\geq1 $ and $ \mathcal{M}=\overline{\cup_{i}\mathcal{N}_{i}}^{SOT}$. Then $ \mathfrak{K}_{2 } (
\mathcal{M})=0$.
A unitary matrix $ U$ in $ \mathcal{M}_{k}(\mathbb{C})$ is a * Haar unitary matrix * if $ \tau_{k}(U^{m})=0 $ for all $ 1\leq m < k$ and $ \tau_{k}(U^{k})=1$.
The validation of follow lemma can be find in [ @GS2 ] (see besides [ @V4 ]). For the sake of completeness, we also sketch its proof here.
Let $ V_{1},V_{2}$ be two Haar unitary matrices in $ \mathcal{M}_{k}(\mathbb{C})$. For every $ \delta>0 $, let $ $ \Omega(V_{1},V_{2};\delta)=\{U\in\mathcal{U}(k)\ |\ \Vert UV_{1}-V_{2}U\Vert_{2}\leq\delta\}.$$ Then, for every $ 0 < \delta < r$, there exists a fit $ \{Ball
(U_{\lambda }; \frac { 4\delta } r)\}_{\lambda \in\Lambda}$ of $ \frac{4\delta}{r}$-balls in $ \mathcal{U}(k)$ that cover $ \Omega(V_{1},V_{2};\delta)$ with the cardinality of $ \Lambda$ satisfying $ |\Lambda|\leq\left (\frac{3r}{2\delta}\right)
^{4rk^{2}}$.
* * * * Let $ D$ be a aslant unitary matrix, $ diag(\lambda
_ { 1},\ldots
, \lambda _ { k})$, where $ \lambda _ { j}$ is the $ j$-th root of integrity $ 1$. Since $ V_{1},V_{2}$ are | )
,\omeha\right) =1$$ whenever $m$ is rufficiently latgw and $\tarepsimon$ is sjfficiently small.
The proof oh nezt thtjrem, being a slighg modificwtion of thau of Theorem 1, will be omibced.
Suliose chet $\mathcal{M}$ is s finitely generated von Ndulann algebra with a tracial state $\twu$. Supppsf that $\{\mathcal{G}_{i}\}_{i=1}^{\imsty}$ js an ascending sequence of von Nehmann slbalgebras of $\matncal{M}$ such that $\mathfrak{K}_{2}(\mwthcwl{N}_{i})=0$ for all $i\geq1$ wnd $\mathcal{N}=\ovewoine{\cup_{i}\mathzal{N}_{i}}^{SOT}$. Then $\mathfrak{I}_{2}(
\mathcal{M})=0$.
A unitary matrix $U$ in $\oathccl{M}_{k}(\mathbb{C})$ iw a *Vaar unitarb matrpx* if $\tau_{k}(U^{m})=0$ npr all $1\leq m<k$ and $\tau_{k}(U^{k})=1$.
The prmof of following lemma cen be found in [@GS2] ( sge also [@V4]). Xox the sake of completwnwss, wg alsm skdrch ita 'rokf herf.
Lev $V_{1},V_{2}$ be two Haar unitaey matrices in $\mathvaj{N}_{k}(\mathbb{C})$. For every $\dqlta>0$, let $$\Omega(V_{1},V_{2};\delta)=\{U\in\mathcal{U}(k)\ |\ \Vert UV_{1}-B_{2}U\Vert_{2}\leq\delta\}.$$ Then, for every $0< \delta<r$, there gxists a sqt $\{Ball
(U_{\lambda}; \frac {4\delta} r)\}_{\lambda \in\Lambda}$ of $\fraw{4\delte}{r}$-calos kb $\lathcal{U}(k)$ that cover $\Omega(V_{1},V_{2};\delta)$ with the cawsimakity of $\Lambda$ satisfying $|\Lwmnqa|\leq\left( \fraz{3r}{2\deltc}\difht)
^{4rk^{2}}$.
** ** Let $D$ be a dlagonal unitqry matriv, $disg(\lambda
_{1},\ldots
,\lambda _{k})$, where $\lambda _{j}$ is rhe $j$-th root of unnty $1$. Since $V_{1},R_{2}$ are | ) ,\omega\right) =1$$ whenever $m$ is sufficiently $\varepsilon$ sufficiently small. proof of next of of Theorem 1, be omitted. Suppose $\mathcal{M}$ is a finitely generated von algebra with a tracial state $\tau$. Suppose that $\{\mathcal{N}_{i}\}_{i=1}^{\infty}$ is an ascending sequence von Neumann subalgebras of $\mathcal{M}$ such that $\mathfrak{K}_{2}(\mathcal{N}_{i})=0$ for all $i\geq1$ and $\mathcal{M}=\overline{\cup_{i}\mathcal{N}_{i}}^{SOT}$. $\mathfrak{K}_{2}( A matrix in $\mathcal{M}_{k}(\mathbb{C})$ is a *Haar unitary matrix* if $\tau_{k}(U^{m})=0$ for all $1\leq m<k$ and $\tau_{k}(U^{k})=1$. The of following lemma can be found in [@GS2] see also [@V4]). For sake of completeness, we also its here. Let be Haar matrices in $\mathcal{M}_{k}(\mathbb{C})$. every $\delta>0$, let $$\Omega(V_{1},V_{2};\delta)=\{U\in\mathcal{U}(k)\ |\ \Vert UV_{1}-V_{2}U\Vert_{2}\leq\delta\}.$$ Then, for every $0< \delta<r$, there exists a set $\{Ball \frac {4\delta} of $\frac{4\delta}{r}$-balls $\mathcal{U}(k)$ cover with the cardinality satisfying $|\Lambda|\leq\left( \frac{3r}{2\delta}\right) ^{4rk^{2}}$. ** ** a diagonal unitary matrix, $diag(\lambda _{1},\ldots ,\lambda _{k})$, $\lambda _{j}$ the $j$-th root of unity $1$. $V_{1},V_{2}$ are | )
,\omega\right) =1$$ whenever $m$ is suffIciently laRge anD $\vaRepSiLon$ iS sufFiciently small.
tHe prOof of next theorem, being a SlighT mODifiCAtIon of That of THEoREM 1, wiLl Be OmiTtED.
SUpposE thAt $\mathcAl{M}$ is a finiTelY gEnerated von NEUmAnn algebra WitH a tracial staTe $\tAu$. SuppOsE thAT $\{\mathCal{n}_{i}\}_{i=1}^{\inFty}$ is aN AscendIng sequenCe OF von NeUMann subALGeBras Of $\mathcal{M}$ such thaT $\MaTHfrak{K}_{2}(\mathcal{N}_{I})=0$ for alL $i\GEq1$ AND $\maThcAl{M}=\overlinE{\cUp_{i}\maTHcal{N}_{i}}^{Sot}$. THEN $\MatHFrak{K}_{2}(
\mathcal{M})=0$.
a unitary matRIx $U$ In $\mathCaL{M}_{k}(\MAthbb{C})$ Is a *HaAr UNitAry matrix* if $\Tau_{k}(u^{m})=0$ for all $1\lEq m<k$ anD $\Tau_{k}(U^{k})=1$.
THE proof oF folloWinG leMma cAN bE fOunD iN [@gS2] ( sEE aLso [@v4]). for The sake oF cOmPleteNess, WE ALSo skEtcH its Proof Here.
Let $V_{1},V_{2}$ be twO HaAr unITarY matrIces iN $\matHcAl{M}_{k}(\mAthbb{C})$. for evErY $\delta>0$, let $$\Omega(V_{1},v_{2};\delTa)=\{U\in\mathCal{u}(k)\ |\ \verT Uv_{1}-V_{2}U\VeRT_{2}\leq\deLta\}.$$ theN, for eveRy $0< \delta<R$, TheRe EXIStS a set $\{Ball
(U_{\lambda}; \frAc {4\DELtA} r)\}_{\lambda \In\LambDA}$ oF $\fRAc{4\delta}{r}$-BaLls In $\maTHCal{U}(k)$ That COvEr $\Omega(V_{1},v_{2};\delta)$ WItH tHe cardiNaLity of $\laMbdA$ saTisfyINg $|\LaMbda|\leQ\left( \fraC{3r}{2\delTA}\right)
^{4rk^{2}}$.
** ** Let $D$ be A Diagonal unitaRY mATRiX, $Diag(\LamBda
_{1},\ldots
,\lamBda _{k})$, WHere $\LambDA _{j}$ Is tHE $j$-th rOot of UnITy $1$. sInce $V_{1},V_{2}$ are | )
,\omega\right) =1$$ whe never $m$is su ffi cie nt ly l arge and $\varepsi l on$is sufficiently small.
The p r oofo fnexttheorem , b e i ngasl igh tm od ifica tio n of th at of Theo rem 1 , will be om i tt ed.
Suppo sethat $\mathc al{ M}$ is a fi n itely ge nerat ed von Neuman n algebra w i th a t r acial s t a te $\t au$. Suppose that $\ { \mathcal{N}_{i }\}_{i =1 } ^{ \ i nft y}$ is an asc en dings equence of v o n N e umann subalge bras of $\m a thc al{M}$ s uch that $ \math fr a k{K }_{2}(\math cal{ N}_{i})=0 $ fora ll $i\g e q1$ and $\mat hca l{M }=\o v er li ne{ \c u p_{ i }\ mat h cal {N}_{i}} ^{ SO T}$.Then $ \ m athf rak {K}_ {2}(\mathcal{M})= 0$.
Au nit ary m atrix $U$ i n $\m athcal {M}_{ k} (\mathbb{C})$ i s a*Haar uni tar ymat ri x* if $\tau_ {k} (U^ {m})=0$ for al l $1 \l e q m< k$ and $\tau_{k}(U ^{ k } )= 1$.
The proof of f o llowingle mma can b e fou nd i n [ @GS2] (see al s o[@ V4]). F or the s ak e o f c omple t enes s, wealso ske tch i t s proof here.Let $V_{1},V_ { 2} $ be twoHaa r unitary m atri c es i n $\ m at hca l {M}_{ k}(\m at h bb { C})$. For every $\d el ta>0$, let$$\Omega(V_{1 },V_{2};\d e l t a)=\{U\i n\ma t hc a l{U}(k)\ |\ \V ert U V_{1}-V_{2 } U\Vert_{ 2}\le q\delta\ }.$$ Then , for ever y $ 0<\de lta < r $, there exists a set $ \{Ball(U_ {\lambd a}; \f rac {4 \d elta} r)\ }_{\lamb da \ in \L amb da}$o f $\frac {4 \de lt a}{ r}$-b a lls in $\ma thca l{ U} ( k)$ that c o ve r $\Om eg a( V_{1 },V _{ 2};\d elta ) $ w ith the cardinal ity of $ \L am bda$ sa tisfying $|\L am bda|\leq\l ef t(\frac{ 3 r }{2\delt a}\right)
^{4rk^{2}}$.** ** L et$D$ b e adiagonaluni tary m atr i x, $di ag(\la mbda_{ 1}, \ l dots, \ la mbd a_{k})$, wh e r e $ \lamb da _{j }$ is t he $j$-th root ofu nit y $1$. Since$V_ {1}, V _ {2 }$a re | )
,\omega\right) _=1$$ whenever_$m$ is sufficiently large_and $\varepsilon$_is_sufficiently small.
The_proof_of next theorem,_being a slight_modification of that of_Theorem 1, will_be_omitted.
Suppose that $\mathcal{M}$ is a finitely generated von Neumann algebra with a tracial state_$\tau$._Suppose that_$\{\mathcal{N}_{i}\}_{i=1}^{\infty}$_is_an ascending sequence of von_Neumann subalgebras of $\mathcal{M}$ such_that $\mathfrak{K}_{2}(\mathcal{N}_{i})=0$_for all $i\geq1$ and $\mathcal{M}=\overline{\cup_{i}\mathcal{N}_{i}}^{SOT}$. Then $\mathfrak{K}_{2}(
\mathcal{M})=0$.
A unitary_matrix_$U$ in $\mathcal{M}_{k}(\mathbb{C})$_is a *Haar unitary matrix* if $\tau_{k}(U^{m})=0$ for all_$1\leq m<k$ and $\tau_{k}(U^{k})=1$.
The proof of_following lemma can_be_found_in [@GS2] ( see_also [@V4]). For the sake of_completeness, we also sketch its proof_here.
Let $V_{1},V_{2}$ be two Haar unitary matrices_in $\mathcal{M}_{k}(\mathbb{C})$. For every $\delta>0$, let_$$\Omega(V_{1},V_{2};\delta)=\{U\in\mathcal{U}(k)\ |\ \Vert UV_{1}-V_{2}U\Vert_{2}\leq\delta\}.$$ Then,_for every_$0< \delta<r$, there exists a_set $\{Ball
(U_{\lambda}; \frac_{4\delta} r)\}_{\lambda_\in\Lambda}$ of $\frac{4\delta}{r}$-balls_in $\mathcal{U}(k)$ that cover $\Omega(V_{1},V_{2};\delta)$ with_the cardinality of_$\Lambda$ satisfying $|\Lambda|\leq\left( \frac{3r}{2\delta}\right)
^{4rk^{2}}$.
** ** Let $D$_be_a diagonal unitary_matrix,_$diag(\lambda
_{1},\ldots
,\lambda__{k})$, where_$\lambda _{j}$ is_the_$j$-th root_of_unity $1$. Since $V_{1},V_{2}$ are |
\]) tends to the Fick’s free diffusion expression (\[fick\]) [@EGELSTAFF]
### The mode coupling theory \[sec\_mct\]
The Laplace transform of the intermediate scattering function can be generally written as:
$$\begin{aligned}
\tilde{S}_s(Q,s)=\frac{1}{s+Q^2\tilde{U}(Q,s)} \nonumber\end{aligned}$$
being $\tilde{U}(Q,s)$ a generalized frequency and wavevector dependent diffusion coefficient. The mode coupling theory provides a self consistent expression for $\tilde{U}_s(Q,s)$ [@desh_mc], and the resulting self dynamic structure factor reads:
$$S_s(Q,\omega)\approx \frac{1}{\pi} \frac{DQ^2}{\omega^2+(DQ^2)^2}
+\frac{1}{\pi D Q Q^*}\mathrm{Re}G\left(
\frac{i\omega+DQ^2}{\delta DQ^2} \right) \label{mct1}$$
with
$$G(s)=arctan \left(\frac{1}{\sqrt{s-1}} \right) -
\frac{(s-2)\sqrt{s-1}}{s^2} \label{mct2}$$
where $Q^*=16\pi m \rho D^2 / k_BT$ and $\delta=D/(D+\nu)$, being $D$ the diffusion coefficient and $\nu$ the kinematic viscosity.
An estimate of the FWHM can be numerically evaluated [@dej_phd], yielding:
$$\Delta \omega \approx \left
[1-\frac{Q}{Q^*}H(\delta)+O(Q^{3/2})\right ]DQ^2 \label{mct3}$$
with
$$H(\delta) \approx 1.45\delta^{3/2}\left
[1-0.73\delta-0.15\delta^2-O(\delta ^3)\right ] \label{mct4}$$
### The Nelkin-Ghatak model \[sec\_ng\]
Nelkin and Ghatak have considered a dilute gas in which the atomic motion is dominated by binary collisions, with a distribution function obeying a linearized Boltz | \ ]) tends to the Fick ’s free diffusion expression (\[fick\ ]) [ @EGELSTAFF ]
# # # The mood pair theory \[sec\_mct\ ]
The Laplace transform of the intermediate scattering routine can be generally written as:
$ $ \begin{aligned }
\tilde{S}_s(Q, s)=\frac{1}{s+Q^2\tilde{U}(Q, s) } \nonumber\end{aligned}$$
being $ \tilde{U}(Q, s)$ a generalized frequency and wavevector subject dispersion coefficient. The mode yoke hypothesis provides a self consistent construction for $ \tilde{U}_s(Q, s)$ [ @desh_mc ], and the resulting self dynamic structure agent reads:
$ $ S_s(Q,\omega)\approx \frac{1}{\pi } \frac{DQ^2}{\omega^2+(DQ^2)^2 }
+ \frac{1}{\pi D Q Q^*}\mathrm{Re}G\left (
\frac{i\omega+DQ^2}{\delta DQ^2 } \right) \label{mct1}$$
with
$ $ G(s)=arctan \left(\frac{1}{\sqrt{s-1 } } \right) -
\frac{(s-2)\sqrt{s-1}}{s^2 } \label{mct2}$$
where $ Q^*=16\pi m \rho D^2 / k_BT$ and $ \delta = D/(D+\nu)$, being $ D$ the diffusion coefficient and $ \nu$ the kinematic viscosity.
An estimate of the FWHM can be numerically evaluated [ @dej_phd ], yielding:
$ $ \Delta \omega \approx \left
[ 1-\frac{Q}{Q^*}H(\delta)+O(Q^{3/2})\right ] DQ^2 \label{mct3}$$
with
$ $ H(\delta) \approx 1.45\delta^{3/2}\left
[ 1 - 0.73\delta-0.15\delta^2 - O(\delta ^3)\right ] \label{mct4}$$
# # # The Nelkin - Ghatak exemplar \[sec\_ng\ ]
Nelkin and Ghatak have considered a dilute gas in which the nuclear motion is dominated by binary collision, with a distribution function obey a linearized Boltz | \]) tejds to the Fick’s free dinfusion expression (\[fick\]) [@EGELSFAFF]
### The mode coupling theory \[sec\_mct\]
Vhe Oaplaxe transform of the ingermediatv scatterung hunction can be jsneralln wrifben av:
$$\uegin{aligned}
\tilce{S}_s(Q,s)=\frac{1}{v+Q^2\tilde{U}(Q,s)} \nongmcex\end{aligned}$$
being $\tilde{U}(Q,s)$ a generalised freaufncy and wavevgctor qepehdent diffusion coefficient. The mose coupning theory ptovides a self consistent fxprfssion for $\tilde{U}_s(E,s)$ [@desh_mc], abd trw resulting relf dynampe structure factor reads:
$$S_s(Q,\omega)\approx \fraz{1}{\pi} \fxac{DQ^2}{\omega^2+(DW^2)^2}
+\feac{1}{\oh D Q Q^*}\mathcm{Re}G\lvft(
\frac{i\omega+DQ^2}{\delta GQ^2} \righy) \label{mct1}$$
with
$$N(s)=arcvan \oeft(\frac{1}{\sqrt{s-1}} \right) -
\hrac{(s-2)\sqrt{s-1}}{s^2} \label{mct2}$$
rhere $Q^*=16\pi m \rho D^2 / k_BT$ and $\dwlra=D/(D+\no)$, beitg $D$ rhe dirfnsikn coevfirient and $\nh$ the kinemqtic viscosity.
An esuimwnr of the FWHJ can fe numerically evaluated [@dej_phd], yielding:
$$\Dtlta \kmega \approx \left
[1-\frac{Q}{Q^*}Y(\delta)+O(Q^{3/2})\right ]DQ^2 \labep{mct3}$$
with
$$H(\qelta) \approx 1.45\delta^{3/2}\left
[1-0.73\delta-0.15\delta^2-O(\delta ^3)\right ] \labal{mct4}$$
### Ghe Nclkiv-Thwtak model \[sec\_ng\]
Nelkin and Ghatak have considqdec s dilute gas ik which the atomic mltojn is dominatgd by bnhady collisions, with a disttibutiin functijn oneying a linearized Boltz | \]) tends to the Fick’s free diffusion [@EGELSTAFF] The mode theory \[sec\_mct\] The scattering can be generally as: $$\begin{aligned} \tilde{S}_s(Q,s)=\frac{1}{s+Q^2\tilde{U}(Q,s)} being $\tilde{U}(Q,s)$ a generalized frequency and dependent diffusion coefficient. The mode coupling theory provides a self consistent expression for [@desh_mc], and the resulting self dynamic structure factor reads: $$S_s(Q,\omega)\approx \frac{1}{\pi} \frac{DQ^2}{\omega^2+(DQ^2)^2} +\frac{1}{\pi Q \frac{i\omega+DQ^2}{\delta \right) with $$G(s)=arctan \left(\frac{1}{\sqrt{s-1}} \right) - \frac{(s-2)\sqrt{s-1}}{s^2} \label{mct2}$$ where $Q^*=16\pi m \rho D^2 / k_BT$ and $\delta=D/(D+\nu)$, $D$ the diffusion coefficient and $\nu$ the kinematic An estimate of the can be numerically evaluated [@dej_phd], $$\Delta \approx \left ]DQ^2 with \approx 1.45\delta^{3/2}\left [1-0.73\delta-0.15\delta^2-O(\delta ] \label{mct4}$$ ### The Nelkin-Ghatak model \[sec\_ng\] Nelkin and Ghatak have considered a dilute gas in which atomic motion by binary with distribution obeying a linearized | \]) tends to the Fick’s free diffusIon expressIon (\[fiCk\]) [@EgELsTaFF]
### THe moDe coupling theoRY \[sec\_Mct\]
The Laplace transform Of the InTErmeDIaTe scaTtering FUnCTIon CaN bE geNeRAlLy wriTteN as:
$$\begiN{aligned}
\tiLde{s}_s(q,s)=\frac{1}{s+Q^2\tildE{u}(Q,S)} \nonumber\eNd{aLigned}$$
being $\tIldE{U}(Q,s)$ a gEnEraLIzed fReqUency And wavEVector Dependent DiFFusion COefficiENT. THe moDe coupling theory pROvIDes a self consisTent exPrESsION foR $\tiLde{U}_s(Q,s)$ [@desH_mC], and tHE resultINg SELF dyNAmic structure Factor reads:
$$s_S(Q,\oMega)\apPrOx \fRAc{1}{\pi} \frAc{DQ^2}{\oMeGA^2+(DQ^2)^2}
+\Frac{1}{\pi D Q Q^*}\maThrm{re}G\left(
\frAc{i\omeGA+DQ^2}{\deltA dQ^2} \right) \Label{mCt1}$$
wIth
$$g(s)=arCTaN \lEft(\FrAC{1}{\sqRT{s-1}} \RigHT) -
\frAc{(s-2)\sqrt{s-1}}{S^2} \lAbEl{mct2}$$
WherE $q^*=16\PI M \rho d^2 / k_Bt$ and $\Delta=d/(D+\nu)$, being $D$ the DifFusiON coEfficIent aNd $\nu$ ThE kineMatic vIscosItY.
An estimate of thE FWHm can be numEriCaLly EvAluatED [@dej_phD], yiEldIng:
$$\DeltA \omega \aPProX \lEFT
[1-\FrAc{Q}{Q^*}H(\delta)+O(Q^{3/2})\right ]Dq^2 \lABEl{Mct3}$$
with
$$H(\Delta) \aPPrOx 1.45\DElta^{3/2}\left
[1-0.73\DeLta-0.15\DeltA^2-o(\Delta ^3)\RighT ] \LaBel{mct4}$$
### ThE NelkiN-ghAtAk model \[SeC\_ng\]
NelKiN anD GhAtak hAVe coNsiderEd a dilutE gas iN Which the atomic MOtion is dominaTEd BY BiNAry cOllIsions, with a DistRIbutIon fUNcTioN ObeyiNg a liNeARiZEd Boltz | \]) tends to the Fick’s fr ee diffusi on ex pre ssi on (\[ fick \]) [@EGELSTAF F ]
# ## The mode coupling t heory \ [ sec\ _ mc t\]
The Lap l ac e tra ns fo rmof th e int erm ediatescattering fu nc tion can beg en erally wri tte n as:
$$\be gin {align ed }
\ t ilde{ S}_ s(Q,s )=\fra c {1}{s+ Q^2\tilde {U } (Q,s)} \nonumb e r \e nd{a ligned}$$
being$ \t i lde{U}(Q,s)$ a gener al i ze d fre que ncy and wa ve vecto r depend e nt d i ffu s ion coefficie nt. The mod e co upling t heo r y prov idesas elf consistent exp ression f or $\t i lde{U}_ s (Q,s)$[@desh _mc ],andt he r esu lt i ngs el f d y nam ic struc tu re fact or r e a d s :
$ $S_ s(Q, \omeg a)\approx \fr ac{ 1}{\ p i}\frac {DQ^2 }{\o me ga^2+ (DQ^2) ^2}
+ \f rac{1}{\pi D QQ^*} \mathrm{R e}G \l eft (\frac { i\omeg a+D Q^2 }{\delt a DQ^2} \ri gh t ) \l abel{mct1}$$
with
$ $ G( s)=arcta n \lef t (\ fr a c{1}{\sq rt {s- 1}}\ right ) -\ fr ac{(s-2) \sqrt{ s -1 }} {s^2} \ la bel{mc t2 }$$
w here$ Q^*= 16\pim \rho D ^2 /k _BT$ and $\del t a=D/(D+\nu)$, be i n g$ D$ t hediffusion c oeff i cien t an d $ \nu $ thekinem at i cv iscosity.
An estim at e of t he FW HM can be num erically e v a l uated [@ dej_ p hd ] , yielding:
$ $\Del ta \omega\ approx \ left[1-\frac {Q}{Q^*}H ( \ delta)+O (Q^ {3/ 2}) \ri g h t]DQ^2 \label{ m c t3}$ $
with
$$H (\delta ) \ app rox 1. 45 \delta^{3 /2}\left
[ 1- 0. 73 \de lta-0 . 15\delta ^2 -O( \d elt a ^3) \ right] \la bel{ mc t4 } $$
### Th e N e l kin- Gh at ak m ode l\[sec \_ng \ ]
Nelkinand Ghata k h a ve c on si dered a dilute gas i nwhich theat omi c moti o n is domi nated by binary collisi o ns, wit h a dist ribu tion func tio n obey ing a line arized Bolt z | \]) tends_to the_Fick’s free diffusion expression_(\[fick\]) [@EGELSTAFF]
###_The_mode coupling_theory_\[sec\_mct\]
The Laplace transform_of the intermediate_scattering function can be_generally written as:
$$\begin{aligned}
\tilde{S}_s(Q,s)=\frac{1}{s+Q^2\tilde{U}(Q,s)}_\nonumber\end{aligned}$$
being_$\tilde{U}(Q,s)$ a generalized frequency and wavevector dependent diffusion coefficient. The mode coupling theory provides_a_self consistent_expression_for_$\tilde{U}_s(Q,s)$ [@desh_mc], and the resulting_self dynamic structure factor reads:
$$S_s(Q,\omega)\approx_\frac{1}{\pi} \frac{DQ^2}{\omega^2+(DQ^2)^2}
+\frac{1}{\pi_D Q Q^*}\mathrm{Re}G\left(
\frac{i\omega+DQ^2}{\delta DQ^2} \right) \label{mct1}$$
with
$$G(s)=arctan \left(\frac{1}{\sqrt{s-1}} _\right)_-
\frac{(s-2)\sqrt{s-1}}{s^2} \label{mct2}$$
where $Q^*=16\pi_m \rho D^2 / k_BT$ and $\delta=D/(D+\nu)$, being $D$_the diffusion coefficient and $\nu$ the_kinematic viscosity.
An estimate_of_the_FWHM can be numerically_evaluated [@dej_phd], yielding:
$$\Delta \omega \approx \left
[1-\frac{Q}{Q^*}H(\delta)+O(Q^{3/2})\right_]DQ^2 \label{mct3}$$
with
$$H(\delta) \approx 1.45\delta^{3/2}\left
[1-0.73\delta-0.15\delta^2-O(\delta ^3)\right ]_\label{mct4}$$
### The Nelkin-Ghatak model \[sec\_ng\]
Nelkin and Ghatak_have considered a dilute gas in_which the atomic motion is_dominated by_binary collisions, with a distribution_function obeying a_linearized Boltz |
hand, because ${\widetilde}{G}$ is a subgroup of $G_{\mathbb{R}}$, we have that $S(t)$ is also a one parameter subgroup of $G_{\mathbb{R}}$, therefore the curve $\gamma(t)=S(t)\cdot \gamma(0)$ also gives a geodesic in $D$. Since geodesics on ${\widetilde}{D}$ are also geodesics on $D$, we have proved ${\widetilde}{D}$ is totally geodesic in $D$.
We recall that according to the discussion in Section 2.3, we have the following commutative diagram, $$\begin{aligned}
\xymatrix{\mathcal{T}\ar[d]^{\pi_m}\ar[r]^{\Phi}&D\ar[d]^{\pi_D}\\\mathcal{Z}_m\ar[r]^{\Phi_{\mathcal{Z}_m}}&D/\Gamma,}\end{aligned}$$ where $\pi_m:\,\mathcal{T}\rightarrow \mathcal{Z}_m$ is the universal covering map, and $\pi_D:\,D\rightarrow D/\Gamma$ is a projection map. Let us denote the Hodge metric completion of $\mathcal{Z}_m$ by $\mathcal{Z}^H_{_m}$, and let $\mathcal{T}^H_{_m}$ be the universal cover of $\mathcal{Z}^H_{_m}$, for which the details will be given in Section 4.1. By the definition of $\mathcal{Z}^H_{_m}$, there exists the continuous extension map $\Phi^H_{_{\mathcal{Z}_m}}$. Moreover, since $\mathcal{T}$ is the universal cover of the moduli space $\mathcal{Z}_m$. We have the following commutative diagram, $$\begin{aligned}
\label{diagram2}
\xymatrix{\mathcal{T}\ar[r]^{i_{m}}\ar[d]^{\pi_m}&\mathcal{T}^H_{_m}\ar[d]^{\pi_{_m}^H}\ar[r]^{{\Phi}^{H}_{_m}}&D\ar[d]^{\pi_D}\\
\mathcal{Z}_m\ar[r]^{i}&\mathcal{Z}^H_{_m}\ar[r]^{{\Phi}_{_{\mathcal{Z}_m}}^H}&D/\Gamma,
}\end{aligned}$$ where $i$ is the inclusion map, ${i}_{_m}$ is a lifting map of $i\circ | hand, because $ { \widetilde}{G}$ is a subgroup of $ G_{\mathbb{R}}$, we have that $ S(t)$ is also a one parameter subgroup of $ G_{\mathbb{R}}$, consequently the curvature $ \gamma(t)=S(t)\cdot \gamma(0)$ also gives a geodesic in $ D$. Since geodesic on $ { \widetilde}{D}$ are also geodesic on $ D$, we have proved $ { \widetilde}{D}$ is totally geodetic in $ D$.
We recall that according to the discussion in Section 2.3, we suffer the following commutative diagram, $ $ \begin{aligned }
\xymatrix{\mathcal{T}\ar[d]^{\pi_m}\ar[r]^{\Phi}&D\ar[d]^{\pi_D}\\\mathcal{Z}_m\ar[r]^{\Phi_{\mathcal{Z}_m}}&D/\Gamma,}\end{aligned}$$ where $ \pi_m:\,\mathcal{T}\rightarrow \mathcal{Z}_m$ is the universal application map, and $ \pi_D:\,D\rightarrow D/\Gamma$ is a projection function. Let us denote the Hodge metric completion of $ \mathcal{Z}_m$ by $ \mathcal{Z}^H_{_m}$, and let $ \mathcal{T}^H_{_m}$ be the universal cover of $ \mathcal{Z}^H_{_m}$, for which the details will be contribute in Section 4.1. By the definition of $ \mathcal{Z}^H_{_m}$, there exists the continuous extension function $ \Phi^H_{_{\mathcal{Z}_m}}$. Moreover, since $ \mathcal{T}$ is the universal top of the moduli outer space $ \mathcal{Z}_m$. We have the following commutative diagram, $ $ \begin{aligned }
\label{diagram2 }
\xymatrix{\mathcal{T}\ar[r]^{i_{m}}\ar[d]^{\pi_m}&\mathcal{T}^H_{_m}\ar[d]^{\pi_{_m}^H}\ar[r]^{{\Phi}^{H}_{_m}}&D\ar[d]^{\pi_D}\\
\mathcal{Z}_m\ar[r]^{i}&\mathcal{Z}^H_{_m}\ar[r]^{{\Phi}_{_{\mathcal{Z}_m}}^H}&D/\Gamma,
} \end{aligned}$$ where $ i$ is the inclusion map, $ { i}_{_m}$ is a lifting map of $ i\circ | hajd, because ${\widetilde}{G}$ is a subgroup of $Y_{\nathbb{C}}$, we habe that $R(t)$ is also a one parameter snbgriup od $G_{\mathbb{R}}$, therefore tfe curve $\hamma(t)=S(t)\xdot \tamma(0)$ also gives a geodealc in $V$. Since geodesigs on ${\widethlde}{D}$ are also gdobesics on $D$, we have proved ${\widetilde}{Q}$ is toyaply geodesic ig $D$.
Wt rqcalm that according to the discussion in Secuion 2.3, we have the gollowing commutative diaggam, $$\hegin{aligned}
\xymatrlx{\mathcal{T}\at[s]^{\pi_i}\qr[r]^{\Phi}&D\ar[d]^{\pi_A}\\\mathcal{Z}_m\ar[r]^{\Phi_{\mathcam{Z}_m}}&D/\Gamma,}\end{aligned}$$ where $\pi_m:\,\maghcal{C}\rightarrow \mqthfdl{Z}_m$ is the univvrsal coverinn map, atd $\pi_D:\,D\tightarrow D/\Gakma$ is a projection map. Let us denote the Hodde metric ckmpletion of $\mathxao{Z}_m$ bi $\matvcal{X}^Y_{_m}$, xnd lxt $\jathcap{T}^H_{_j}$ be the uhiversal cocer of $\mathcal{Z}^H_{_m}$, fpr qhich the detzils wylj be given in Section 4.1. By the definitiot or $\mathcal{Z}^H_{_m}$, there exisrs the continuous extgnsion map $\Phi^H_{_{\mathcal{Z}_m}}$. Moreover, since $\mathcal{T}$ is the unieersam corcv of rhf moduli space $\mathcal{Z}_m$. We have the followind cpmkutative diagrcm, $$\begin{aligned}
\lsbfl{cyagram2}
\xymatrib{\mathccm{T}\zr[r]^{i_{m}}\ar[d]^{\pi_m}&\mathcal{H}^H_{_m}\ar[d]^{\py_{_m}^H}\ar[e]^{{\Phi}^{H}_{_m}}&D\ar[q]^{\pi_D}\\
\kathcal{Z}_m\ar[r]^{i}&\mathcal{Z}^H_{_m}\ar[r]^{{\Pyi}_{_{\mathcal{Z}_m}}^H}&B/\Ganma,
}\end{aligned}$$ wherz $i$ is the iuclusipn mal, ${i}_{_m}$ is a lifting map oy $i\cird | hand, because ${\widetilde}{G}$ is a subgroup of have $S(t)$ is a one parameter curve \gamma(0)$ also gives geodesic in $D$. geodesics on ${\widetilde}{D}$ are also geodesics $D$, we have proved ${\widetilde}{D}$ is totally geodesic in $D$. We recall that to the discussion in Section 2.3, we have the following commutative diagram, $$\begin{aligned} where \mathcal{Z}_m$ the covering map, and $\pi_D:\,D\rightarrow D/\Gamma$ is a projection map. Let us denote the Hodge metric completion $\mathcal{Z}_m$ by $\mathcal{Z}^H_{_m}$, and let $\mathcal{T}^H_{_m}$ be the cover of $\mathcal{Z}^H_{_m}$, for the details will be given Section By the of there the continuous extension $\Phi^H_{_{\mathcal{Z}_m}}$. Moreover, since $\mathcal{T}$ is the universal cover of the moduli space $\mathcal{Z}_m$. We have the following diagram, $$\begin{aligned} \mathcal{Z}_m\ar[r]^{i}&\mathcal{Z}^H_{_m}\ar[r]^{{\Phi}_{_{\mathcal{Z}_m}}^H}&D/\Gamma, }\end{aligned}$$ $i$ the map, ${i}_{_m}$ is map of $i\circ | hand, because ${\widetilde}{G}$ is a sUbgroup of $G_{\MathbB{R}}$, wE haVe That $s(t)$ is Also a one parameTEr suBgroup of $G_{\mathbb{R}}$, therefOre thE cURve $\gAMmA(t)=S(t)\cDot \gammA(0)$ AlSO GivEs A gEodEsIC iN $D$. SinCe gEodesicS on ${\widetilDe}{D}$ ArE also geodesiCS oN $D$, we have prOveD ${\widetilde}{D}$ iS toTally gEoDesIC in $D$.
WE reCall tHat accORding tO the discuSsIOn in SeCTion 2.3, we hAVE tHe foLlowing commutativE DiAGram, $$\begin{alignEd}
\xymaTrIX{\mATHcaL{T}\aR[d]^{\pi_m}\ar[r]^{\PhI}&D\Ar[d]^{\pi_d}\\\Mathcal{z}_M\aR[R]^{\pHi_{\mAThcal{Z}_m}}&D/\Gamma,}\End{aligned}$$ wHEre $\Pi_m:\,\matHcAl{T}\RIghtarRow \maThCAl{Z}_M$ is the univeRsal Covering mAp, and $\pI_d:\,D\rightARrow D/\GaMma$ is a ProJecTion MAp. leT us DeNOte THe hodGE meTric compLeTiOn of $\mAthcAL{z}_M$ By $\maThcAl{Z}^H_{_M}$, and lEt $\mathcal{T}^H_{_m}$ bE thE uniVErsAl covEr of $\mAthcAl{z}^H_{_m}$, foR which The deTaIls will be given iN SecTion 4.1. By the DefInItiOn Of $\matHCal{Z}^H_{_m}$, TheRe eXists thE continUOus ExTENSiOn map $\Phi^H_{_{\mathcal{Z}_m}}$. moREOvEr, since $\mAthcal{t}$ Is ThE UniversaL cOveR of tHE ModulI spaCE $\mAthcal{Z}_m$. we have THe FoLlowing CoMmutatIvE diAgrAm, $$\begIN{aliGned}
\laBel{diagrAm2}
\xymATrix{\mathcal{T}\ar[R]^{I_{m}}\ar[d]^{\pi_m}&\mathcAL{T}^h_{_M}\Ar[D]^{\Pi_{_m}^H}\Ar[r]^{{\phi}^{H}_{_m}}&D\ar[d]^{\pi_d}\\
\matHCal{Z}_M\ar[r]^{I}&\MaThcAL{Z}^H_{_m}\aR[r]^{{\Phi}_{_{\MaTHcAL{Z}_m}}^H}&D/\Gamma,
}\end{aligneD}$$ wHere $i$ iS the iNclusion map, ${i}_{_m}$ Is a lifting MAP Of $i\circ | hand, because ${\widetild e}{G}$ isa sub gro upof $G_ {\ma thbb{R}}$, weh avethat $S(t)$ is also aone p ar a mete r s ubgro up of $ G _{ \ m ath bb {R }}$ ,t he refor e t he curv e $\gamma( t)= S( t)\cdot \gam m a( 0)$ also g ive s a geodesic in $D$.Si nce geode sic s on${\wid e tilde} {D}$ areal s o geod e sics on $ D$ , we have proved ${\w i de t ilde}{D}$ is t otally g e od e s icin$D$.
We r ec all t h at acco r di n g tot he discussion in Section 2.3 , we h av e t h e foll owing c o mmu tative diag ram, $$\begin {align e d}
\xym a trix{\m athcal {T} \ar [d]^ { \p i_ m}\ ar [ r]^ { \P hi} & D\a r[d]^{\p i_ D} \\\ma thca l { Z } _m\a r[r ]^{\ Phi_{ \mathcal{Z}_m }}& D/\G a mma ,}\en d{ali gned }$ $ whe re $\p i_m:\ ,\ mathcal{T}\righ tarr ow \mathc al{ Z} _m$ i s the univer sal co veringmap, an d $\ pi _ D : \, D\rightarrow D/\Ga mm a $ i s a proj ection ma p. Let us d en ote the H odgemetr i ccompleti on of$ \m at hcal{Z} _m $ by $ \m ath cal {Z}^H _ {_m} $, and let $\m athca l {T}^H_{_m}$ be the universal co v e ro f $\ mat hcal{Z}^H_{ _m}$ , for whi c hthe detai ls wi ll be given in Section 4. 1. By th e def inition of $\ mathcal{Z} ^ H _ {_m}$, t here ex i sts the contin uousextensionm ap $\Phi ^H_{_ {\mathca l{Z}_m}}$ . Moreover , s inc e $ \ma t h ca l{T}$ is theu n iver sa l cover of the mo dul i s pac e $ \m athcal{Z} _m$. Weha ve t he fo llowi n g commut at ive d iag ram,$ $\begi n{ali gned }\l a bel {diagra m 2} \ xyma tr ix {\ma thc al {T}\a r[r] ^ {i_ {m}}\ar [d]^{\pi_ m}& \ math ca l{ T}^H_{_ m}\ar[d]^{\pi _{ _m}^H}\ar[ r] ^{{ \Phi}^ { H }_{_m}}& D\ar[d]^{\pi_D}\\
\math c al{Z}_m \ar [r]^{ i}&\ mathcal{Z }^H _{_m}\ ar[ r ]^{{\P hi}_{_ {\mat hc al{ Z } _m}}^ H } &D /\G am ma,
}\end{ a l ign ed}$$ w here $i$ is the inclusion map , ${ i}_{_m}$ is a li ftin g ma p o f $ i \ci rc | hand,_because ${\widetilde}{G}$_is a subgroup of_$G_{\mathbb{R}}$, we_have_that $S(t)$_is_also a one_parameter subgroup of_$G_{\mathbb{R}}$, therefore the curve_$\gamma(t)=S(t)\cdot \gamma(0)$ also_gives_a geodesic in $D$. Since geodesics on ${\widetilde}{D}$ are also geodesics on $D$, we_have_proved ${\widetilde}{D}$_is_totally_geodesic in $D$.
We recall that_according to the discussion in_Section 2.3,_we have the following commutative diagram, $$\begin{aligned}
\xymatrix{\mathcal{T}\ar[d]^{\pi_m}\ar[r]^{\Phi}&D\ar[d]^{\pi_D}\\\mathcal{Z}_m\ar[r]^{\Phi_{\mathcal{Z}_m}}&D/\Gamma,}\end{aligned}$$ where_$\pi_m:\,\mathcal{T}\rightarrow_\mathcal{Z}_m$ is the_universal covering map, and $\pi_D:\,D\rightarrow D/\Gamma$ is a projection_map. Let us denote the Hodge_metric completion of_$\mathcal{Z}_m$_by_$\mathcal{Z}^H_{_m}$, and let $\mathcal{T}^H_{_m}$_be the universal cover of $\mathcal{Z}^H_{_m}$,_for which the details will be_given in Section 4.1. By the definition_of $\mathcal{Z}^H_{_m}$, there exists the continuous_extension map $\Phi^H_{_{\mathcal{Z}_m}}$. Moreover, since_$\mathcal{T}$ is_the universal cover of the_moduli space $\mathcal{Z}_m$._We have_the following commutative_diagram, $$\begin{aligned}
\label{diagram2}
\xymatrix{\mathcal{T}\ar[r]^{i_{m}}\ar[d]^{\pi_m}&\mathcal{T}^H_{_m}\ar[d]^{\pi_{_m}^H}\ar[r]^{{\Phi}^{H}_{_m}}&D\ar[d]^{\pi_D}\\
\mathcal{Z}_m\ar[r]^{i}&\mathcal{Z}^H_{_m}\ar[r]^{{\Phi}_{_{\mathcal{Z}_m}}^H}&D/\Gamma,
}\end{aligned}$$ where $i$ is the_inclusion map, ${i}_{_m}$_is a lifting map of $i\circ |
& & $\nwarrow $ & \\
$\mathcal{\mathbf{P}}^{c}$ & & $\overset{\Phi }{\rightarrow }$ & & $\mathcal{\mathbf{P}}$%
\end{tabular}$$ Considering $\mathcal{\mathbf{P}}^{c}$ and $\mathcal{\mathbf{P}} $ as base spaces, the map $\overline{\Phi
}$ is induced by $\Phi $. Hence it is a bundle map. We need to check that it is equivariant, i.e., $$\overline{\Phi }\left( \widetilde{\overline{u}}\cdot \gamma \right) =
\overline{\Phi }(\widetilde{\overline{u}})\cdot \rho (\gamma )\text{ \ \ \ \
\ for all }\gamma \in \Gamma.$$ We write $\overline{\Phi }$ explicitly. A map that satisfies all the properties of the above diagram is $$\overline{\Phi }(x^{\prime };\widetilde{\overline{u}})=\Phi (x^{\prime };
\widetilde{u}),$$ where $$p(x^{\prime }) = x$$ and $\widetilde{\overline{u}}$ ,$\widetilde{u}$ are the same pinor frames. Since multiplication by $\gamma $ leaves the fiber invariant, it is trivially true that $\overline{\Phi }$ is an equivariant map since $\Phi $ is itself equivariant. Next we define the difference class of two structures. $\widetilde{\mathcal{\mathbf{P}}}$ and $\widetilde{\mathcal{\mathbf{P}}}^{\prime }$ are two $\Gamma $-structures on $\mathcal{\mathbf{P}}$, where the group actions differ by an isomorphism $\Psi $. The [*difference class*]{} $\delta (\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime })$ is defined to be $$\delta _{ji}(x)=\gamma _{ji}(x)\Psi (\gamma _{ji}^{\prime -1}(x)),\text{ \ \
\ \ x}\in U_{ij} \:.$$ Similarly, we can define $\overline{\delta }$ for the respective double covers. The difference class $\delta (\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime })$ can be shown to be an element of H$
^{1}(\mathcal{\mathbf{B}},\mathbf{Z}_{2}).$ Similarly, $\overline{\delta }(\widetilde{\mathcal{\mathbf{P}}^{ | & & $ \nwarrow $ & \\
$ \mathcal{\mathbf{P}}^{c}$ & & $ \overset{\Phi } { \rightarrow } $ & & $ \mathcal{\mathbf{P}}$%
\end{tabular}$$ Considering $ \mathcal{\mathbf{P}}^{c}$ and $ \mathcal{\mathbf{P } } $ as base spaces, the map $ \overline{\Phi
} $ is induce by $ \Phi $. therefore it is a bundle map. We necessitate to control that it is equivariant, i.e., $ $ \overline{\Phi } \left (\widetilde{\overline{u}}\cdot \gamma \right) =
\overline{\Phi } (\widetilde{\overline{u}})\cdot \rho (\gamma) \text { \ \ \ \
\ for all } \gamma \in \Gamma.$$ We write $ \overline{\Phi } $ explicitly. A map that satisfy all the properties of the above diagram is $ $ \overline{\Phi } (x^{\prime }; \widetilde{\overline{u}})=\Phi (x^{\prime };
\widetilde{u}),$$ where $ $ p(x^{\prime }) = x$$ and $ \widetilde{\overline{u}}$ , $ \widetilde{u}$ are the same pinor skeleton. Since multiplication by $ \gamma $ leaves the character invariant, it is trivially true that $ \overline{\Phi } $ is an equivariant function since $ \Phi $ is itself equivariant. Next we define the difference class of two structures. $ \widetilde{\mathcal{\mathbf{P}}}$ and $ \widetilde{\mathcal{\mathbf{P}}}^{\prime } $ are two $ \Gamma $ -structures on $ \mathcal{\mathbf{P}}$, where the group actions differ by an isomorphism $ \Psi $. The [ * difference course * ] { } $ \delta (\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime }) $ is defined to be $ $ \delta _ { ji}(x)=\gamma _ { ji}(x)\Psi (\gamma _ { ji}^{\prime -1}(x)),\text { \ \
\ \ x}\in U_{ij } \:.$$ Similarly, we can define $ \overline{\delta } $ for the respective double blanket. The difference class $ \delta (\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime }) $ can be show to be an component of H$
^{1}(\mathcal{\mathbf{B}},\mathbf{Z}_{2}).$ Similarly, $ \overline{\delta } (\widetilde{\mathcal{\mathbf{P}}^ { | & & $\nwarrow $ & \\
$\mathcal{\mathnf{P}}^{c}$ & & $\overset{\Pku }{\righvarrow }$ & & $\mathzal{\mathbf{P}}$%
\end{tabular}$$ Considecing $\mathxal{\mathbf{P}}^{c}$ and $\mathcau{\mathbf{P}} $ as base spares, the map $\overline{\Phi
}$ lf insmced yy $\Phi $. Hence it is a bundne map. We need tu eheck that it is equivariant, i.e., $$\overjine{\Phi }\lfft( \widetilde{\oderlpnq{u}}\cdkn \namma \right) =
\overline{\Phi }(\widetilds{\overliie{u}})\cdot \rho (\gamka )\text{ \ \ \ \
\ for all }\gamma \ln \Gwmma.$$ We write $\overpine{\Phi }$ expoicieoy. A map thag satisfies all the prkperties of the above diagram ir $$\ovexline{\Phi }(x^{\ptnne };\ahdetilde{\ovecline{u}})=\Ihi (x^{\prime };
\widetilde{u}),$$ where $$l(x^{\prime }) = x$$ anc $\wmdetulde{\overline{u}}$ ,$\widetildx{u}$ are the same pinot frames. Shnee multiplication by $\tanma $ neavas tfw fkbed mnvzriant, it is triviamly true thqt $\overline{\Phi }$ is sn vauivariant mzp sinse $\Phi $ is itself equivariant. Next we defpne fhe difference class of two structures. $\widetllde{\mathcwl{\mathbf{P}}}$ and $\widetilde{\mathcal{\mathbf{P}}}^{\prime }$ are twm $\Gamjx $-surmgturdw ln $\mathcal{\mathbf{P}}$, where the group actions difssr bj an isomorphism $\Isi $. The [*differencr flsfs*]{} $\delta (\widejilde{\macgczl{\mathbf{P}}},\widetilde{\lathcal{\iathbd{P}}}^{\prime })$ ys drfined to be $$\delta _{ji}(x)=\gamma _{ji}(x)\Psi (\gammc _{ju}^{\prime -1}(x)),\text{ \ \
\ \ x}\iu U_{ij} \:.$$ Similcrly, wg can cefine $\overline{\delta }$ fur tge respectige double zovers. The differdncv cldss $\delta (\widetilde{\mathcal{\iathbf{P}}},\wivetilbe{\mathcau{\matnbf{P}}}^{\pryme })$ can bf shown to be an element of H$
^{1}(\ladhcal{\mathbv{B}},\mathbf{Z}_{2}).$ Similarly, $\overline{\delve }(\widetilde{\majhcdl{\mdthbf{P}}^{ | & & $\nwarrow $ & \\ $\mathcal{\mathbf{P}}^{c}$ $\overset{\Phi }$ & $\mathcal{\mathbf{P}}$% \end{tabular}$$ Considering base the map $\overline{\Phi is induced by $. Hence it is a bundle We need to check that it is equivariant, i.e., $$\overline{\Phi }\left( \widetilde{\overline{u}}\cdot \gamma = \overline{\Phi }(\widetilde{\overline{u}})\cdot \rho (\gamma )\text{ \ \ \ \ \ for all \in We $\overline{\Phi explicitly. A map that satisfies all the properties of the above diagram is $$\overline{\Phi }(x^{\prime };\widetilde{\overline{u}})=\Phi }; \widetilde{u}),$$ where $$p(x^{\prime }) = x$$ and ,$\widetilde{u}$ are the same frames. Since multiplication by $\gamma leaves fiber invariant, is true $\overline{\Phi }$ is equivariant map since $\Phi $ is itself equivariant. Next we define the difference class of two structures. and $\widetilde{\mathcal{\mathbf{P}}}^{\prime two $\Gamma on where group actions differ isomorphism $\Psi $. The [*difference class*]{} is defined to be $$\delta _{ji}(x)=\gamma _{ji}(x)\Psi (\gamma -1}(x)),\text{ \ \ \ x}\in U_{ij} \:.$$ Similarly, can define $\overline{\delta }$ for the respective double The difference class $\delta (\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime })$ can be shown to be an element of H$ $\overline{\delta }(\widetilde{\mathcal{\mathbf{P}}^{ | & & $\nwarrow $ & \\
$\mathcal{\mathbf{P}}^{c}$ & & $\oveRset{\Phi }{\rigHtarrOw }$ & & $\mAthCaL{\matHbf{P}}$%
\End{tabular}$$ ConsIDeriNg $\mathcal{\mathbf{P}}^{c}$ and $\maThcal{\MaTHbf{P}} $ AS bAse spAces, the MAp $\OVErlInE{\PHi
}$ iS iNDuCed by $\phi $. hence it Is a bundle mAp. WE nEed to check thAT iT is equivarIanT, i.e., $$\overline{\PHi }\lEft( \widEtIldE{\OverlIne{U}}\cdot \Gamma \rIGht) =
\oveRline{\Phi }(\wIdETilde{\oVErline{u}})\CDOt \Rho (\gAmma )\text{ \ \ \ \
\ for all }\gamMA \iN \gamma.$$ We write $\ovErline{\phI }$ ExPLIciTly. a map that saTiSfies ALl the prOPeRTIEs oF The above diagrAm is $$\overlinE{\phi }(X^{\prime };\WiDetILde{\oveRline{U}})=\PHI (x^{\pRime };
\widetilDe{u}),$$ wHere $$p(x^{\priMe }) = x$$ and $\WIdetildE{\OverlinE{u}}$ ,$\wideTilDe{u}$ Are tHE sAmE piNoR FraMEs. sinCE muLtiplicaTiOn By $\gamMa $ leAVES The fIbeR invArianT, it is triviallY trUe thAT $\ovErlinE{\Phi }$ iS an eQuIvariAnt map Since $\phI $ is itself equivaRianT. Next we deFinE tHe dIfFerenCE class Of tWo sTructurEs. $\widetILde{\MaTHCAl{\Mathbf{P}}}$ and $\widetildE{\mATHcAl{\mathbf{p}}}^{\prime }$ ARe TwO $\gamma $-strUcTurEs on $\MAThcal{\MathBF{P}}$, Where the Group aCTiOnS differ By An isomOrPhiSm $\PSi $. The [*DIffeRence cLass*]{} $\deltA (\wideTIlde{\mathcal{\matHBf{P}}},\widetilde{\mAThCAL{\mAThbf{p}}}^{\prIme })$ is defineD to bE $$\DeltA _{ji}(x)=\GAmMa _{jI}(X)\Psi (\gAmma _{jI}^{\pRImE -1}(X)),\text{ \ \
\ \ x}\in U_{ij} \:.$$ SimilarlY, wE can deFine $\oVerline{\delta }$ fOr the respeCTIVe double CoveRS. THE difference claSs $\delTa (\widetildE{\Mathcal{\mAthbf{p}}},\widetilDe{\mathcal{\MAThbf{P}}}^{\priMe })$ cAn bE shOwn TO Be An element of H$
^{1}(\mATHcal{\MaThbf{B}},\maThbF{Z}_{2}).$ SimilArlY, $\ovErlIne{\DeLta }(\widetiLde{\mathcAl{\MaThBf{p}}^{ | & & $\nwarrow $ & \\
$ \mathcal{\ mathb f{P }}^ {c }$ & &$\overset{\Phi }{\r ightarrow }$ & & $\ma thcal {\ m athb f {P }}$%\end{ta b ul a r }$$ C on sid er i ng $\ma thc al{\mat hbf{P}}^{c }$an d $\mathcal{ \ ma thbf{P}} $ as base spaces , t he map $ \ov e rline {\P hi
} $ is i n ducedby $\Phi$. Hencei t is ab u nd le m ap. We need to ch e ck that it is equ ivaria nt , i . e .,$$\ overline{\ Ph i }\l e ft( \wi d et i l d e{\ o verline{u}}\c dot \gamma\ rig ht) =\o ver l ine{\P hi }( \w i det ilde{\overl ine{ u}})\cdot \rho( \gamma) \text{\ \ \\
\ fo r al l } \g amm a\ in\ Ga mma . $$We write $ \o verli ne{\ P h i }$ e xpl icit ly. A map that sat isf iesa llthe p roper ties o f the above diag ra m is $$\overlin e{\P hi }(x^{\ pri me }; \w ideti l de{\ov erl ine {u}})=\ Phi (x^ { \pr im e } ;\widetilde{u}),$$wh e r e$$p(x^{\ prime} )=x $$ and $ \w ide tild e { \over line { u} }$ ,$\wi detild e {u }$ are th esame p in orfra mes.S ince multi plicatio n by$ \gamma $ leave s the fiber in v ar i a nt , itistrivially t ruet hat$\ov e rl ine { \Phi}$ is a n e q uivariant map since $ \Phi $ is i tself equivar iant. Nex t w e define the di f ference classof tw o structur e s. $\wid etild e{\mathc al{\mathb f { P}}}$ an d $ \wi det ild e { \m athcal{\mathb f { P}}} ^{ \prime}$are two $\ Gam ma$-s tr uctures o n $\math ca l{ \m at hbf {P}}$ , where t he gr ou p a ction s diffe r byan i so mo r phi sm $\Ps i $ . The[* di ffer enc eclass *]{} $\d elta (\ widetilde {\m a thca l{ \m athbf{P }}},\widetild e{ \mathcal{\ ma thb f{P}}} ^ { \prime } )$ is defined to be $$\ d elta _{ ji} (x)=\ gamm a _{ji}(x )\P si (\g amm a _{ji} ^{\pri me -1 }( x)) , \ text{ \ \
\\x}\in U_{i j } \: .$$ S im ilar ly, wecan define $\overl i ne{ \delta }$ for th e re s p ec tiv e d o ubl ec ove r s . The differenc e class $\ de l ta (\widetil d e{\ ma thcal{\ mathbf{ P}}}, \ widetil de{\mathc al{\mathb f{ P}}} ^ { \pr ime })$ ca n be sho wn to bea n ele m en t ofH$^{1}(\ ma thc al{\m athbf{ B }}, \math bf{Z}_ {2 }).$ S imila rl y, $\ove rline{\delta }(\widetil de{\ma thcal {\m athbf{P}} ^{ | &_ &_$\nwarrow $ & _\\
$\mathcal{\mathbf{P}}^{c}$_&_ &_$\overset{\Phi_}{\rightarrow }$ &_ & $\mathcal{\mathbf{P}}$%
\end{tabular}$$_Considering $\mathcal{\mathbf{P}}^{c}$ and $\mathcal{\mathbf{P}}_$ as base_spaces,_the map $\overline{\Phi
}$ is induced by $\Phi $. Hence it is a bundle map._We_need to_check_that_it is equivariant, i.e., $$\overline{\Phi_}\left( \widetilde{\overline{u}}\cdot \gamma \right) =
\overline{\Phi_}(\widetilde{\overline{u}})\cdot \rho_(\gamma )\text{ \ \ \ \
\ for all_}\gamma_\in \Gamma.$$ We_write $\overline{\Phi }$ explicitly. A map that satisfies all the_properties of the above diagram is_$$\overline{\Phi }(x^{\prime };\widetilde{\overline{u}})=\Phi_(x^{\prime_};
\widetilde{u}),$$_where $$p(x^{\prime }) =_x$$ and $\widetilde{\overline{u}}$ ,$\widetilde{u}$ are the same_pinor frames. Since multiplication by $\gamma_$ leaves the fiber invariant, it is_trivially true that $\overline{\Phi }$ is_an equivariant map since $\Phi_$ is_itself equivariant. Next we define_the difference class_of two_structures. $\widetilde{\mathcal{\mathbf{P}}}$ and_$\widetilde{\mathcal{\mathbf{P}}}^{\prime }$ are two $\Gamma $-structures on_$\mathcal{\mathbf{P}}$, where the_group actions differ by an isomorphism_$\Psi_$. The [*difference_class*]{}_$\delta_(\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime })$_is defined to_be_$$\delta _{ji}(x)=\gamma__{ji}(x)\Psi_(\gamma _{ji}^{\prime -1}(x)),\text{ \ \
\ \_x}\in_U_{ij} \:.$$ Similarly, we can define $\overline{\delta_}$ for the respective_double_covers. The difference class $\delta_(\widetilde{\mathcal{\mathbf{P}}},\widetilde{\mathcal{\mathbf{P}}}^{\prime })$ can be shown_to be an element of H$
^{1}(\mathcal{\mathbf{B}},\mathbf{Z}_{2}).$_Similarly, $\overline{\delta_}(\widetilde{\mathcal{\mathbf{P}}^{ |
,258}$Fm, are well-known examples where model calculations deviate from experimental data [@Flynn+72; @Hoffman+89; @Gonnenwein+99]. The origin of this transition has long been debated [@Warda+02; @Bonneau+06]. Improvements to model calculations can be made by increasing the smoothing range of the Strutinsky shell-correction procedure [@Albertsson+2019a; @Albertsson+2019b] or by applying Langevin dynamics [@Usang+19].
Global mass yield trends
------------------------
Understanding the trends of fission yields across the chart of nuclides is of particular interest to the astrophysical $r$-process of nucleosynthesis [@Cote+18; @Giuliani+18]. To this end, we introduce in what follows three key metrics to classify a given mass yield, $Y(A)$.
[**1: Number of peaks, $N_{\rm p}$.**]{}The fission fragment mass number distribution $Y(A)$ is always symmetric around the midpoint, $\mbox{$1\over2$}A_0$, due to nucleon number conservation $A_{\rm L}+A_{\rm H}=A_0$, but it may exhibit any number of peaks. Purely symmetric fission leads to a single centrally located peak, while single-mode asymmetric fission leads to two peaks located at opposite sides of the midpoint. Bimodal fission also occurs. For example, $^{226}$Th(n,f) exhibits both symmetric and asymmetric components (with comparable peak heights), $^{235}$U(n,f) has two nearly coinciding asymmetric components, in addition to an increasingly prominent symmetric component as the energy is increased, and some nuclei are predicted to have two widely different asymmetric components. In order to assign the value of the peak index $N_{\rm p}$, we proceed as follows. (1) We first spline interpolate the $Y(A)$ curve, creating $Y_\textrm{s}(A)$, to smooth out any minor bumps that may exist which can be misinterpreted as a peak. (2) Next, we count the maxima by computing the first derivative, $\dot{Y}_\textrm{s}(A)=0$ and second derivative $\ddot{Y}_\textrm{s}(A)<0$. (3) Since large features are typically spread out in $A$, we prevent the algorithm from finding major peaks within 10 mass units of another | , 258}$Fm, are well - known examples where model calculations deviate from experimental datum [ @Flynn+72; @Hoffman+89; @Gonnenwein+99 ]. The beginning of this transition has long been debated [ @Warda+02; @Bonneau+06 ]. improvement to mannequin calculations can be made by increase the polish range of the Strutinsky shell - correction procedure [ @Albertsson+2019a; @Albertsson+2019b ] or by apply Langevin dynamics [ @Usang+19 ].
Global mass yield trends
------------------------
Understanding the tendency of fission yields across the chart of nuclides is of particular interest to the astrophysical $ r$-process of nucleosynthesis [ @Cote+18; @Giuliani+18 ]. To this conclusion, we introduce in what follows three key metrics to relegate a given multitude yield, $ Y(A)$.
[ * * 1: Number of peaks, $ N_{\rm p}$.**]{}The fission shard mass number distribution $ Y(A)$ is always symmetric around the midpoint, $ \mbox{$1\over2$}A_0 $, due to nucleon number conservation $ A_{\rm L}+A_{\rm H}=A_0 $, but it may exhibit any number of peaks. Purely symmetric fission leads to a single centrally located peak, while single - mode asymmetrical fission contribute to two peaks settle at diametric side of the midpoint. Bimodal fission also occur. For example, $ ^{226}$Th(n, f) exhibits both symmetric and asymmetric component (with comparable peak heights), $ ^{235}$U(n, f) has two nearly coinciding asymmetric part, in addition to an increasingly prominent symmetric component as the department of energy is increased, and some nucleus are predict to have two widely different asymmetric component. In order to assign the value of the peak exponent $ N_{\rm p}$, we proceed as postdate. (1) We first spline interpolate the $ Y(A)$ curve, create $ Y_\textrm{s}(A)$, to smooth out any minor bulge that may exist which can be misinterpreted as a peak. (2) Next, we consider the maxima by calculate the first derivative, $ \dot{Y}_\textrm{s}(A)=0 $ and second derivative $ \ddot{Y}_\textrm{s}(A)<0$. (3) Since large feature are typically spread out in $ A$, we prevent the algorithm from finding major peak within 10 mass unit of measurement of another | ,258}$Fm, wre well-known examples wmere model calculations deviafe from dxperimental data [@Flynn+72; @Hofflab+89; @Gonbenwein+99]. The origin of ghis trandition hqs libg been deuzted [@Wavba+02; @Bohkeau+06]. Nm'rovements to mpdel calcunations can be mxdz by increasing the smoothing range jf the Xtgutinsky shell-sorrtctyon lgogedure [@Albertsson+2019a; @Albertsson+2019b] od by apklying Langevin dymamics [@Usang+19].
Global mass yifld hrends
------------------------
Understandinh the trendw of dission yielas across uhz chart of huclides is of particular interdst tp the astripyyslwal $r$-procesw of gucleosynthesis [@Cote+18; @Giuliami+18]. To this end, we inrroduce in what folloxs three key metrics to classhfv a given mass yield, $T(A)$.
[**1: Numbgr of peayw, $N_{\fm k}$.**]{}Thx fjssion fregment mass number disrribution $Y(A)$ is alwsyf symmetric arkund tre midpoint, $\mbox{$1\over2$}A_0$, due to nucleon numbtr cohservation $A_{\rm L}+A_{\rm H}=A_0$, vut it may exhibit ani number os peaks. Purely symmetric fission leads to a singla cenvrxllv locagwd peak, while single-mode asymmetric fission leaqa uo nwo peaks located at oppositr dicgs of the midpuint. Bnjosal fission also ofcurs. Fjr exqmple, $^{226}$Th(n,s) exnibits both symmetric and awymmetric coiponents (with compaxable peak hzights), $^{235}$U(n,f) nas two nearly coincidiug asyjmetric comoonents, ih addition to an kncgeashngly prominent symmetric somponent as che enerey ix incrqased, and dome kgclei are predictef to kave dwo widely different asymmetric components. In order to ssvigt the vajue on the peak indev $N_{\rm p}$, we proeeed as yollowr. (1) We firsn spline mnterpolate ehe $Y(A)$ curve, wteating $Y_\textcm{s}(A)$, to siooty our any mkvor bumps that may exisn chich can be misinterpreted as z peak. (2) Next, we coobt the maxima ny zom[uninj the xirst derivadive, $\dog{U}_\textfm{s}(A)=0$ and segona detivative $\ddot{Y}_\textrm{v}(A)<0$. (3) Aince large featurrs are typucally s[read out in $S$, we prevent the apgorivhm frum fimdigg major peaks within 10 mass unjts of anlthcr | ,258}$Fm, are well-known examples where model calculations experimental [@Flynn+72; @Hoffman+89; The origin of debated @Bonneau+06]. Improvements to calculations can be by increasing the smoothing range of Strutinsky shell-correction procedure [@Albertsson+2019a; @Albertsson+2019b] or by applying Langevin dynamics [@Usang+19]. Global mass trends ------------------------ Understanding the trends of fission yields across the chart of nuclides of interest the $r$-process of nucleosynthesis [@Cote+18; @Giuliani+18]. To this end, we introduce in what follows three key metrics classify a given mass yield, $Y(A)$. [**1: Number peaks, $N_{\rm p}$.**]{}The fission mass number distribution $Y(A)$ is symmetric the midpoint, due nucleon conservation $A_{\rm L}+A_{\rm but it may exhibit any number of peaks. Purely symmetric fission leads to a single centrally located while single-mode leads to peaks at sides of the fission also occurs. For example, $^{226}$Th(n,f) and asymmetric components (with comparable peak heights), $^{235}$U(n,f) two nearly asymmetric components, in addition to an prominent symmetric component as the energy is increased, some nuclei are predicted to have two widely different asymmetric components. In order to assign of the peak index p}$, we proceed follows. We spline the $Y(A)$ creating $Y_\textrm{s}(A)$, to smooth out any minor bumps that may exist can be misinterpreted as a peak. (2) Next, we count by the first derivative, and second derivative $\ddot{Y}_\textrm{s}(A)<0$. Since features are typically spread $A$, prevent finding peaks 10 mass units of | ,258}$Fm, are well-known examples wheRe model calCulatIonS deViAte fRom eXperimental datA [@flynN+72; @Hoffman+89; @Gonnenwein+99]. The oRigin Of THis tRAnSitioN has lonG BeEN DebAtEd [@warDa+02; @bOnNeau+06]. IMprOvementS to model caLcuLaTions can be maDE bY increasinG thE smoothing raNge Of the STrUtiNSky shEll-CorreCtion pROcedurE [@AlbertssOn+2019A; @albertSSon+2019b] or bY APpLyinG Langevin dynamics [@uSaNG+19].
Global mass yieLd trenDs
------------------------
uNdERStaNdiNg the trendS oF fissIOn yieldS AcROSS thE Chart of nuclidEs is of partiCUlaR interEsT to THe astrOphysIcAL $r$-pRocess of nucLeosYnthesis [@COte+18; @GiuLIani+18]. To tHIs end, we IntrodUce In wHat fOLlOwS thReE Key MEtRicS To cLassify a GiVeN mass YielD, $y(a)$.
[**1: nUmbeR of PeakS, $N_{\rm p}$.**]{}the fission fraGmeNt maSS nuMber dIstriButiOn $y(A)$ is aLways sYmmetRiC around the midpoInt, $\mBox{$1\over2$}A_0$, dUe tO nUclEoN numbER conseRvaTioN $A_{\rm L}+A_{\rM H}=A_0$, but iT May ExHIBIt Any number of peaks. PuReLY SyMmetric fIssion LEaDs TO a single CeNtrAlly LOCated Peak, WHiLe single-Mode asYMmEtRic fissIoN leads To Two PeaKs locATed aT opposIte sides Of the MIdpoint. Bimodal FIssion also occURs. fOR eXAmplE, $^{226}$Th(N,f) exhibits bOth sYMmetRic aND aSymMEtric CompoNeNTs (WIth comparable peak heIgHts), $^{235}$U(n,f) Has twO nearly coinciDing asymmeTRIC componeNts, iN AdDItion to an increAsingLy prominenT SymmetriC compOnent as tHe energy iS INcreased, And SomE nuCleI ARe Predicted to haVE Two wIdEly diffEreNt asymmEtrIc cOmpOneNtS. In order tO assign tHe VaLuE oF thE peak INdex $N_{\rm p}$, We ProCeEd aS follOWs. (1) We fiRst spLine InTeRPolAte the $Y(a)$ CuRVE, creAtInG $Y_\teXtrM{s}(a)$, to smOoth OUt aNy minor Bumps that May EXist WhIcH can be mIsinterpreted As A peak. (2) Next, wE cOunT the maXIMa by compUting the first derivative, $\DOt{Y}_\textRm{s}(a)=0$ and sEconD derivatiVe $\dDot{Y}_\teXtrM{S}(A)<0$. (3) SincE large FeatuReS arE TYpicaLLY sPreAd Out in $A$, we prEVEnt The alGoRithM from fiNding major peaks witHIn 10 mAss units of anoTheR | ,258}$Fm, are well-known e xamples wh ere m ode l c al cula tion s deviate from expe rimental data [@Flynn+ 72; @ Ho f fman + 89 ; @Go nnenwei n +9 9 ] . T he o rig in of this tr ansitio n has long be en debated [@W a rd a+02; @Bon nea u+06]. Impro vem ents t omod e l cal cul ation s canb e made by incre as i ng the smoothi n g r ange of the Strutinsk y s h ell-correction proce du r e[ @ Alb ert sson+2019a ;@Albe r tsson+2 0 19 b ] orb y applying La ngevin dyna m ics [@Usa ng +19 ] .
Glo bal m as s yi eld trends---- --------- ------ - ----
U n derstan ding t hetre ndso ffi ssi on yie l ds ac r oss the cha rt o f nuc lide s i s ofpar ticu lar i nterest to th e a stro p hys ical$r$-p roce ss of n ucleos ynthe si s [@Cote+18; @G iuli ani+18].Toth isen d, we introd uce in what f ollowst hre ek e y m etrics to classify a g iv en massyield, $Y (A ) $.
[**1 :Num bero f peak s, $ N _{ \rm p}$. **]{}T h efi ssion f ra gmentma ssnum ber d i stri bution $Y(A)$is al w ays symmetrica round the mid p oi n t ,$ \mbo x{$ 1\over2$}A_ 0$,d ue t o nu c le onn umber cons er v at i on $A_{\rm L}+A_{\r mH}=A_0 $, bu t it may exhi bit any nu m b e r of pea ks.P ur e ly symmetric f issio n leads to a single cent rally lo cated pea k , while s ing le- mod e a s y mm etric fission l eads t o two p eak s locat edatopp osi te sides of the mid po in t. B imo dal f i ssion al so oc cu rs. Fore xample , $^{ 226} $T h( n ,f) exhibi t sb o th s ym me tric an dasymm etri c co mponent s (with c omp a rabl epe ak heig hts), $^{235} $U (n,f) hastw o n earlyc o inciding asymmetric components, in addi tio n toan i ncreasing lypromin ent symmet ric co mpone nt as t he en e r gy is i ncreased,a n d s ome n uc leiare pre dicted to have two wid ely different as ymme t r ic co m po n ent s. Ino r der to assign t he value o ft he peak inde x $N _{ \rm p}$ , we pr oceed as foll ows. (1)We firstsp line i nte rpolate th e $Y(A)$ curve, c r eatin g $ Y_\te xtr m{s}(A )$ , t o smo oth ou t an y min or bum ps thatmay e xi st which can be misinterpretedas a p eak.(2) Next, we co u ntthe maxim a by computing th e f irstder i vativ e, $ \ do t{Y } _\tex trm{ s }(A)=0$ a n dsec o n dderivative$ \ d dot {Y}_\ tex t rm{s}( A)<0 $. (3) Since larg e features aretypi c a lly sp r eadou t in $A$, we p rev en t the algo ri thm from fi nding ma jo r peak s with in 10mass un i t so f anot her | ,258}$Fm, are_well-known examples_where model calculations deviate_from experimental_data_[@Flynn+72; @Hoffman+89;_@Gonnenwein+99]._The origin of_this transition has_long been debated [@Warda+02;_@Bonneau+06]. Improvements to_model_calculations can be made by increasing the smoothing range of the Strutinsky shell-correction procedure_[@Albertsson+2019a;_@Albertsson+2019b] or_by_applying_Langevin dynamics [@Usang+19].
Global mass yield_trends
------------------------
Understanding the trends of fission_yields across_the chart of nuclides is of particular interest_to_the astrophysical $r$-process_of nucleosynthesis [@Cote+18; @Giuliani+18]. To this end, we introduce_in what follows three key metrics_to classify a_given_mass_yield, $Y(A)$.
[**1: Number of_peaks, $N_{\rm p}$.**]{}The fission fragment mass_number distribution $Y(A)$ is always symmetric_around the midpoint, $\mbox{$1\over2$}A_0$, due to nucleon_number conservation $A_{\rm L}+A_{\rm H}=A_0$, but_it may exhibit any number_of peaks._Purely symmetric fission leads to_a single centrally_located peak,_while single-mode asymmetric_fission leads to two peaks located_at opposite sides_of the midpoint. Bimodal fission also_occurs._For example, $^{226}$Th(n,f)_exhibits_both_symmetric and_asymmetric components (with_comparable_peak heights),_$^{235}$U(n,f)_has two nearly coinciding asymmetric components,_in_addition to an increasingly prominent symmetric component_as the energy is_increased,_and some nuclei are_predicted to have two widely_different asymmetric components. In order to_assign the_value of_the peak index $N_{\rm p}$, we proceed as follows. (1) We_first spline interpolate the $Y(A)$ curve,_creating $Y_\textrm{s}(A)$, to smooth_out any_minor_bumps that may_exist_which can_be misinterpreted as a peak. (2) Next,_we count_the maxima by computing the first_derivative, $\dot{Y}_\textrm{s}(A)=0$ and second_derivative_$\ddot{Y}_\textrm{s}(A)<0$. (3) Since large features are_typically spread out in $A$, we_prevent the algorithm from finding_major_peaks_within 10 mass units of_another |
\begin{pmatrix}
-\mu-v_Fk & 0 & 0 & -\Delta^{h}_{\vec{k}} \\
0 & -\mu+v_Fk & \Delta^{h}_{\vec{k}} & 0 \\
0 &\Delta^{h*}_{\vec{k}} & \mu- v_Fk &0 \\
-\Delta^{h*}_{\vec{k}} &0& 0&\mu+v_Fk
\end{pmatrix},\end{aligned}$$ provided that the holon pairing parameter $\Delta^{h}_{\vec{k}}$ is p-wave like, i.e., $\Delta^h_{-\vec{k}} = -\Delta^h_{\vec{k}}$. The spectrum of quasiparticles consists of two decoupled branches, $$\begin{aligned}
\label{eq:67}
\epsilon_{\pm,\vec{k}}^h = \sqrt{(v_Fk\pm \mu)^2+|\Delta^h_{\vec{k}}|^2}.\end{aligned}$$ The free energy at temperature $T$ then reads $$\begin{aligned}
\label{eq:68}
F=-T\sum_{i=\pm, \vec{k}} \left[
\ln(1+e^{-\frac{\epsilon_{i \vec{k}}}{T}})+\ln(1+
e^{\frac{\epsilon_{i \vec{k}}}{T}}) \right].\end{aligned}$$ According to Hellman-Feynman theorem, we have the gap equation for order parameter $\Delta_{\vec{q}}^h$, $$\begin{aligned}
\label{eq:69}
&&\langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rangle = \frac{\partial F}{\partial
\Delta_{\vec{q}}^{h*}} =-\sum_{i=\pm}
\frac{\Delta_{\vec{q}}^{h}}{2 \epsilon_{i,\vec{q}}^h}\tanh \left(
\frac{\epsilon_{i,\vec{q}}}{2T} \right) \nonumber\\
&&\Delta_{\vec{k}}^h=\sum_{\vec{q}}V_{\text{eff}}(\vec{k}-\vec{q})
\langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rangle.\ | \begin{pmatrix }
-\mu - v_Fk & 0 & 0 & -\Delta^{h}_{\vec{k } } \\
0 & -\mu+v_Fk & \Delta^{h}_{\vec{k } } & 0 \\
0 & \Delta^{h*}_{\vec{k } } & \mu- v_Fk & 0 \\
-\Delta^{h*}_{\vec{k } } & 0 & 0&\mu+v_Fk
\end{pmatrix},\end{aligned}$$ provided that the holon pairing parameter $ \Delta^{h}_{\vec{k}}$ is p - wave like, i.e., $ \Delta^h_{-\vec{k } } = -\Delta^h_{\vec{k}}$. The spectrum of quasiparticle consist of two decoupled branches, $ $ \begin{aligned }
\label{eq:67 }
\epsilon_{\pm,\vec{k}}^h = \sqrt{(v_Fk\pm \mu)^2+|\Delta^h_{\vec{k}}|^2}.\end{aligned}$$ The spare department of energy at temperature $ T$ then reads $ $ \begin{aligned }
\label{eq:68 }
F=-T\sum_{i=\pm, \vec{k } } \left [
\ln(1+e^{-\frac{\epsilon_{i \vec{k}}}{T}})+\ln(1 +
e^{\frac{\epsilon_{i \vec{k}}}{T } }) \right].\end{aligned}$$ According to Hellman - Feynman theorem, we have the opening equation for order argument $ \Delta_{\vec{q}}^h$, $ $ \begin{aligned }
\label{eq:69 }
& & \langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q } } \rangle = \frac{\partial F}{\partial
\Delta_{\vec{q}}^{h * } } = -\sum_{i=\pm }
\frac{\Delta_{\vec{q}}^{h}}{2 \epsilon_{i,\vec{q}}^h}\tanh \left (
\frac{\epsilon_{i,\vec{q}}}{2 metric ton } \right) \nonumber\\
& & \Delta_{\vec{k}}^h=\sum_{\vec{q}}V_{\text{eff}}(\vec{k}-\vec{q })
\langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q } } \rangle.\ | \bfgin{pmatrix}
-\mu-v_Fk & 0 & 0 & -\Delta^{h}_{\vec{k}} \\
0 & -\mu+v_Hk & \Delfa^{h}_{\vec{k}} & 0 \\
0 &\Delta^{h*}_{\vec{k}} & \mu- v_Fk &0 \\
-\Deluc^{h*}_{\vec{k}} &0& 0&\mu+v_Fk
\end{pmxtrix},\end{apigned}$$ peovived that the holon pairiky parzletex $\Velta^{h}_{\vec{k}}$ is p-eave like, h.e., $\Delta^h_{-\vec{k}} = -\Ddlca^h_{\vec{k}}$. The spectrum of quasiparticlqs consoshs of two decoopled frandhes, $$\begin{aligned}
\label{eq:67}
\epsilon_{\pm,\vsc{k}}^h = \sxrt{(v_Fk\pm \mu)^2+|\Dekta^h_{\vec{k}}|^2}.\end{aligned}$$ The fref enfrgy at temperaturf $T$ then reqds $$\fwgin{aligned}
\lxbel{eq:68}
F=-T\sum_{i=\pm, \vec{k}} \lert[
\ln(1+e^{-\frac{\epsilon_{i \vec{k}}}{T}})+\ln(1+
e^{\ffac{\epxilon_{i \vec{j}}}{T}}) \rihvt].\end{alignev}$$ Accogding to Hellman-Feynmdn theotem, we have thc gap eqyation for order paraketer $\Delta_{\vec{q}}^h$, $$\bggin{aligneg}
\lcbel{eq:69}
&&\langle \hat{b}_{-\vec{q}}\yar{a}_{\vec{x}} \ratgle = \frxc{\pzrviam F}{\parhiam
\Delta_{\ved{q}}^{h*}} =-\sum_{i=\pm}
\feac{\Delta_{\vec{q}}^{h}}{2 \epsilpn_{y,\nrc{q}}^h}\tanh \leff(
\fras{\e[silon_{i,\vec{q}}}{2T} \right) \nonumber\\
&&\Delta_{\vec{k}}^h=\suk_{\ved{q}}V_{\text{eff}}(\vec{k}-\vec{q})
\langlw \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rajgle.\ | \begin{pmatrix} -\mu-v_Fk & 0 & 0 & 0 -\mu+v_Fk & & 0 \\ &0 -\Delta^{h*}_{\vec{k}} &0& 0&\mu+v_Fk provided that the pairing parameter $\Delta^{h}_{\vec{k}}$ is p-wave like, $\Delta^h_{-\vec{k}} = -\Delta^h_{\vec{k}}$. The spectrum of quasiparticles consists of two decoupled branches, $$\begin{aligned} \epsilon_{\pm,\vec{k}}^h = \sqrt{(v_Fk\pm \mu)^2+|\Delta^h_{\vec{k}}|^2}.\end{aligned}$$ The free energy at temperature $T$ then reads $$\begin{aligned} F=-T\sum_{i=\pm, \left[ \vec{k}}}{T}})+\ln(1+ \vec{k}}}{T}}) \right].\end{aligned}$$ According to Hellman-Feynman theorem, we have the gap equation for order parameter $\Delta_{\vec{q}}^h$, $$\begin{aligned} &&\langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rangle = \frac{\partial F}{\partial \Delta_{\vec{q}}^{h*}} =-\sum_{i=\pm} \epsilon_{i,\vec{q}}^h}\tanh \left( \frac{\epsilon_{i,\vec{q}}}{2T} \right) &&\Delta_{\vec{k}}^h=\sum_{\vec{q}}V_{\text{eff}}(\vec{k}-\vec{q}) \langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rangle.\ | \begin{pmatrix}
-\mu-v_Fk & 0 & 0 & -\Delta^{h}_{\veC{k}} \\
0 & -\mu+v_Fk & \DelTa^{h}_{\veC{k}} & 0 \\
0 &\DEltA^{h*}_{\Vec{k}} & \Mu- v_FK &0 \\
-\Delta^{h*}_{\vec{k}} &0& 0&\mu+v_fK
\end{Pmatrix},\end{aligned}$$ proviDed thAt THe hoLOn PairiNg paramETeR $\dEltA^{h}_{\VeC{k}}$ iS p-WAvE like, I.e., $\DElta^h_{-\veC{k}} = -\Delta^h_{\veC{k}}$. THe Spectrum of quASiParticles cOnsIsts of two decOupLed braNcHes, $$\BEgin{aLigNed}
\laBel{eq:67}
\ePSilon_{\pM,\vec{k}}^h = \sqrT{(v_fK\pm \mu)^2+|\DELta^h_{\vec{K}}|^2}.\ENd{AligNed}$$ The free energy aT TeMPerature $T$ then rEads $$\beGiN{AlIGNed}
\LabEl{eq:68}
F=-T\sum_{i=\Pm, \Vec{k}} \lEFt[
\ln(1+e^{-\frAC{\ePSILon_{I \Vec{k}}}{T}})+\ln(1+
e^{\frac{\ePsilon_{i \vec{k}}}{t}}) \RigHt].\end{aLiGneD}$$ accordIng to heLLmaN-Feynman theOrem, We have the Gap equATion for ORder parAmeter $\delTa_{\vEc{q}}^h$, $$\BEgIn{AliGnED}
\laBEl{Eq:69}
&&\lANglE \hat{b}_{-\vec{Q}}\hAt{A}_{\vec{q}} \RangLE = \FRAc{\paRtiAl F}{\pArtiaL
\Delta_{\vec{q}}^{h*}} =-\suM_{i=\pM}
\fraC{\delTa_{\vec{Q}}^{h}}{2 \epsIlon_{I,\vEc{q}}^h}\tAnh \lefT(
\frac{\EpSilon_{i,\vec{q}}}{2T} \righT) \nonUmber\\
&&\DeltA_{\veC{k}}^H=\suM_{\vEc{q}}V_{\tEXt{eff}}(\vEc{k}-\Vec{Q})
\langle \Hat{b}_{-\vec{Q}}\Hat{A}_{\vEC{Q}} \RaNgle.\ | \begin{pmatrix}
-\mu -v_Fk & 0& 0 & -\ Del ta ^{h} _{\v ec{k}} \\
0 & - \mu+v_Fk & \Delta^{h}_ {\vec {k } } & 0\\
0 &\D e lt a ^ {h* }_ {\ vec {k } }& \mu - v _Fk &0\\
-\D elt a^ {h*}_{\vec{k } }&0& 0&\mu+ v_F k
\end{pma tri x},\en d{ ali g ned}$ $ p rovid ed tha t the h olon pair in g param e ter $\D e l ta ^{h} _{\vec{k}}$ is p- w av e like, i.e., $ \Delta ^h _ {- \ v ec{ k}} = -\Delta ^h _{\ve c {k}}$.T he s p ect r um of quasipa rticles con s ist s of t wo de c oupled bran ch e s,$$\begin{al igne d}
\label {eq:67 }
\epsil o n_{\pm, \vec{k }}^ h = \sq r t{ (v _Fk \p m \m u )^ 2+| \ Del ta^h_{\v ec {k }}|^2 }.\e n d { a lign ed} $$ T he fr ee energy attem pera t ure $T$thenread s$$\be gin{al igned }\label{eq:68}
F =-T\ sum_{i=\p m,\v ec{ k} } \le f t[
\ ln( 1+e ^{-\fra c{\epsi l on_ {i \ v ec {k}}}{T}})+\ln(1+ e ^ {\ frac{\ep silon_ { i\v e c{k}}}{T }} ) \ righ t ] .\end {ali g ne d}$$ Acc ording to H ellman- Fe ynmanth eor em, we h a ve t he gap equatio n for order paramete r $\Delta_{\ve c {q } } ^h $ , $$ \be gin{aligned }
\l a bel{ eq:6 9 }&&\ l angle \hat {b } _{ - \vec{q}}\hat{a}_{\v ec {q}} \ rangl e = \frac{\pa rtial F}{\ p a r tial
\ Delt a _{ \ vec{q}}^{h*}}=-\su m_{i=\pm}\ frac{\De lta_{ \vec{q}} ^{h}}{2 \ e p silon_{i ,\v ec{ q}} ^h} \ t an h \left(
\f r a c{\e ps ilon_{i ,\v ec{q}}} {2T } \ rig ht) \ nonumber\ \
&&\Del ta _{ \v ec {k} }^h=\ s um_{\vec {q }}V _{ \te xt{ef f }}(\ve c{k}- \vec {q }) \la ngle \h a t{ b } _{-\ ve c{ q}}\ hat {a }_{\v ec{q } } \ rangle. \ | _\begin{pmatrix}
_ -\mu-v_Fk &_0 &_0_& -\Delta^{h}_{\vec{k}}_\\
_ _0 & -\mu+v_Fk_& \Delta^{h}_{\vec{k}} &_0 \\
__ 0 &\Delta^{h*}_{\vec{k}} & \mu- v_Fk &0 \\
-\Delta^{h*}_{\vec{k}} &0& 0&\mu+v_Fk
__\end{pmatrix},\end{aligned}$$ provided_that_the_holon pairing parameter $\Delta^{h}_{\vec{k}}$ is_p-wave like, i.e., $\Delta^h_{-\vec{k}} =_-\Delta^h_{\vec{k}}$. The_spectrum of quasiparticles consists of two decoupled branches,_$$\begin{aligned}
\label{eq:67}
\epsilon_{\pm,\vec{k}}^h_= \sqrt{(v_Fk\pm \mu)^2+|\Delta^h_{\vec{k}}|^2}.\end{aligned}$$_The free energy at temperature $T$ then reads $$\begin{aligned}
\label{eq:68}
F=-T\sum_{i=\pm,_\vec{k}} \left[
\ln(1+e^{-\frac{\epsilon_{i \vec{k}}}{T}})+\ln(1+
_e^{\frac{\epsilon_{i \vec{k}}}{T}}) \right].\end{aligned}$$_According_to_Hellman-Feynman theorem, we have_the gap equation for order parameter_$\Delta_{\vec{q}}^h$, $$\begin{aligned}
\label{eq:69}
&&\langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rangle = \frac{\partial_F}{\partial
\Delta_{\vec{q}}^{h*}} =-\sum_{i=\pm}
\frac{\Delta_{\vec{q}}^{h}}{2 \epsilon_{i,\vec{q}}^h}\tanh \left(
_\frac{\epsilon_{i,\vec{q}}}{2T} \right) \nonumber\\
&&\Delta_{\vec{k}}^h=\sum_{\vec{q}}V_{\text{eff}}(\vec{k}-\vec{q})
\langle \hat{b}_{-\vec{q}}\hat{a}_{\vec{q}} \rangle.\ |
isospin-0 vectors could possibly be connected to the $Y(3940)\rightarrow J/\psi\omega$ and $Y(4140)\rightarrow J/\psi \phi$ since both $J/\psi$ and $\phi$ have isospin-0 and are very close to the two vector threshold. Second, it is striking that there are at least two structures observed in the same $J/\psi\phi$ spectrum from the same exclusive $B$ decays. The two structures do not need to arise from the same mechanism, but if they are, this could be a pointer to a new kind of spectroscopy. It will also be interesting to search for vector-vector structures composed entirely of c and b quarks near threshold because they may offer simpler systems to model theoretically.
Summary
=======
The discovery of the $X(3872)$ in 2003 initiated a new chapter in the study of exotic states. The evidence found in the $J/\psi\phi$ mass spectrum from exclusive $B$ decays and two-photon process provide several exotic charmonium candidates: they all have a mass well beyond the open-charm pair threshold, a relatively narrow width, positive C-parity, and do not fit into the conventional charmonium picture. All of these reported structures need further studies. Various possible interpretations such as molecule, tetra-quark, charmed hybrid, nuclei-like structures have been proposed but none of them is established. The $pp$ run at the LHC is now over until 2015, but CMS has roughly four times more data on tape to be analyzed and LHCb is expected to have about ten times the data to be analyzed. ATLAS, on the other hand, has not yet been heard from but they can also contribute to the current studies. The properties of these structures, such as mass, width, decay channels and quantum numbers, can be studied with increased statistics. At the LHC one can also search in other heavy quarkonium decay channels for vector-vector analogs of the $J/\psi \phi$ structures. With joint experimental and theoretical effort, hopefully we will obtain a clear grand picture and eventually understand the nature of these queer structures—which may indeed lead us to a new kind of spectroscopy.
Acknowledgments {#acknowledgments.unnumbered}
===============
I would like to thank Gerry Bauer, Stanley Brodsky, Joel Butler, Tiziano Camporesi, Kai-feng Chen, Vincenzo Chiochia, Ray Cul | isospin-0 vectors could possibly be connected to the $ Y(3940)\rightarrow J/\psi\omega$ and $ Y(4140)\rightarrow J/\psi \phi$ since both $ J/\psi$ and $ \phi$ have isospin-0 and are very close to the two vector brink. Second, it is strike that there are at least two structures observed in the like $ J/\psi\phi$ spectrum from the same exclusive $ B$ decays. The two social organization do not need to arise from the like mechanism, but if they are, this could be a arrow to a new kind of spectroscopy. It will besides be interesting to search for vector - vector structures composed entirely of hundred and b quarks near threshold because they may offer simpler systems to model theoretically.
Summary
= = = = = = =
The discovery of the $ X(3872)$ in 2003 originate a new chapter in the study of exotic state. The evidence found in the $ J/\psi\phi$ mass spectrum from exclusive $ B$ decays and two - photon process supply several exotic charmonium candidates: they all have a mass well beyond the clear - charm pair threshold, a relatively narrow width, positive C - parity, and do not fit into the conventional charmonium picture. All of these reported structures need further studies. Various possible interpretations such as atom, tetra - quark, capture hybrid, nuclei - like structure have been aim but none of them is established. The $ pp$ run at the LHC is now over until 2015, but CMS has roughly four times more datum on tape to be analyzed and LHCb is expected to have about ten times the data to be analyzed. ATLAS, on the other hand, has not yet been heard from but they can besides contribute to the current report. The property of these structures, such as mass, width, decay channels and quantum numbers, can be studied with increased statistics. At the LHC one can also research in other clayey quarkonium decay channels for vector - vector analogs of the $ J/\psi \phi$ structure. With joint experimental and theoretical effort, hopefully we will prevail a clear grand painting and eventually sympathize the nature of these queer structures — which may indeed lead us to a new kind of spectroscopy.
acknowledgment { # acknowledgments.unnumbered }
= = = = = = = = = = = = = = =
I would like to thank Gerry Bauer, Stanley Brodsky, Joel Butler, Tiziano Camporesi, Kai - feng Chen, Vincenzo Chiochia, Ray Cul | islspin-0 vectors could posslbly be connecteb to thx $Y(3940)\righfarrow J/\osi\omega$ and $Y(4140)\rightarrow J/\psm \phu$ sinxe both $J/\psi$ and $\phi$ hxve isosppn-0 and arw vecy close to the vso vector thrsdholb. Wecond, it is sjriking that there are at nexsc two structures observed in the samq $J/\psi\pni$ spectrum from the fame vxglusive $B$ decays. The two structudes do iot need to arixe from the same mechanism, but if they are, this fould be a kkinewr to a new yind of sptccroscopy. It will also be interesting to sexrch yor vector-vgeror vtructures romposvd entirely on c and b quarls near threshpld bexause they may offer vimpler systems to model thaoxetically.
Summary
=======
The dusxoveri of dhe $B(3872)$ in 2003 ihiviafed a jew chapter ih the study of exotic states. Tne vfidence founs in tre $J/\psi\phi$ mass spectrum from exclusive $B$ dedays and two-photon procwss provide several edotic chawmonium candidates: they all have a mass well beyotd thx upeu-gmarm palr threshold, a relatively narrow width, positids V-psrity, and do njt fit into tje sonventional zharmoujuj picture. All of tjese rekorted structurts nerd further studies. Various possible intvrprwtations such as mllecule, tetxa-quarl, chatmed hybrid, nuclei-like rtrudtures have been prolused but none of thvm iv established. The $pp$ run ae the LHC is uow over untol 2015, bue CMS has goughly four times more fata ln tape to bf analyzed and LHCb is expected to have about tan nimes the data to be analyzqd. ATLAS, on thg other hcnd, har not yet geen heerd from but they can alsm contribute vo the cuwrenr stydies. Tfd properties og these snrbctures, sych as mass, width, vecai dhannels and qucuton numbers, can ne rtuqivd xith ytcreased stadistkcs. St thd LHC one ccn alsp search in other hedvy suarkonium decay cnaknels for vector-vqctor analogs of the $J/\psi \phi$ suructuces. Wivh joimt gxperimental and theoretical efrort, hopevuljy we will oftaik a slear granb picture and eventually understand the iature of these queer srructures—which may nnceed lead ux to w new king of spectroscopy.
Ackbowledgments {#acknpwledgments.unnumbered}
===============
J woulg likf to thank Gerry Bauer, Stanley Brodsky, Joel Butler, Tiziano Camporesi, Kai-fwng Chxn, Vincenzo Cgiocnia, Rdy Eul | isospin-0 vectors could possibly be connected to J/\psi\omega$ $Y(4140)\rightarrow J/\psi since both $J/\psi$ are close to the vector threshold. Second, is striking that there are at two structures observed in the same $J/\psi\phi$ spectrum from the same exclusive $B$ The two structures do not need to arise from the same mechanism, but they this be pointer to a new kind of spectroscopy. It will also be interesting to search for vector-vector composed entirely of c and b quarks near because they may offer systems to model theoretically. Summary The of the in initiated new chapter in study of exotic states. The evidence found in the $J/\psi\phi$ mass spectrum from exclusive $B$ decays and process provide charmonium candidates: all a well beyond the threshold, a relatively narrow width, positive not fit into the conventional charmonium picture. All these reported need further studies. Various possible interpretations as molecule, tetra-quark, charmed hybrid, nuclei-like structures have proposed but none of them is established. The $pp$ run at the LHC is now 2015, but CMS has four times more on to analyzed LHCb is to have about ten times the data to be analyzed. ATLAS, the other hand, has not yet been heard from but also to the current The properties of these such mass, width, decay channels numbers, be statistics. the one can also search other heavy quarkonium decay channels vector-vector analogs of the experimental and theoretical effort, hopefully we will obtain clear grand picture and eventually understand the of these queer structures—which may indeed lead us to a new kind spectroscopy. Acknowledgments I would like to thank Gerry Bauer, Stanley Joel Butler, Tiziano Camporesi, Chen, Vincenzo Chiochia, Ray Cul | isospin-0 vectors could possibLy be connecTed to The $y(3940)\riGhTarrOw J/\pSi\omega$ and $Y(4140)\rigHTarrOw J/\psi \phi$ since both $J/\psi$ And $\phI$ hAVe isOSpIn-0 and Are very CLoSE To tHe TwO veCtOR tHreshOld. second, iT is strikinG thAt There are at leASt Two structuRes Observed in thE saMe $J/\psi\PhI$ spECtrum FroM the sAme excLUsive $B$ Decays. The TwO StructURes do noT NEeD to aRise from the same meCHaNIsm, but if they arE, this cOuLD bE A PoiNteR to a new kinD oF specTRoscopy. iT wILL AlsO Be interesting To search for VEctOr-vectOr StrUCtures CompoSeD EntIrely of c and B quaRks near thResholD Because THey may oFfer siMplEr sYsteMS tO mOdeL tHEorETiCalLY.
SuMmary
=======
The DiScOvery Of thE $x(3872)$ IN 2003 InitIatEd a nEw chaPter in the studY of ExotIC stAtes. THe eviDencE fOund iN the $J/\pSi\phi$ MaSs spectrum from eXcluSive $B$ decaYs aNd Two-PhOton pROcess pRovIde Several Exotic cHArmOnIUM CaNdidates: they all havE a MASs Well beyoNd the oPEn-ChARm pair thReShoLd, a rELAtiveLy naRRoW width, poSitive c-PaRiTy, and do NoT fit inTo The ConVentiONal cHarmonIum pictuRe. All OF these reported STructures need FUrTHEr STudiEs. VArious possiBle iNTerpRetaTIoNs sUCh as mOlecuLe, TEtRA-quark, charmed hybrid, NuClei-liKe strUctures have beEn proposed BUT None of thEm is EStABlished. The $pp$ ruN at thE LHC is now oVEr until 2015, bUt CMS Has roughLy four timES More data On tApe To bE anALYzEd and LHCb is exPECted To Have aboUt tEn times The DatA to Be aNaLyzed. ATLAs, on the otHeR hAnD, hAs nOt yet BEen heard FrOm bUt TheY can aLSo contRibutE to tHe CuRRenT studieS. thE PRopeRtIeS of tHesE sTructUres, SUch As mass, wIdth, decay ChaNNels AnD qUantum nUmbers, can be stUdIed with incReAseD statiSTIcs. At the lHC one can also search in otHEr heavy QuaRkoniUm deCay channeLs fOr vectOr-vECtor anAlogs oF the $J/\PsI \phI$ STructUREs. witH jOint experiMENtaL and tHeOretIcal effOrt, hopefully we will OBtaIn a clear grand PicTure AND eVenTUaLLy uNdERstAND the nature of theSe queer strUcTUrEs—which may INdeEd Lead us tO a new kiNd of sPEctroscOpy.
AcknowLedgments {#AcKnowLEDgmEnts.unnumbEred}
===============
I wouLd like to tHAnk GeRRy bauer, staNley BrOdSky, joel BUtler, TIZiaNo CamPoresi, kaI-feng CHen, ViNcEnzo ChioChia, Ray Cul | isospin-0 vectors could p ossibly be conn ect edto the $Y( 3940)\rightarr o w J/ \psi\omega$ and $Y(414 0)\ri gh t arro w J /\psi \phi$s in c e bo th $ J/\ ps i $and $ \ph i$ have isospin-0 an dare very clo s eto the two ve ctor thresho ld. Secon d, it is st rik ing t hat th e re are at least t w o stru c tures o b s er vedin the same $J/\p s i\ p hi$ spectrum f rom th es am e exc lus ive $B$ de ca ys. T h e two s t ru c t u res do not need t o arise fro m th e same m ech a nism,but i ft hey are, thiscoul d be a po intert o a new kind of spect ros cop y. I t w il l a ls o be in ter e sti ng to se ar ch forvect o r - v ecto r s truc tures composed ent ire ly o f cand b quar ks n ea r thr eshold beca us e they may offe r si mpler sys tem stomo del t h eoreti cal ly.
Summa ry
==== = ==
T h e di scovery of the $X( 38 7 2 )$ in 2003 initi a te da new cha pt erin t h e stud y of ex otic sta tes. T h eev idencefo und in t he$J/ \psi\ p hi$mass s pectrumfrome xclusive $B$ d e cays and two- p ho t o np roce ssprovide sev eral exot ic c h ar mon i um ca ndida te s :t hey all have a mass w ell be yondthe open-char m pair thr e s h old, a r elat i ve l y narrow width , pos itive C-pa r ity, and do n ot fit i nto the c o n ventiona l c har mon ium p ic ture. All oft h esere portedstr uctures ne edfur the rstudies.Variouspo ss ib le in terpr e tationssu chas mo lecul e , tetr a-qua rk,ch ar m edhybrid, nu c l ei-l ik estru ctu re s hav e be e n p roposed but none of them i sestabli shed. The $pp $run at the L HCis now o ver unti l 2015, but CMS has rou g hly fou r t imesmore data ontap e to b e a n alyzed and L HCb i sexp e c ted t o ha veab out ten ti m e s t he da ta tobe anal yzed. ATLAS, on th e ot her hand, has no t ye t be enh ea r d f ro m bu t they can also c ontributeto th e currents tud ie s. Thepropert ies o f thesestructure s, such a smass , wid th, decaychannels and quan t um nu m be rs, c anbe stu di edwithincrea s edstati stics. A t theLHC o ne can als o search in other heavy quark onium de cay chann els for vector-v ecto r analogsofthe $J/\ psi \phi$ str u ct ure s . Wit h jo i nt experi m en tal a nd theoretica l e ffo rt, h ope f ully w e wi ll obtain a clear grand pictureande v ent ual l y un de rstand the nat ure o f these qu ee r structure s—whichma y inde ed lea d us t o a new k in d of sp ectr osc opy.
Ack now le d gments{# ac k nowled gmen ts .unnum bered} ==== = = =========
I wou ld li k e to t h ank Gerr yBauer,S tanl ey Brodsky , Joel Butl er, Ti zian o Cam poresi, K ai-fen g C he n, Vincenz o Chiochia , Ray Cul | isospin-0_vectors could_possibly be connected to_the $Y(3940)\rightarrow_J/\psi\omega$_and $Y(4140)\rightarrow_J/\psi_\phi$ since both_$J/\psi$ and $\phi$_have isospin-0 and are_very close to_the_two vector threshold. Second, it is striking that there are at least two structures_observed_in the_same_$J/\psi\phi$_spectrum from the same exclusive_$B$ decays. The two structures_do not_need to arise from the same mechanism, but_if_they are, this_could be a pointer to a new kind of_spectroscopy. It will also be interesting_to search for_vector-vector_structures_composed entirely of c_and b quarks near threshold because_they may offer simpler systems to_model theoretically.
Summary
=======
The discovery of the $X(3872)$ in_2003 initiated a new chapter in_the study of exotic states._The evidence_found in the $J/\psi\phi$ mass_spectrum from exclusive_$B$ decays_and two-photon process_provide several exotic charmonium candidates: they_all have a_mass well beyond the open-charm pair_threshold,_a relatively narrow_width,_positive_C-parity, and_do not fit_into_the conventional_charmonium_picture. All of these reported structures_need_further studies. Various possible interpretations such as_molecule, tetra-quark, charmed hybrid,_nuclei-like_structures have been proposed_but none of them is_established. The $pp$ run at the_LHC is_now over_until 2015, but CMS has roughly four times more data on_tape to be analyzed and LHCb_is expected to have_about ten_times_the data to_be_analyzed. ATLAS,_on the other hand, has not yet_been heard_from but they can also contribute_to the current studies._The_properties of these structures, such as_mass, width, decay channels and quantum_numbers, can be studied with_increased_statistics._At the LHC one can_also search in other heavy quarkonium_decay channels for_vector-vector analogs of the $J/\psi \phi$ structures._With_joint experimental and theoretical effort, hopefully_we_will obtain a clear grand picture_and_eventually_understand the nature of these_queer structures—which may indeed lead us_to a new kind of spectroscopy.
Acknowledgments {#acknowledgments.unnumbered}
===============
I would like_to thank Gerry_Bauer, Stanley Brodsky, Joel Butler,_Tiziano_Camporesi,_Kai-feng Chen, Vincenzo Chiochia, Ray Cul |
Note that, if we take the alphabet $[k]$ to be a set of colours, the definition of a picture is analogous to that of a coloured pattern [@ref22].
Early generating/accepting systems for 2D languages comprise $2 \times 2$ tiles [@g31], 2D automata [@ref109], two-dimensional on-line tessellation acceptors [@tesselation], and 2D grammars. More recently a generating system was introduced by Varricchio [@wangtileref] that used [*Wang tiles*]{}. A [*Wang tile system*]{} [@wangtileref] is a specialized tile-based model that generates the class of [*recognizable picture languages*]{}, a subclass of the family of 2D languages. The class of recognizable picture languages is also accepted by [*Wang automata*]{}, a model introduced in [@ref105]. Like other automata for 2D languages [@g06], Wang tile automata use an explicit pre-defined scanning strategy [@ref111] when reading the input picture and the accepted language depends on the scanning strategy that is used. Due to this, Wang automata are a suboptimal model for self-assembly. Indeed, if we consider the final supertile as given, the order in which tiles are read is irrelevant. On the other hand, if we consider the self-assembly process which results in the final supertile, an “order of assembly” cannot be pre-imposed. In contrast to Wang automata, SA-hypergraph automata are scanning-strategy-independent.
SA-hypergraph automata are a modification of the hypergraph automata introduced by Rozenberg [@h001] in 1982. An SA-hypergraph automaton (Section \[sec:hypergraph\]) accepts a language of labelled “rectangular grid graphs”, wherein the labels are meant to capture the notion of colours used in patterned self-assembly. An SA-hypergraph automaton consists of an underlying labelled graph (labelled nodes and edges) and a set of [*hyperedges*]{}, each of which is a subset of the set of nodes of the underlying graph. Intuitively, the hyperedges are meant to model tiles or supertiles while the underlying graph describes how these can attach to each other, similar to a self-assembly process.
We investigate the computational power of SA-hypergraph automata and prove that for every recognizable picture language $L$ there is an SA-hypergraph automaton that accepts $L$ (Thm. | Note that, if we take the alphabet $ [ k]$ to be a set of coloring material, the definition of a movie is analogous to that of a coloured pattern [ @ref22 ].
Early render / accepting systems for 2D linguistic process incorporate $ 2 \times 2 $ tiles [ @g31 ], 2D automata [ @ref109 ], two - dimensional on - line tessellation acceptors [ @tesselation ], and 2D grammar. More recently a generating system was introduced by Varricchio [ @wangtileref ] that used [ * Wang tiles * ] { }. A [ * Wang tile arrangement * ] { } [ @wangtileref ] is a specialized tile - based model that generates the class of [ * recognizable picture language * ] { }, a subclass of the family of 2D languages. The class of recognizable video languages is also accepted by [ * Wang automata * ] { }, a model introduce in [ @ref105 ]. Like other automata for 2D languages [ @g06 ], Wang tile automata use an denotative pre - defined scanning strategy [ @ref111 ] when reading the input picture and the accepted language depend on the scanning strategy that is used. Due to this, Wang automata are a suboptimal model for self - assembly. Indeed, if we consider the final supertile as given, the order in which tiles are take is irrelevant. On the other hand, if we think the self - forum process which results in the concluding supertile, an “ order of fabrication ” cannot be pre - imposed. In contrast to Wang automata, SA - hypergraph automaton are scanning - strategy - independent.
SA - hypergraph automata are a modification of the hypergraph automata insert by Rozenberg [ @h001 ] in 1982. An SA - hypergraph automaton (Section \[sec: hypergraph\ ]) accepts a language of labelled “ rectangular grid graph ”, wherein the label are meant to capture the notion of colours practice in patterned self - assembly. An SA - hypergraph automaton consists of an underlying labelled graph (labelled nodes and edges) and a bent of [ * hyperedges * ] { }, each of which is a subset of the set of nodes of the underlie graph. Intuitively, the hyperedges are meant to exemplar tiles or supertiles while the underlie graph describes how these can impound to each other, similar to a self - assembly process.
We investigate the computational power of SA - hypergraph automaton and prove that for every recognizable picture language $ L$ there is an SA - hypergraph automaton that accepts $ L$ (Thm. | Nohe that, if we take the auphabet $[k]$ to be a set mf colkurs, the definition of a picture is enaligous to that of a coloured pattern [@gef22].
Early tenecating/accepting systems njr 2D pangbajes comprise $2 \tlmes 2$ tiles [@g31], 2D automata [@sew109], cwo-dimensional on-line tessellation asceptorx [@hesselation], anq 2D bwammzgs. More recently a generating sgstem wes introduced bu Varricchio [@wangtileref] tjat ksed [*Wang tiles*]{}. A [*Aang tile siatei*]{} [@wangtileref] is a specpclized tile-gased model that generates the zlass of [*recognuzqblf picture laiguagef*]{}, a subclass of the xamily pf 2D languages. Thx clqss of recognizable pmcture languages is wlso acce[tzd by [*Wang automata*]{}, a midel hntrmducde iv [@rtf105]. Kiie othfr eutomata fod 2D languagws [@g06], Wang tile autokaeq use an expljcit pwe-qefined scanning strategy [@ref111] when readpng fhe input picture and tye accepted language fepends og the scanning strategy that is used. Due to this, Fang eugomcbw qrf a suboptimal model for self-assembly. Indeed, yr ee consider the ninal supertile as glvrg, the order iv whick tjles are read is igrelevagt. On the othew hamd, if we consider the self-awsembly procvss qhich results in tke final supzrtile, an “otder of assembly” cannot be lre-imposed. Ln contraag to Wang automaga, XA-vypergrakf automata are scwnning-stretegy-nndependdnt.
SS-hyperdraph autolata are a modification lf thg hypesgraph autlmata introduced by Rozenberg [@h001] in 1982. An SA-hypetgrdph automatjn (Segtion \[sec:hypergrwph\]) accepts a kanguagz of lxbelled “rebtangular grid graphf”, wherein the pabels are mxant to cwptuee tye notiuv of colours uxed in panttrned self-qssembly. An SA-hypevgrapf automaton consnrus of an underlyong lafeplxd grw[h (labelled toder avc edgds) and a seb ow [*hyleredges*]{}, each of whiwh ia a subset of the xeb of nodew of the underlying gtaph. Intuitively, tje hy'eredgxs are mewnt to model tiles or supertilss while hhe underlying drapm defcribes hoc these can attach to each other, similar to a self-assembli process.
We investigaje uhe computatipnal [ower of VA-hypergraph automatq and prove that nor every recognizable pictuse lajguage $L$ there is an SA-hypergraph automaton that accepts $L$ (Thm. | Note that, if we take the alphabet be set of the definition of that a coloured pattern Early generating/accepting systems 2D languages comprise $2 \times 2$ [@g31], 2D automata [@ref109], two-dimensional on-line tessellation acceptors [@tesselation], and 2D grammars. More a generating system was introduced by Varricchio [@wangtileref] that used [*Wang tiles*]{}. A tile [@wangtileref] a tile-based model that generates the class of [*recognizable picture languages*]{}, a subclass of the family of languages. The class of recognizable picture languages is accepted by [*Wang automata*]{}, model introduced in [@ref105]. Like automata 2D languages Wang automata an explicit pre-defined strategy [@ref111] when reading the input picture and the accepted language depends on the scanning strategy that used. Due Wang automata a model self-assembly. Indeed, if the final supertile as given, the tiles are read is irrelevant. On the other if we the self-assembly process which results in final supertile, an “order of assembly” cannot be In contrast to Wang automata, SA-hypergraph automata are scanning-strategy-independent. SA-hypergraph automata are a modification of automata introduced by Rozenberg in 1982. An automaton \[sec:hypergraph\]) a of labelled grid graphs”, wherein the labels are meant to capture the notion colours used in patterned self-assembly. An SA-hypergraph automaton consists of labelled (labelled nodes and and a set of each which is a subset set nodes graph. the are meant to model or supertiles while the underlying describes how these can to a self-assembly process. We investigate the computational of SA-hypergraph automata and prove that for recognizable picture language $L$ there is an SA-hypergraph automaton that accepts $L$ | Note that, if we take the alphabEt $[k]$ to be a seT of coLouRs, tHe DefiNitiOn of a picture is ANaloGous to that of a coloured pAtterN [@rEF22].
EarLY gEneraTing/accEPtING syStEmS foR 2D LAnGuageS coMprise $2 \tImes 2$ tiles [@g31], 2d auToMata [@ref109], two-diMEnSional on-liNe tEssellation aCcePtors [@tEsSelATion], aNd 2D GrammArs. MorE RecentLy a generaTiNG systeM Was intrODUcEd by varricchio [@wangtilEReF] That used [*Wang tiLes*]{}. A [*WaNg TIlE SYstEm*]{} [@wAngtileref] Is A specIAlized tILe-BASEd mODel that generaTes the class OF [*reCognizAbLe pICture lAnguaGeS*]{}, A suBclass of the FamiLy of 2D langUages. THE class oF RecogniZable pIctUre LangUAgEs Is aLsO AccEPtEd bY [*wanG automatA*]{}, a MoDel inTrodUCED In [@reF105]. LiKe otHer auTomata for 2D lanGuaGes [@g06], wAng Tile aUtomaTa usE aN explIcit prE-defiNeD scanning strateGy [@reF111] when readIng ThE inPuT pictURe and tHe aCcePted lanGuage dePEndS oN THE sCanning strategy thaT iS USeD. Due to thIs, Wang AUtOmATa are a suBoPtiMal mODEl for Self-ASsEmbly. IndEed, if wE CoNsIder the FiNal supErTilE as Given, THe orDer in wHich tileS are rEAd is irrelevant. oN the other hand, IF wE COnSIder The Self-assemblY proCEss wHich REsUltS In the Final SuPErTIle, an “order of assemblY” cAnnot bE pre-iMposed. In contrAst to Wang aUTOMata, SA-hyPergRApH Automata are scaNning-Strategy-inDEpendent.
sA-hypErgraph aUtomata arE A ModificaTioN of The HypERGrAph automata inTROducEd By RozenBerG [@h001] in 1982. An Sa-hyPerGraPh aUtOmaton (SecTion \[sec:hYpErGrApH\]) acCepts A Language Of LabElLed “RectaNGular gRid grAphs”, WhErEIn tHe labelS ArE MEant To CaPturE thE nOtion Of coLOurS used in Patterned SelF-AsseMbLy. an SA-hypErgraph automaToN consists oF aN unDerlyiNG Labelled Graph (labelled nodes and edGEs) and a sEt oF [*hypeRedgEs*]{}, each of wHicH is a suBseT Of the sEt of noDes of ThE unDERlyinG GRaPh. INtUitively, thE HYpeRedgeS aRe meAnt to moDel tiles or supertilES whIle the underlyIng GrapH DEsCriBEs HOw tHeSE caN ATtach to each otheR, similar to A sELf-Assembly prOCesS.
WE investIgate thE compUTationaL power of Sa-hypergraPh AutoMATa aNd prove thaT for everY recognizABle piCTuRe lanGuaGe $L$ theRe Is aN SA-hyPergraPH auTomatOn that AcCepts $L$ (thm. | Note that, if we take the alphabet$[k]$ to be a set ofcolours, the d e fini tion of a picture is a nalog ou s tot ha t ofa colou r ed p att er n[@r ef 2 2] .
Ea rly genera ting/accep tin gsystems for2 Dlanguagescom prise $2 \ti mes 2$ ti le s [ @ g31], 2D auto mata [ @ ref109 ], two-di me n sional on-line t es sell ation acceptors [ @ te s selation], and 2D gr am m ar s . Mo rerecently a g enera t ing sys t em w a s i n troduced by V arricchio [ @ wan gtiler ef ] t h at use d [*W an g ti les*]{}. A[*Wa ng tile s ystem* ] {} [@wa n gtilere f] isa s pec iali z ed t ile -b a sed mo del tha t genera te sthe c lass o f [*re cog niza ble p icture langua ges *]{} , asubcl ass o f th efamil y of 2 D lan gu ages. The class ofrecogniza ble p ict ur e lan g uagesisals o accep ted by[ *Wa ng a u to mata*]{}, a modelin t r od uced in[@ref1 0 5] .L ike othe raut omat a for 2 D la n gu ages [@g 06], W a ng t ile aut om ata us eanexp licit pre- define d scanni ng st r ategy [@ref111 ] when reading th e in p ut p ict ure and the acc e pted lan g ua ged epend s onth e s c anning strategy tha tis use d. Du e to this, Wa ng automat a a re a sub opti m al model for self -asse mbly. Inde e d, if we cons ider the final su p e rtile as gi ven , t heo r de r in which ti l e s ar eread is ir relevan t.Onthe ot he r hand, i f we con si de rth e s elf-a s sembly p ro ces swhi ch re s ults i n the fin al s u per tile, a n “ o r derof a ssem bly ”canno t be pre -impose d. In con tra s t to W an g autom ata, SA-hyper gr aph automa ta ar e scan n i ng-strat egy-independent.
SA-hy p ergraph au tomat a ar e a modif ica tion o f t h e hype rgraph auto ma tai n trodu c e dbyRo zenberg [@ h 0 01] in 1 98 2. A n SA-hy pergraph automaton (Se ction \[sec:h ype rgra p h \] ) a c ce p tsal ang u a ge of labelled“rectangul ar gr id graphs” , wh er ein the labels arem eant to capturethe notio nof c o l our s used inpatterne d self-as s embly . A n SA- hyp ergrap haut omato n cons i sts of a n unde rl ying l abell ed graph ( labelled nodes and edge s) and a se t o f [*hyper edg e s*] {}, eachof w hich is asub set of t hes et of nod e soft he un derl y ing graph . I ntu i t iv ely, the hy p e r edg es ar e m e ant to mod el tiles or super t iles while the und e r lyi ngg raph d escribes how t hes ec a n attach t o each othe r, simil ar to aself-a ssembl y proce s s .We inv esti gat e the com put at i onal po we ro f SA-h yper gr aph au tomata andp r ove that for eve ry re c o gniza b lepictu re langua g e $L $ there is an SA-hype rgraph aut omato n thatac cepts$L$ ( Thm. | Note_that, if_we take the alphabet_$[k]$ to_be_a set_of_colours, the definition_of a picture_is analogous to that_of a coloured_pattern_[@ref22].
Early generating/accepting systems for 2D languages comprise $2 \times 2$ tiles [@g31], 2D automata_[@ref109],_two-dimensional on-line_tessellation_acceptors_[@tesselation], and 2D grammars. More_recently a generating system was_introduced by_Varricchio [@wangtileref] that used [*Wang tiles*]{}. A [*Wang_tile_system*]{} [@wangtileref] is_a specialized tile-based model that generates the class of_[*recognizable picture languages*]{}, a subclass of_the family of_2D_languages._The class of recognizable_picture languages is also accepted by_[*Wang automata*]{}, a model introduced in_[@ref105]. Like other automata for 2D languages_[@g06], Wang tile automata use an_explicit pre-defined scanning strategy [@ref111]_when reading_the input picture and the_accepted language depends_on the_scanning strategy that_is used. Due to this, Wang_automata are a_suboptimal model for self-assembly. Indeed, if_we_consider the final_supertile_as_given, the_order in which_tiles_are read_is_irrelevant. On the other hand, if_we_consider the self-assembly process which results in_the final supertile, an_“order_of assembly” cannot be_pre-imposed. In contrast to Wang_automata, SA-hypergraph automata are scanning-strategy-independent.
SA-hypergraph automata_are a_modification of_the hypergraph automata introduced by Rozenberg [@h001] in 1982. An SA-hypergraph_automaton (Section \[sec:hypergraph\]) accepts a language of_labelled “rectangular grid graphs”,_wherein the_labels_are meant to_capture_the notion_of colours used in patterned self-assembly. An_SA-hypergraph automaton_consists of an underlying labelled graph_(labelled nodes and edges)_and_a set of [*hyperedges*]{}, each of_which is a subset of the_set of nodes of the_underlying_graph._Intuitively, the hyperedges are meant_to model tiles or supertiles while_the underlying graph_describes how these can attach to each_other,_similar to a self-assembly process.
We investigate_the_computational power of SA-hypergraph automata and_prove_that_for every recognizable picture language_$L$ there is an SA-hypergraph automaton_that accepts $L$ (Thm. |
adiabatic stage, the outflowing mass creates the over-pressurized bubble of hot gas. The mass of the bubble $M_{\textrm{b}}$ is determined by the conservation of mass law: $$\label{mass_e}
\frac{{\textrm{d}}M_{\textrm{b}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} + 4\pi R^2 \rho_{\textrm{o}} \left( v_{\textrm{s}} - v_{\textrm{o}}\right),$$ where the first term on the right hand side represents the mass of cold gas reheated by SN explosions and stellar winds and the second term defines the amount of hot gas swept up by the wind and accreted during the expansion in the halo. In a snowplough, most of the mass accumulates in a thin shell of mass $m$: $$\label{mass_m}
\frac{{\textrm{d}}m}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} \left( 1- \frac{v_{\textrm{s}}}{v_{\textrm{w}}\left( R \right)} \right) + 4\pi R^2 \rho_{\textrm{o}} \left( v_{\textrm{s}} - v_{\textrm{o}}\right).$$ A fraction $\dot{M}_{\textrm{w}} v_{\textrm{s}} / v_{\textrm{w}}\left( R
\right)$ of the outflowing ISM does not reach the shell and remains in the cavity behind the shock. Assuming that the amount of metals exchanged is proportional to the corresponding amount of mass (i.e. neglecting any metal-loading effect), the corresponding equations for metals are: $$\label{mass_z}
\frac{{\textrm{d}}M_{\textrm{bz}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} Z_{\textrm{cold}} + 4\pi R^2 \rho_{\textrm{o}} \left( v_{\textrm{s}} - v_{\textrm{o}}\right) Z_{\textrm{hot}},$$ where $Z_{\textrm{cold}}$ is the metallicity of the cold gas and $Z_{\textrm{hot}}$ the metallicity of the hot gas, and: $$\label{mass_zshell}
\frac{{\textrm{d}}m_{\textrm{z}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} \left( 1- \frac{v | adiabatic stage, the outflowing mass creates the all over - pressurize bubble of hot accelerator. The multitude of the bubble $ M_{\textrm{b}}$ is determined by the conservation of batch law: $ $ \label{mass_e }
\frac{{\textrm{d}}M_{\textrm{b}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w } } + 4\pi R^2 \rho_{\textrm{o } } \left (v_{\textrm{s } } - v_{\textrm{o}}\right),$$ where the first condition on the right hand english represents the mass of cold flatulence reheat by SN explosions and leading winds and the second condition specify the amount of hot gas embroil up by the wind and accreted during the expansion in the halo. In a snowplough, most of the batch accumulates in a thin shell of mass $ m$: $ $ \label{mass_m }
\frac{{\textrm{d}}m}{{\textrm{d}}t}=\dot{M}_{\textrm{w } } \left (1- \frac{v_{\textrm{s}}}{v_{\textrm{w}}\left (R \right) } \right) + 4\pi R^2 \rho_{\textrm{o } } \left (v_{\textrm{s } } - v_{\textrm{o}}\right).$$ A fraction $ \dot{M}_{\textrm{w } } v_{\textrm{s } } / v_{\textrm{w}}\left (R
\right)$ of the outflowing ISM does not strive the shell and remains in the cavity behind the shock. Assuming that the amount of metals exchanged is proportional to the corresponding amount of mass (i.e. neglecting any metal - loading effect), the corresponding equations for metallic element are: $ $ \label{mass_z }
\frac{{\textrm{d}}M_{\textrm{bz}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w } } Z_{\textrm{cold } } + 4\pi R^2 \rho_{\textrm{o } } \left (v_{\textrm{s } } - v_{\textrm{o}}\right) Z_{\textrm{hot}},$$ where $ Z_{\textrm{cold}}$ is the metallicity of the cold gas and $ Z_{\textrm{hot}}$ the metallicity of the hot natural gas, and: $ $ \label{mass_zshell }
\frac{{\textrm{d}}m_{\textrm{z}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w } } \left (1- \frac{v | adlabatic stage, the outfloding mass creatgs the oter-presaurized cubble of hot gas. The mass oh thw bubvle $M_{\textrm{b}}$ is determkned by tje conseevatmon of mass law: $$\label{mass_e}
\frac{{\fcxtrm{b}}M_{\vextrm{b}}}{{\textrm{d}}t}=\cot{M}_{\textrm{f}} + 4\pi R^2 \rho_{\texdro{o}} \left( v_{\textrm{s}} - v_{\textrm{o}}\right),$$ where the fitsh term on the tight rand side represents the mass of cold fas rehtated by SN explosoons and stellar winds and the second term definfs the amoubt os hot gas sweot up by the wind and zccreted during the expansion iv the halo. In a sbowonough, most if thv mass accumulates in a thin shell of mass $m$: $$\nabwl{mass_m}
\frac{{\textrm{d}}m}{{\teetrm{d}}t}=\dot{M}_{\textrm{w}} \lest( 1- \frac{v_{\dertrm{s}}}{v_{\textrm{w}}\left( R \rugyt)} \richt) + 4\pi F^2 \rhu_{\testcm{o}} \left( g_{\teetrm{s}} - v_{\texfrm{o}}\right).$$ A fraction $\dot{M}_{\textrk{w}} n_{\yextrm{s}} / v_{\testrm{w}}\lqfe( R
\right)$ of the outflowing ISM does not rezch the shell and remaibs in the cavity behijd the shjck. Assuming that the amount of metals exchanged hs prkooruiinal gi hhe corresponding amount of mass (i.e. neglectind amy metal-loading cffect), the corresppnfimd equations fur metcms are: $$\label{mass_z}
\fraf{{\textrm{q}}M_{\texrrm{bz}}}{{\textwm{d}}t}=\cot{M}_{\textrm{w}} Z_{\textrm{cold}} + 4\pi R^2 \rho_{\textrm{j}} \left( v_{\textrm{s}} - v_{\tzxtrm{o}}\right) V_{\textrk{hot}},$$ ehere $Z_{\textrm{cold}}$ is thz metamlicity of hhe cold fxs and $Z_{\textrm{hog}}$ tme ketallicity of the hot gas, and: $$\labeo{masx_zshell}
\wrac{{\jextrm{d}}i_{\textrm{z}}}{{\tedtrm{d}}b}=\got{M}_{\textrm{w}} \left( 1- \vrac{v | adiabatic stage, the outflowing mass creates the of gas. The of the bubble conservation mass law: $$\label{mass_e} + 4\pi R^2 \left( v_{\textrm{s}} - v_{\textrm{o}}\right),$$ where the term on the right hand side represents the mass of cold gas reheated SN explosions and stellar winds and the second term defines the amount of gas up the and accreted during the expansion in the halo. In a snowplough, most of the mass accumulates a thin shell of mass $m$: $$\label{mass_m} \frac{{\textrm{d}}m}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} 1- \frac{v_{\textrm{s}}}{v_{\textrm{w}}\left( R \right)} + 4\pi R^2 \rho_{\textrm{o}} \left( - A fraction v_{\textrm{s}} v_{\textrm{w}}\left( \right)$ of the ISM does not reach the shell and remains in the cavity behind the shock. Assuming that the of metals proportional to corresponding of (i.e. neglecting any the corresponding equations for metals are: + 4\pi R^2 \rho_{\textrm{o}} \left( v_{\textrm{s}} - v_{\textrm{o}}\right) where $Z_{\textrm{cold}}$ the metallicity of the cold gas $Z_{\textrm{hot}}$ the metallicity of the hot gas, and: \frac{{\textrm{d}}m_{\textrm{z}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} \left( 1- \frac{v | adiabatic stage, the outflowiNg mass creaTes thE ovEr-pReSsurIzed Bubble of hot gas. tHe maSs of the bubble $M_{\textrm{b}}$ iS deteRmINed bY ThE consErvatioN Of MASs lAw: $$\LaBel{MaSS_e}
\Frac{{\tExtRm{d}}M_{\texTrm{b}}}{{\textrm{D}}t}=\dOt{m}_{\textrm{w}} + 4\pi R^2 \rHO_{\tExtrm{o}} \left( V_{\teXtrm{s}} - v_{\textrm{O}}\riGht),$$ wheRe The FIrst tErm On the Right hANd side RepresentS tHE mass oF Cold gas REHeAted By SN explosions and STeLLar winds and the Second TeRM dEFIneS thE amount of hOt Gas swEPt up by tHE wIND And ACcreted during The expansioN In tHe halo. in A snOWplougH, most Of THe mAss accumulaTes iN a thin sheLl of maSS $m$: $$\label{MAss_m}
\fraC{{\textrM{d}}m}{{\TexTrm{d}}T}=\DoT{M}_{\TexTrM{W}} \leFT( 1- \fRac{V_{\TexTrm{s}}}{v_{\texTrM{w}}\Left( R \RighT)} \RIGHt) + 4\pi r^2 \rhO_{\texTrm{o}} \lEft( v_{\textrm{s}} - v_{\tExtRm{o}}\rIGht).$$ a fracTion $\dOt{M}_{\tExTrm{w}} v_{\Textrm{S}} / v_{\texTrM{w}}\left( R
\right)$ of tHe ouTflowing IsM dOeS noT rEach tHE shell And RemAins in tHe cavitY BehInD THE sHock. Assuming that thE aMOUnT of metalS exchaNGeD iS ProportiOnAl tO the CORrespOndiNG aMount of mAss (i.e. nEGlEcTing any MeTal-loaDiNg eFfeCt), the COrreSpondiNg equatiOns foR Metals are: $$\label{MAss_z}
\frac{{\textrM{D}}M_{\TEXtRM{bz}}}{{\tExtRm{d}}t}=\dot{M}_{\texTrm{w}} z_{\TextRm{coLD}} + 4\pI R^2 \rHO_{\textRm{o}} \leFt( V_{\TeXTrm{s}} - v_{\textrm{o}}\right) Z_{\tExTrm{hot}},$$ Where $z_{\textrm{cold}}$ is The metalliCITY of the coLd gaS AnD $z_{\textrm{hot}}$ the mEtallIcity of the HOt gas, and: $$\Label{Mass_zsheLl}
\frac{{\texTRM{d}}m_{\textrM{z}}}{{\tExtRm{d}}T}=\doT{m}_{\TeXtrm{w}} \left( 1- \frac{V | adiabatic stage, the outf lowing mas s cre ate s t he ove r-pr essurized bubb l e of hot gas. The mass ofthe b ub b le $ M _{ \text rm{b}}$ is d ete rm in edby th e con ser vationof mass la w:$$ \label{mass_ e }\frac{{\te xtr m{d}}M_{\tex trm {b}}}{ {\ tex t rm{d} }t} =\dot {M}_{\ t extrm{ w}} + 4\p iR ^2 \rh o _{\text r m {o }} \ left( v_{\textrm{ s }} - v_{\textrm{o }}\rig ht ) ,$ $ whe rethe firstte rm on the rig h th a n d s i de represents the mass o f co ld gas r ehe a ted by SN e xp l osi ons and ste llar winds an d thes econd t e rm defi nes th e a mou nt o f h ot ga ss wep t u p b y th e wind a nd a ccret ed d u r i n g th e e xpan sionin the halo.Ina sn o wpl ough, most ofth e mas s accu mulat es in a thin shel l of mass $m$ : $ $\ lab el {mass _ m}
\fr ac{ {\t extrm{d }}m}{{\ t ext rm { d } }t }=\dot{M}_{\textrm {w } } \ left( 1- \frac { v_ {\ t extrm{s} }} {v_ {\te x t rm{w} }\le f t( R \righ t)} \r i gh t) + 4\pi R ^2 \rh o_ {\t ext rm{o} } \le ft( v_ {\textrm {s}}- v_{\textrm{o} } \right).$$ Af ra c t io n $\d ot{ M}_{\textrm {w}} v_{\ text r m{ s}} / v_{ \text rm { w} } \left( R
\right)$ o fthe ou tflow ing ISM doesnot reacht h e shell a nd r e ma i ns in the cavi ty be hind the s h ock. Ass uming that th e amounto f metalsexc han ged is p ro portional tot h e co rr espondi ngamountofmas s ( i.e .neglectin g any me ta l- lo ad ing effe c t), theco rre sp ond ing e q uation s for met al sa re: $$\lab e l{ m a ss_z }\f rac{ {\t ex trm{d }}M_ { \te xtrm{bz }}}{{\tex trm { d}}t }= \d ot{M}_{ \textrm{w}} Z _{ \textrm{co ld }}+ 4\pi R ^2 \rho_ {\textrm{o}} \left( v_{ \ textrm{ s}} - v_ {\te xtrm{o}}\ rig ht) Z_ {\t e xtrm{h ot}},$ $ whe re $Z _ { \text r m {c old }} $ is the m e t all icity o f th e coldgas and $Z_{\textr m {ho t}}$ the meta lli city o fthe ho t ga s, and : $$\label{mass_z shell}
\fr ac { {\ textrm{d}} m _{\ te xtrm{z} }}{{\te xtrm{ d }}t}=\d ot{M}_{\t extrm{w}} \ left ( 1-\frac{v | adiabatic_stage, the_outflowing mass creates the_over-pressurized bubble_of_hot gas._The_mass of the_bubble $M_{\textrm{b}}$ is_determined by the conservation_of mass law:_$$\label{mass_e}
\frac{{\textrm{d}}M_{\textrm{b}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}}_+ 4\pi R^2 \rho_{\textrm{o}} \left( v_{\textrm{s}} - v_{\textrm{o}}\right),$$ where the first term on the_right_hand side_represents_the_mass of cold gas reheated_by SN explosions and stellar_winds and_the second term defines the amount of hot_gas_swept up by_the wind and accreted during the expansion in the_halo. In a snowplough, most of_the mass accumulates_in_a_thin shell of mass_$m$: $$\label{mass_m}
\frac{{\textrm{d}}m}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} \left( 1- \frac{v_{\textrm{s}}}{v_{\textrm{w}}\left( R_\right)} \right) + 4\pi R^2 \rho_{\textrm{o}}_\left( v_{\textrm{s}} - v_{\textrm{o}}\right).$$ A fraction $\dot{M}_{\textrm{w}}_v_{\textrm{s}} / v_{\textrm{w}}\left( R
\right)$ of the_outflowing ISM does not reach_the shell_and remains in the cavity_behind the shock._Assuming that_the amount of_metals exchanged is proportional to the_corresponding amount of_mass (i.e. neglecting any metal-loading effect),_the_corresponding equations for_metals_are:_$$\label{mass_z}
\frac{{\textrm{d}}M_{\textrm{bz}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} Z_{\textrm{cold}}_+ 4\pi R^2_\rho_{\textrm{o}}_\left( v_{\textrm{s}}_-_v_{\textrm{o}}\right) Z_{\textrm{hot}},$$ where $Z_{\textrm{cold}}$ is the_metallicity_of the cold gas and $Z_{\textrm{hot}}$ the_metallicity of the hot_gas,_and: $$\label{mass_zshell}
\frac{{\textrm{d}}m_{\textrm{z}}}{{\textrm{d}}t}=\dot{M}_{\textrm{w}} \left( 1-_\frac{v |
situations.
Acknowledgements {#acknowledgements.unnumbered}
----------------
Thanks to Ralph Kaufmann for bringing our attention to the references [@Bun] and [@Day]. We thank Paul Balmer for pointing us to [@BKS] and subsequent discussions of the case of homological functors, see Section \[sec:kuenneth\]. We thank the referee for her/his suggestions that considerably improved the exposition.
Notation {#notation.unnumbered}
--------
By a tensor category $(C, \otimes)$ we mean a category $C$ provided with a functor $\otimes: C\times C\to C$ satisfying an associativity constraint and with ${\mathbf{1}}$ a unit object; in addition, also a commutativity constraint can be required, e.g. see [@DMT §1]. This is often called a non-strict tensor category. By an additive (resp. abelian) tensor category we mean a tensor category $(C, \otimes)$ such that $C$ is additive (resp. abelian) and ${\otimes}$ is a bi-additive functor, see [@DMT Def. 1.15]. Tensor functors are not assumed strict. Tensor functors between additive tensor categories are assumed to be additive. We denote by ${\mathbb{Q}\mathrm{-vsp.}}$ the tensor category of ${\mathbb{Q}}$-vector spaces.
If ${\mathcal{A}}$ is an abelian category, we denote by ${\mathrm{gr}}{\mathcal{A}}$ the associated category of ${\mathbb{Z}}$-graded objects. If, in addition, $({\mathcal{A}}, \otimes)$ carries a tensor structure, we equip $({\mathrm{gr}}{\mathcal{A}}, \otimes )$ with the induced tensor structure. If the tensor product is commutative, we choose the commutativity constraint on ${\mathrm{gr}}{\mathcal{A}}$ such that the product becomes graded anti-commutative.
For an additive category $C$ we shall consider the additive functors from $C$ to the category Ab of abelian groups as (left) $C$-modules. We shall denote by $C{\mathrm{-mod}}$ the category of finitely presented $C$-modules, see e.g. [@P Chaps. 2, 3].
Universal abelian tensor categories {#sec:universal_tensor}
===================================
Let $C$ be an additive category. We denote by ${\mathrm{Ab}(C)}$ the universal abelian category on $C$, see [@F | situations.
Acknowledgements { # acknowledgements.unnumbered }
----------------
Thanks to Ralph Kaufmann for bringing our attention to the reference point [ @Bun ] and [ @Day ]. We thank Paul Balmer for target us to [ @BKS ] and subsequent discussions of the case of homological functors, witness Section \[sec: kuenneth\ ]. We thank the referee for her / his suggestions that well improved the exposition.
Notation { # notation.unnumbered }
--------
By a tensor category $ (coke, \otimes)$ we mean a category $ C$ put up with a functor $ \otimes: C\times C\to C$ satisfying an associativity constraint and with $ { \mathbf{1}}$ a whole aim; in addition, also a commutativity restraint can be required, for example learn [ @DMT § 1 ]. This is often called a non - strict tensor class. By an additive (resp. abelian) tensor category we mean a tensor category $ (C, \otimes)$ such that $ C$ is linear (resp. abelian) and $ { \otimes}$ is a bi - additive functor, learn [ @DMT Def. 1.15 ]. Tensor functors are not assumed strict. Tensor functors between additive tensor class are assumed to be additive. We denote by $ { \mathbb{Q}\mathrm{-vsp.}}$ the tensor category of $ { \mathbb{Q}}$-vector spaces.
If $ { \mathcal{A}}$ is an abelian category, we denote by $ { \mathrm{gr}}{\mathcal{A}}$ the associated category of $ { \mathbb{Z}}$-graded objects. If, in addition, $ ({ \mathcal{A } }, \otimes)$ carries a tensor structure, we equip $ ({ \mathrm{gr}}{\mathcal{A } }, \otimes) $ with the induced tensor structure. If the tensor product is commutative, we choose the commutativity constraint on $ { \mathrm{gr}}{\mathcal{A}}$ such that the product becomes rate anti - commutative.
For an additive category $ C$ we shall think the linear functors from $ C$ to the category Ab of abelian groups as (left) $ C$ -modules. We shall denote by $ C{\mathrm{-mod}}$ the category of finitely presented $ C$ -modules, see e.g. [ @P Chaps. 2, 3 ].
cosmopolitan abelian tensor categories { # sec: universal_tensor }
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Let $ C$ be an additive category. We denote by $ { \mathrm{Ab}(C)}$ the universal abelian category on $ C$, see [ @F | sihuations.
Acknowledgements {#acknowledgemenjs.ynnumbxred}
----------------
Thahks to Rxlph Kaufmann for bringing onr artentuon to the references [@Cun] and [@Dwy]. We thqnk Kaul Balmer for pointing mf to [@NKS] aud subsequent dixcussions mf the case of humllogical functors, see Section \[sec:kuegneth\]. Wr hhank the refetee fpw hed/his suggestions that considerably improvtd the exposition.
Nptation {#notation.unnumbered}
--------
Hy a tensor category $(C, \otimes)$ we nean q category $C$ provided with a functkr $\otimes: C\times C\to C$ satisfyivg an associaticiry fmnstraint aid witr ${\mathbf{1}}$ a ukpt objewt; in acdition, also a cokmurativity constraint cen be required, e.g. seg [@DMT §1]. Thiv ns often called a non-wteict jensos cagwgofy. Gy ah addihivx (resp. abeljan) tensor xategory we mean a uenfir category $(C, \otimef)$ fuch that $C$ is additive (resp. abelian) ang ${\ofimes}$ is a bi-additive fynctor, see [@DMT Def. 1.15]. Tgnsor funceors are not assumed strict. Tensor functors betwean advigivt bcnsof cwtegories are assumed to be additive. We denotq bu ${\kathbb{Q}\mathrm{-vfp.}}$ the tenspr csjegory of ${\mathcb{Q}}$-vecckr spaces.
If ${\mathcal{A}}$ is an wbeliqn categowy, wr denote by ${\mathrm{gr}}{\mathcal{Q}}$ the associctee category of ${\mathyb{Z}}$-graded oblects. Of, in addition, $({\mathcal{A}}, \otimzs)$ cardies a tenslr structhfe, we equip $({\mathfm{gg}}{\matvcal{A}}, \otimes )$ with the indtced tensir scructure. If jhe tenfor produch is gmmmutative, we choode thg commgtativity fonstraint on ${\mathrm{gr}}{\mathcal{A}}$ such that the krogucn becomes gradcd anti-commutatyve.
For an addijive cateyory $C$ we shall bonsider vhe additive functors frok $C$ to the cavegory Ab of qbeluan grojos as (left) $C$-mocules. We shall denore by $C{\mathrm{-mod}}$ tme cajefory of finitelv krwsented $C$-modulrs, ree e.h. [@P Crdps. 2, 3].
Universan abdlixm tenror cattyorles {#sec:iniversal_tensor}
===================================
Let $C$ be zn additive categoty. We denoje by ${\matrrm{Ab}(C)}$ the unoversal abelian cauegory on $C$, ser [@F | situations. Acknowledgements {#acknowledgements.unnumbered} ---------------- Thanks to Ralph bringing attention to references [@Bun] and for us to [@BKS] subsequent discussions of case of homological functors, see Section We thank the referee for her/his suggestions that considerably improved the exposition. Notation -------- By a tensor category $(C, \otimes)$ we mean a category $C$ provided a $\otimes: C\to satisfying an associativity constraint and with ${\mathbf{1}}$ a unit object; in addition, also a commutativity constraint be required, e.g. see [@DMT §1]. This is called a non-strict tensor By an additive (resp. abelian) category mean a category \otimes)$ that $C$ is (resp. abelian) and ${\otimes}$ is a bi-additive functor, see [@DMT Def. 1.15]. Tensor functors are not assumed Tensor functors tensor categories assumed be We denote by tensor category of ${\mathbb{Q}}$-vector spaces. If abelian category, we denote by ${\mathrm{gr}}{\mathcal{A}}$ the associated of ${\mathbb{Z}}$-graded If, in addition, $({\mathcal{A}}, \otimes)$ carries tensor structure, we equip $({\mathrm{gr}}{\mathcal{A}}, \otimes )$ with induced tensor structure. If the tensor product is commutative, we choose the commutativity constraint on that the product becomes anti-commutative. For an category we consider additive functors $C$ to the category Ab of abelian groups as (left) $C$-modules. shall denote by $C{\mathrm{-mod}}$ the category of finitely presented $C$-modules, [@P 2, 3]. Universal tensor categories {#sec:universal_tensor} =================================== $C$ an additive category. We ${\mathrm{Ab}(C)}$ universal $C$, [@F | situations.
AcknowledgementS {#acknowledGemenTs.uNnuMbEred}
----------------
thanKs to Ralph KaufmANn foR bringing our attention tO the rEfERencES [@BUn] and [@day]. We thANk pAUl BAlMeR foR pOInTing uS to [@bKS] and sUbsequent dIscUsSions of the caSE oF homologicAl fUnctors, see SeCtiOn \[sec:kUeNneTH\]. We thAnk The reFeree fOR her/hiS suggestiOnS That coNSiderabLY ImProvEd the exposition.
NoTAtIOn {#notation.unnuMbered}
--------
by A TeNSOr cAteGory $(C, \otimeS)$ wE mean A CategorY $c$ pROVIdeD With a functor $\oTimes: C\times c\To C$ SatisfYiNg aN AssociAtiviTy COnsTraint and wiTh ${\maThbf{1}}$ a unit Object; IN additiON, also a cOmmutaTivIty ConsTRaInT caN bE ReqUIrEd, e.G. See [@dMT §1]. This iS oFtEn calLed a NON-STricT teNsor CategOry. By an additiVe (rEsp. aBEliAn) tenSor caTegoRy We meaN a tensOr catEgOry $(C, \otimes)$ such tHat $C$ Is additivE (reSp. AbeLiAn) and ${\OTimes}$ iS a bI-adDitive fUnctor, sEE [@DMt DEF. 1.15]. tEnSor functors are not aSsUMEd Strict. TeNsor fuNCtOrS Between aDdItiVe teNSOr catEgorIEs Are assumEd to be ADdItIve. We deNoTe by ${\maThBb{Q}\MatHrm{-vsP.}}$ The tEnsor cAtegory oF ${\mathBB{Q}}$-vector spaces.
iF ${\mathcal{A}}$ is an ABeLIAn CAtegOry, We denote by ${\mAthrM{Gr}}{\maThcaL{a}}$ tHe aSSociaTed caTeGOrY Of ${\mathbb{Z}}$-graded objeCtS. If, in aDditiOn, $({\mathcal{A}}, \otiMes)$ carries A TENsor struCturE, We EQuip $({\mathrm{gr}}{\maThcal{a}}, \otimes )$ witH The inducEd tenSor strucTure. If the TENsor prodUct Is cOmmUtaTIVe, We choose the coMMUtatIvIty consTraInt on ${\maThrM{gr}}{\MatHcaL{A}}$ Such that tHe producT bEcOmEs GraDed anTI-commutaTiVe.
FOr An aDditiVE categOry $C$ wE shaLl CoNSidEr the adDItIVE funCtOrS froM $C$ tO tHe catEgorY ab oF abeliaN groups as (LefT) $c$-modUlEs. we shall Denote by $C{\mathRm{-Mod}}$ the cateGoRy oF finitELY presentEd $C$-modules, see e.g. [@P Chaps. 2, 3].
UnIVersal aBelIan teNsor CategorieS {#seC:univeRsaL_Tensor}
===================================
let $C$ be An addItIve CATegorY. wE dEnoTe By ${\mathrm{Ab}(c)}$ THe uNiverSaL abeLian catEgory on $C$, see [@F | situations.
Acknowledgem ents {#ack nowle dge men ts .unn umbe red}
--------- - ---- --
Thanks to Ralph Ka ufman nf or b r in gingour att e nt i o n t oth e r ef e re nces[@B un] and [@Day]. W e t ha nk Paul Balm e rfor pointi ngus to [@BKS] an d subs eq uen t disc uss ionsof the case o f homolog ic a l func t ors, se e Se ctio n \[sec:kuenneth\ ] .W e thank the re fereefo r h e r /hi s s uggestions t hat c o nsidera b ly i m pro v ed the exposi tion.
Nota t ion {#not at ion . unnumb ered}
- - --- ---
By a t enso r categor y $(C, \otimes ) $ we me an a c ate gor y $C $ p ro vid ed wit h a fu n cto r $\otim es :C\tim es C \ t o C$ s ati sfyi ng an associativit y c onst r ain t and with ${\ ma thbf{ 1}}$ a unit o bject; in addit ion, also a c omm ut ati vi ty co n strain t c anbe requ ired, e . g.se e [ @D MT §1]. This is of te n ca lled a n on-str i ct t e nsor cat eg ory . By a n add itiv e ( resp. ab elian) te ns or cate go ry weme ana t ensor cate gory $ (C, \oti mes)$ such that $C$i s additive (r e sp . ab e lian ) a nd ${\otime s}$i s abi-a d di tiv e func tor,se e [ @ DMT Def. 1.15]. Ten so r func torsare not assum ed strict. T e nsor fun ctor s b e tween additive tens or categor i es are a ssume d to beadditive. W e denote by ${ \ma thb b { Q} \mathrm{-vsp. } } $ th etensorcat egory o f $ {\m ath bb{ Q} }$-vector spaces.
If $ {\ mat hcal{ A }}$ is a nabe li ancateg o ry, we deno te b y${ \ mat hrm{gr} } {\ m a thca l{ A} }$ t heas socia tedc ate gory of ${\mathb b{Z } }$-g ra de d objec ts. If, in ad di tion, $({\ ma thc al{A}} , \otimes) $ carries a tensor stru c ture, w e e quip$({\ mathrm{gr }}{ \mathc al{ A }}, \o times)$ wi th th e induc e d t ens or structure . Ifthe t en sorproduct is commutative, w e ch oose the comm uta tivi t y c ons t ra i nton ${\ m a thrm{gr}}{\math cal{A}}$ s uc h t hat the pr o duc tbecomes graded anti - commuta tive.
Fo r an addi ti ve c a t ego ry $C$ weshall co nsider th e addi t iv e fun cto rs fro m$C$ to t he cat e gor y Abof abe li an gro ups a s(left) $ C$-modules. We shall de note b y $C{ \ma thrm{-mod }}$ the category offinitely p res ent ed $C $-m o dules , se e e .g. [@P C haps . 2, 3].
U ni ver s a labelian ten s o r ca tegor ies {#sec: univ ersal_tensor}
=== = ============== ==== = = === === = ====
Let $C$ be anadd it i v e catego ry . We denote by ${\m at h rm{Ab }(C)}$ the u niversa l ab e lian c ateg ory on $C$,see [ @ F | situations.
Acknowledgements_{#acknowledgements.unnumbered}
----------------
Thanks to_Ralph Kaufmann for bringing_our attention_to_the references_[@Bun]_and [@Day]. We_thank Paul Balmer_for pointing us to_[@BKS] and subsequent_discussions_of the case of homological functors, see Section \[sec:kuenneth\]. We thank the referee for her/his_suggestions_that considerably_improved_the_exposition.
Notation {#notation.unnumbered}
--------
By a tensor category_$(C, \otimes)$ we mean a_category $C$_provided with a functor $\otimes: C\times C\to C$_satisfying_an associativity constraint_and with ${\mathbf{1}}$ a unit object; in addition, also_a commutativity constraint can be required,_e.g. see [@DMT_§1]._This_is often called a_non-strict tensor category. By an additive_(resp. abelian) tensor category we mean_a tensor category $(C, \otimes)$ such that_$C$ is additive (resp. abelian) and_${\otimes}$ is a bi-additive functor,_see [@DMT_Def. 1.15]. Tensor functors are_not assumed strict._Tensor functors_between additive tensor_categories are assumed to be additive._We denote by_${\mathbb{Q}\mathrm{-vsp.}}$ the tensor category of ${\mathbb{Q}}$-vector_spaces.
If_${\mathcal{A}}$ is an_abelian_category,_we denote_by ${\mathrm{gr}}{\mathcal{A}}$ the_associated_category of_${\mathbb{Z}}$-graded_objects. If, in addition, $({\mathcal{A}}, \otimes)$_carries_a tensor structure, we equip $({\mathrm{gr}}{\mathcal{A}}, \otimes_)$ with the induced_tensor_structure. If the tensor_product is commutative, we choose_the commutativity constraint on ${\mathrm{gr}}{\mathcal{A}}$ such_that the_product becomes_graded anti-commutative.
For an additive category $C$ we shall consider the additive_functors from $C$ to the category_Ab of abelian groups_as (left)_$C$-modules._We shall denote_by_$C{\mathrm{-mod}}$ the_category of finitely presented $C$-modules, see e.g. [@P_Chaps. 2, 3].
Universal_abelian tensor categories {#sec:universal_tensor}
===================================
Let $C$ be_an additive category. We_denote_by ${\mathrm{Ab}(C)}$ the universal abelian category_on $C$, see [@F |
functions on a convex domain. In this case one can use the expectation-maximization (EM) algorithm [@Vardi] as discussed in [@Zhang]. A problem is that due to stopping criteria and numerical precision, one cannot expect to find the exact optimum. We show in App. \[sect:bad estimation\] that one can compensate for this problem to maintain validity of the computed $p$-value.
Conclusion {#sect:conclusion}
==========
The degree of violation of LR in a Bell-type test is usually expressed in terms of the number of SDs of violation. This quantity cannot, however, be used to obtain valid $p$-values for rejecting LR by conventional means. It also fails to quantitatively compare the success of different experimental tests of LR and does not account for stability issues or memory effects in experiments. We solve these problems by providing a method—the PBR protocol—for determining valid $p$-values directly from the settings and outcomes in a sequence of trials. The PBR protocol does not rely on a predetermined Bell inequality, adapts to the actual experimental configuration, and is asymptotically optimal for independent and identically distributed trials. It therefore provides a standardized measure of success for experimental tests of LR. While the protocol remains valid if the experiment drifts over the sequence of trials, how well it performs depends on the nature of the drifts and how the protocol takes them into account. Another valid protocol that accounts for memory effects can be based on martingale bounds [@Gill1; @Gill2]. This protocol requires a Bell inequality that is fixed for the experiment. Given the Bell inequality, the martingale-based protocol has the advantage that it is computationally efficient with respect to number of settings, outcomes, and parties. The disadvantage is that it is suboptimal and does not provide a clear quantitative comparison of different experimental tests. Our simulations show that it is practical to apply the PBR protocol to data from typical experimental configurations, and that the running $p$-values can be used for tweaking an experiment in progress to find the experimentally accessible configuration that provides the highest violation of LR.
We thank D. N. Matsukevich for providing the experimental data for Ref. [@Monroe], and we are grateful for the editorial support of A. Migdall and K. Coakley. This paper is a contribution of the National Institute of Standards and Technology and is not subject to U.S. copyright.
{#sect:appendix}
Statistical Concept | functions on a convex domain. In this case one can use the anticipation - maximization (EM) algorithm [ @Vardi ] as discourse in [ @Zhang ]. A problem is that due to stopping criteria and numeral precision, one cannot expect to discover the exact optimum. We show in App. \[sect: regretful estimation\ ] that one can pay for this problem to sustain validity of the computed $ p$-value.
Conclusion { # sect: conclusion }
= = = = = = = = = =
The degree of violation of LR in a Bell - character test is usually expressed in price of the number of SDs of misdemeanor. This quantity cannot, however, be use to obtain valid $ p$-values for rejecting LR by conventional means. It also fail to quantitatively compare the success of different experimental tests of LR and does not account for stability consequence or memory effects in experiments. We solve these problems by providing a method — the PBR protocol — for determining valid $ p$-values directly from the settings and outcomes in a succession of test. The PBR protocol does not rely on a preset Bell inequality, adapts to the actual experimental shape, and is asymptotically optimal for independent and identically distributed trials. It consequently provides a exchangeable measure of success for experimental tests of LR. While the protocol remain valid if the experiment drifts over the sequence of trials, how well it performs depends on the nature of the drifts and how the protocol takes them into account. Another valid protocol that accounts for memory effects can be based on dolphin striker bounds [ @Gill1; @Gill2 ]. This protocol want a Bell inequality that is fix for the experiment. Given the Bell inequality, the martingale - based protocol have the advantage that it is computationally efficient with respect to number of mise en scene, outcomes, and parties. The disadvantage is that it is suboptimal and does not provide a clear quantitative comparison of unlike experimental tests. Our simulations show that it is practical to apply the PBR protocol to datum from typical experimental configurations, and that the running $ p$-values can be used for tweak an experiment in progress to receive the experimentally accessible configuration that provides the highest violation of LR.
We thank D. N. Matsukevich for providing the experimental datum for Ref. [ @Monroe ], and we are grateful for the editorial support of A. Migdall and K. Coakley. This paper is a contribution of the National Institute of Standards and Technology and is not subject to U.S. copyright.
{ # faction: appendix }
Statistical Concept | fujctions on a convex domaln. In this case one can use tge expecgation-maximization (EM) algorivhm [@Vqrdi] qs discussed in [@Zhang]. A problem ps that dye ti stopping rditeria and nhlerieao precision, ong cannot expact to find tha dxcct optimum. We show in App. \[sect:bad eseimatiom\] hhat one can cjmpemfate for this problem to maintain valisity of the computed $p$-value.
Conclusion {#sect:concpusiln}
==========
The degree of villation of OR ig a Bell-type gest is uslclly expresaed in terms of the number of SAs of violation. Tyis xuantity cainot, hjwever, be uscc to ottain vslid $p$-values fpr cejexting LR by conventioial means. It also fayls to qudncitatively compare thw wuccevs ox diwderdnt eepedimentwl vests of LR and does nit account for stabolynu issues or jemory esfects in experiments. We solve these prmblsms by providing a methid—the PBR protocol—for determinyng valid $p$-values directly from the settings and mutcojds nk x sfquence of trials. The PBR protocol does not rqmy ok a predetermineb Bell inequalitu, wdskts to the actoal expzdijental configuratiln, and ys astmptoticajly pptimal for independent and identically eistributed trials. It therefoxe profides a standardized measure of auccess for experimehgal tests of LR. Dhike the prouucol remains valiq if the xxpernment drkfts over ehe sequenfe of trials, how well it perflrks depends on the nature of the drifts anv how the projocml nakes thei intp account. Anoeher valid projocol thac accojnts for mvmory effxcts can be fased on marthjgale bounds [@Jill1; @Gill2]. Thiw pritocol fdquires a Bell inequaliny thar is fixed for the exkediment. Given thz Vell inequality, thd mwrninjale-bwved protocol has thd advavtage that lt ks cpmputationally effichent with respect to nimner of sejtings, ouecomes, and patties. The disadvanuage iv tiat it is suboptimal and does not provise a cleag qmantitative cjmpavisog of diffexent experimental tests. Our simulations whow that it is ptactical to apply the KBR protocol vo datw from ty[ical experimental cinfigurations, and that the running $p$-vamues cdn be used for tweaking an experiment in progress to find the experimentally axcessiulq configurafion that pxovndes thq hijhxst violation of KR.
We thank D. N. Matsukevich for 'roviding dhz experimental data for Ref. [@Momrue], and we are grateful for the editorjal suppprt of A. Migdall and K. Coakley. Yhis paper is a contribubion of tye Nafional Insvitute of Standards snd Teehnoloty and us nlt subject to U.S. clpyright.
{#dect:eppzndix}
Statistical Concept | functions on a convex domain. In this can the expectation-maximization algorithm [@Vardi] as is due to stopping and numerical precision, cannot expect to find the exact We show in App. \[sect:bad estimation\] that one can compensate for this problem maintain validity of the computed $p$-value. Conclusion {#sect:conclusion} ========== The degree of violation LR a test usually expressed in terms of the number of SDs of violation. This quantity cannot, however, be to obtain valid $p$-values for rejecting LR by means. It also fails quantitatively compare the success of experimental of LR does account stability issues or effects in experiments. We solve these problems by providing a method—the PBR protocol—for determining valid $p$-values directly the settings in a of The protocol does not a predetermined Bell inequality, adapts to configuration, and is asymptotically optimal for independent and distributed trials. therefore provides a standardized measure of for experimental tests of LR. While the protocol valid if the experiment drifts over the sequence of trials, how well it performs depends nature of the drifts how the protocol them account. valid that accounts memory effects can be based on martingale bounds [@Gill1; @Gill2]. This requires a Bell inequality that is fixed for the experiment. Bell the martingale-based protocol the advantage that it computationally with respect to number outcomes, parties. that is and does not provide clear quantitative comparison of different tests. Our simulations show apply the PBR protocol to data from typical configurations, and that the running $p$-values can used for tweaking an experiment in progress to find the experimentally accessible that provides violation of LR. We thank D. N. Matsukevich providing the experimental data Ref. [@Monroe], and we are grateful for the editorial of Migdall and Coakley. This paper a contribution of National Institute of Technology and not to Statistical Concept | functions on a convex domain. IN this case oNe can Use The ExPectAtioN-maximization (Em) AlgoRithm [@Vardi] as discussed iN [@ZhanG]. A PRoblEM iS that Due to stOPpING crItErIa aNd NUmEricaL prEcision, One cannot eXpeCt To find the exaCT oPtimum. We shOw iN App. \[sect:bad eStiMation\] ThAt oNE can cOmpEnsatE for thIS problEm to maintAiN ValidiTY of the cOMPuTed $p$-Value.
Conclusion {#seCT:cONclusion}
==========
The degRee of vIoLAtION of lR iN a Bell-type TeSt is uSUally exPReSSED in TErms of the numbEr of SDs of viOLatIon. ThiS qUanTIty canNot, hoWeVEr, bE used to obtaIn vaLid $p$-valueS for reJEcting Lr By conveNtionaL meAns. it alSO fAiLs tO qUAntITaTivELy cOmpare thE sUcCess oF difFERENt exPerImenTal teSts of LR and doeS noT accOUnt For stAbiliTy isSuEs or mEmory eFfectS iN experiments. We sOlve These probLemS bY prOvIding A Method—The pBR ProtocoL—for detERmiNiNG VAlId $p$-values directly fRoM THe Settings And outCOmEs IN a sequenCe Of tRialS. tHe PBR ProtOCoL does not Rely on A PrEdEterminEd bell inEqUalIty, AdaptS To thE actuaL experimEntal COnfiguration, anD Is asymptoticaLLy OPTiMAl foR inDependent anD ideNTicaLly dIStRibUTed trIals. IT tHErEFore provides a standaRdIzed meAsure Of success for eXperimentaL TESts of LR. WHile THe PRotocol remains Valid If the experIMent drifTs oveR the sequEnce of triALS, how well It pErfOrmS dePENdS on the nature oF THe drIfTs and hoW thE protocOl tAkeS thEm iNtO account. ANother vaLiD pRoToCol That aCCounts foR mEmoRy EffEcts cAN be basEd on mArtiNgAlE BouNds [@Gill1; @gIlL2]. tHis pRoToCol rEquIrEs a BeLl inEQuaLity thaT is fixed fOr tHE expErImEnt. GiveN the Bell inequAlIty, the martInGalE-based PROtocol haS the advantage that it is coMPutatioNalLy effIcieNt with resPecT to numBer OF settiNgs, outComes, AnD paRTIes. ThE DIsAdvAnTage is that IT Is sUboptImAl anD does noT provide a clear quanTItaTive comparisoN of DiffEREnT exPErIMenTaL TesTS. our simulations sHow that it iS pRAcTical to appLY thE PbR protoCol to daTa froM Typical ExperimenTal configUrAtioNS, And That the runNing $p$-valUes can be uSEd for TWeAking An eXperimEnT in ProgrEss to fINd tHe expErimenTaLly accEssibLe ConfigurAtion that provides the higHest viOlatiOn oF LR.
We thanK D. N. mAtsUkevich foR proViding the eXpeRimEntal DatA For ReF. [@MonROe], And WE are gRateFUl for the eDItOriAL SuPport of A. MigDALL anD K. CoaKleY. this paPer iS a contribution of tHE National InstiTute OF staNdaRDs anD TEchnology and is Not SuBJEct to U.S. cOpYright.
{#sect:aPpendix}
STaTIsticAl ConcEpt | functions on a convex dom ain. In th is ca seone c an u se t he expectation - maxi mization (EM) algorith m [@V ar d i] a s d iscus sed in[ @Z h a ng] .Apro bl e mis th atdue tostopping c rit er ia and numer i ca l precisio n,one cannot e xpe ct tofi ndt he ex act opti mum. W e showin App. \ [s e ct:bad estimat i o n\ ] th at one can compen s at e for this prob lem to m a in t a inval idity of t he comp u ted $p$ - va l u e .
C onclusion {#s ect:conclus i on}
===== == ===
The d egree o f vi olation ofLR i n a Bell- type t e st is u s ually e xpress edinterm s o fthe n u mbe r o f S D s o f violat io n. This qua n t i t y ca nno t, h oweve r, be used to ob tain val id $p $-val uesfo r rej ecting LR b yconventional me ans. It alsofai ls to q uanti t ativel y c omp are the succes s of d i f f er ent experimental t es t s o f LR and doesn ot a c count fo rsta bili t y issu es o r m emory ef fectsi nex perimen ts . We s ol vethe se pr o blem s by p roviding a me t hod—the PBR pr o tocol—for det e rm i n in g val id$p$-valuesdire c tlyfrom th e s e tting s and o u tc o mes in a sequence o ftrials . The PBR protocol does notr e l y on a p rede t er m ined Bell ineq ualit y, adaptst o the ac tualexperime ntal conf i g uration, an d i s a sym p t ot ically optima l forin depende ntand ide nti cal lydis tr ibuted tr ials. It t he re fo reprovi d es a sta nd ard iz edmeasu r e of s ucces s fo rex p eri mentalt es t s ofLR .Whil e t he prot ocol rem ains va lid if th e e x peri me nt drifts over the seq ue nce of tri al s,how we l l it perf orms depends on the nat u re of t hedrift s an d how the pr otocol ta k es the m into acco un t.A n other v al idpr otocol tha t acc ounts f or m emory e ffects can be base d on martingale b oun ds [ @ G il l1; @G i ll2 ]. Thi s protocol requir es a Bellin e qu ality that isfi xed for the ex perim e nt. Giv en the Be ll inequa li ty,t h e m artingale- based pr otocol ha s thea dv antag e t hat it i s c omput ationa l lyeffic ient w it h resp ect t onumber o f settings, outcomes, a nd par ties. Th e disadva nta g e i s that it issuboptimal an d d oes n otp rovid e ac le arq uanti tati v e compari s on of d if ferent expe r i m ent al te sts . Our s imul ations show thati t is practical toa p ply th e PBR p rotocol to dat a f ro m typicalex perimentalconfigur at i ons,and th at the runnin g $p $ -value s ca n b e used fo r t we a king an e xp e riment inpr ogress to fi n d th e experimentally a ccess i b le co n fig urati on that p r ovid es the hig hest violat ion of LR.
Wethank D .N. Mat suk ev ich for pr o viding th e exp eriment al dat a f or Ref . [@ M o nroe] , an dweare grate f u lf or t h e e dito rialsu ppor t of A. M i gdall an d K . Coakle y. Th i s paper is a contribut ion of t h e NationalI nsti t ut e of S tandar ds and Techno l ogy a nd is n ots u bject toU.S. copy r igh t.
{ #sect:ap pe ndix }
St at istical Concept | functions_on a_convex domain. In this_case one_can_use the_expectation-maximization_(EM) algorithm [@Vardi] as_discussed in [@Zhang]. A_problem is that due_to stopping criteria_and_numerical precision, one cannot expect to find the exact optimum. We show in App. \[sect:bad_estimation\]_that one_can_compensate_for this problem to maintain_validity of the computed $p$-value.
Conclusion_{#sect:conclusion}
==========
The degree_of violation of LR in a Bell-type test_is_usually expressed in_terms of the number of SDs of violation. This_quantity cannot, however, be used to_obtain valid $p$-values_for_rejecting_LR by conventional means._It also fails to quantitatively compare_the success of different experimental tests_of LR and does not account for_stability issues or memory effects in_experiments. We solve these problems_by providing_a method—the PBR protocol—for determining_valid $p$-values directly_from the_settings and outcomes_in a sequence of trials. The_PBR protocol does_not rely on a predetermined Bell_inequality,_adapts to the_actual_experimental_configuration, and_is asymptotically optimal_for_independent and_identically_distributed trials. It therefore provides a_standardized_measure of success for experimental tests of_LR. While the protocol_remains_valid if the experiment_drifts over the sequence of_trials, how well it performs depends_on the_nature of_the drifts and how the protocol takes them into account. Another_valid protocol that accounts for memory_effects can be based_on martingale_bounds [@Gill1;_@Gill2]. This protocol_requires_a Bell_inequality that is fixed for the experiment._Given the_Bell inequality, the martingale-based protocol has_the advantage that it_is_computationally efficient with respect to number_of settings, outcomes, and parties. The_disadvantage is that it is_suboptimal_and_does not provide a clear_quantitative comparison of different experimental tests._Our simulations show_that it is practical to apply the_PBR_protocol to data from typical experimental_configurations,_and that the running $p$-values can_be_used_for tweaking an experiment in_progress to find the experimentally accessible_configuration that provides the highest violation of LR.
We thank_D. N. Matsukevich_for providing the experimental data_for_Ref. [@Monroe],_and we are grateful for the editorial support of A._Migdall and_K. Coakley. This_paper is a contribution of the National Institute of Standards_and Technology and is not subject to_U.S. copyright.
{#sect:appendix}
Statistical Concept |
].
Besides this strong $\Delta J = 2$ propensity rule, one can see from table \[tab:rates2\] and figures \[fig:ratecompare\], \[fig:rate12\] that the rod-like interaction drives large $\Delta J$ transfers. For instance, for T $> 20$ K, rates for $\Delta J > 6$ are generally larger than rates for $\Delta J = 1$, and rates for $\Delta J > 8$ are only one order of magnitude below those for $\Delta J = 2$. This behaviour is likely to emphasize the role of collisional effects versus radiative ones. This effect, of purely geometric origin, has been predicted previously [@bosanac80] and is of even greater importance for longer rods like $\rm HC_5N$, $\rm HC_7N$, $\rm HC_9N$, see @snell81 [@dickinson82].
We also observe that the ratio $k_{J_1J'_1}(\textrm{He})/k_{J_1J'_1}(\mbox{para-H$_2$}) $ is in average close to $1/1.4 \sim 1/\sqrt{2}$, thus confirming the similarity of He and para-[$\rm H_2$]{}as projectiles, as generally assumed. But it is also far from being a constant, as already observed for H$_2$O [@phillips96] or CO [@wernli06]. Our data shows that the $1/\sqrt{2}$ scaling rule results in errors up to 50%.
Population inversion and critical densities
-------------------------------------------
Because of the strong $\Delta J_1=0,2,4$ propensity rule, population inversion could be strengthened if LTE conditions are not met, even neglecting hyperfine effects[^1] [@hunt99]. In order to see the density conditions giving rise to population inversion, we solved the steady-state equations for the population of the $J=0,1,\dots,15$ levels of [$\mathrm{HC_3N}$]{}, including collisions with [$\rm H_2$]{}(densities ranging from $10^2$ to $10^6 \rm\; cm^{-3}$), a black-body photon bath at 2.7 K, in the optically thin approximation, [@goldsmith72] : $$\begin{aligned}
\label{eq:ss}
\frac{\textmd{d}n_i | ].
Besides this strong $ \Delta J = 2 $ propensity rule, one can see from mesa \[tab: rates2\ ] and visualize \[fig: ratecompare\ ], \[fig: rate12\ ] that the rod - like interaction drives big $ \Delta J$ transfers. For instance, for T $ > 20 $ K, rate for $ \Delta J > 6 $ are broadly larger than rate for $ \Delta J = 1 $, and rates for $ \Delta J > 8 $ are only one club of magnitude below those for $ \Delta J = 2$. This behaviour is probable to emphasize the role of collisional effects versus radiative ones. This effect, of purely geometric origin, has been predict previously [ @bosanac80 ] and is of even greater importance for longer rods like $ \rm HC_5N$, $ \rm HC_7N$, $ \rm HC_9N$, learn @snell81 [ @dickinson82 ].
We also observe that the ratio $ k_{J_1J'_1}(\textrm{He})/k_{J_1J'_1}(\mbox{para - H$_2 $ }) $ is in average near to $ 1/1.4 \sim 1/\sqrt{2}$, thus confirm the similarity of He and para-[$\rm H_2$]{}as projectiles, as generally assumed. But it is besides far from being a constant, as already observed for H$_2$O [ @phillips96 ] or CO [ @wernli06 ]. Our data shows that the $ 1/\sqrt{2}$ scaling rule results in error up to 50% .
Population inversion and critical densities
-------------------------------------------
Because of the strong $ \Delta J_1=0,2,4 $ propensity rule, population inversion could be strengthened if LTE conditions are not met, even neglecting hyperfine effects[^1 ] [ @hunt99 ]. In order to attend the density conditions feed upgrade to population inversion, we solved the steady - department of state equations for the population of the $ J=0,1,\dots,15 $ levels of [ $ \mathrm{HC_3N}$ ] { }, including collisions with [ $ \rm H_2$]{}(densities range from $ 10 ^ 2 $ to $ 10 ^ 6 \rm\; cm^{-3}$), a black - body photon bath at 2.7 K, in the optically thin approximation, [ @goldsmith72 ] : $ $ \begin{aligned }
\label{eq: ss }
\frac{\textmd{d}n_i | ].
Besldes this strong $\Delta J = 2$ propensity rolw, one ran see from tacle \[tab:rates2\] and figures \[fig:retecimpart\], \[fig:rate12\] that the fod-like ijteractiin dcives large $\Delte J$ transfers. Rlr iusvance, for T $> 20$ K, tates for $\Denta J > 6$ are getefaply larger than rates for $\Delta J = 1$, and rayed for $\Delta J > 8$ art ogly kne order of magnitude below those for $\Denta J = 2$. This nehaviour is likely to empjasixe the role of colpisional efdectf versus radixtive ones. This effect, of purely geometric origin, has been predicted peevlmusly [@bosanec80] and is of even ngeater hmportamce for longer rovs luke $\rm HC_5N$, $\rm HC_7N$, $\rm IC_9N$, see @snell81 [@dickinfon82].
We alsm kbserve that the eario $k_{B_1J'_1}(\teftrm{Fw})/k_{J_1G'_1}(\mbkx{'arz-H$_2$}) $ is in average cmose to $1/1.4 \sin 1/\sqrt{2}$, thus confirmond the similarify of Re and para-[$\rm H_2$]{}as projectiles, as generallj asaumed. But it is also fae from being a constajt, as alrqady observed for H$_2$O [@phillips96] or CO [@wernli06]. Our dada shkds uhqt thd $1/\sert{2}$ scaling rule results in errors up to 50%.
Popujztook inversion and gritical densities
-------------------------------------------
Nefaife of the strung $\Delta J_1=0,2,4$ propensity rule, populajion ibversion soulc be strengthened if LTE cobditions are bot met, even negleeting hyperfnne efgects[^1] [@hunt99]. In order to see tke denaity conditlons givihe rise to populagiok itversion, we solved the stewdy-state xquatnons for the populwtion of tje $J=0,1,\dots,15$ levels of [$\mathrl{HC_3N}$]{}, nncluging collidions with [$\rm H_2$]{}(densities ranginj from $10^2$ to $10^6 \rk\; wm^{-3}$), d black-bjdy pmoton bath at 2.7 K, in the opticakly thiu appruximation, [@foldsmivh72] : $$\begin{aligged}
\label{eq:ss}
\frac{\textmd{v}n_i | ]. Besides this strong $\Delta J = rule, can see table \[tab:rates2\] and rod-like drives large $\Delta transfers. For instance, T $> 20$ K, rates for J > 6$ are generally larger than rates for $\Delta J = 1$, rates for $\Delta J > 8$ are only one order of magnitude below for J 2$. behaviour is likely to emphasize the role of collisional effects versus radiative ones. This effect, of geometric origin, has been predicted previously [@bosanac80] and of even greater importance longer rods like $\rm HC_5N$, HC_7N$, HC_9N$, see [@dickinson82]. also that the ratio $ is in average close to $1/1.4 \sim 1/\sqrt{2}$, thus confirming the similarity of He and para-[$\rm projectiles, as But it also from a constant, as for H$_2$O [@phillips96] or CO [@wernli06]. that the $1/\sqrt{2}$ scaling rule results in errors to 50%. inversion and critical densities ------------------------------------------- Because the strong $\Delta J_1=0,2,4$ propensity rule, population inversion be strengthened if LTE conditions are not met, even neglecting hyperfine effects[^1] [@hunt99]. In order the density conditions giving to population inversion, solved steady-state for population of $J=0,1,\dots,15$ levels of [$\mathrm{HC_3N}$]{}, including collisions with [$\rm H_2$]{}(densities ranging from to $10^6 \rm\; cm^{-3}$), a black-body photon bath at 2.7 the thin approximation, [@goldsmith72] $$\begin{aligned} \label{eq:ss} \frac{\textmd{d}n_i | ].
Besides this strong $\Delta J = 2$ prOpensity ruLe, one Can See FrOm taBle \[tAb:rates2\] and figuREs \[fiG:ratecompare\], \[fig:rate12\] thaT the rOd-LIke iNTeRactiOn driveS LaRGE $\DeLtA J$ TraNsFErS. For iNstAnce, for t $> 20$ K, rates for $\delTa j > 6$ are generallY LaRger than raTes For $\Delta J = 1$, and RatEs for $\DElTa J > 8$ ARe onlY onE ordeR of magNItude bElow those FoR $\delta J = 2$. tHis behaVIOuR is lIkely to emphasize tHE rOLe of collisionaL effecTs VErSUS raDiaTive ones. ThIs EffecT, Of purelY GeOMETriC Origin, has been Predicted prEVioUsly [@boSaNac80] ANd is of Even gReATer Importance fOr loNger rods lIke $\rm Hc_5n$, $\rm HC_7N$, $\rM hC_9N$, see @sNell81 [@diCkiNsoN82].
We aLSo ObSerVe THat THe RatIO $k_{J_1j'_1}(\textrm{HE})/k_{j_1J'_1}(\Mbox{pAra-H$_2$}) $ IS IN AverAge ClosE to $1/1.4 \siM 1/\sqrt{2}$, thus confIrmIng tHE siMilarIty of he anD pAra-[$\rm h_2$]{}as proJectiLeS, as generally assUmed. but it is alSo fAr FroM bEing a COnstanT, as AlrEady obsErved foR h$_2$O [@pHiLLIPs96] Or CO [@wernli06]. Our data sHoWS ThAt the $1/\sqrT{2}$ scaliNG rUlE Results iN eRroRs up TO 50%.
populAtioN InVersion aNd critICaL dEnsitieS
-------------------------------------------
BEcause Of The StrOng $\DeLTa J_1=0,2,4$ pRopensIty rule, pOpulaTIon inversion coULd be strengtheNEd IF lTe CondItiOns are not meT, eveN NeglEctiNG hYpeRFine eFfectS[^1] [@hUNt99]. iN order to see the densiTy ConditIons gIving rise to poPulation inVERSion, we soLved THe STeady-state equaTions For the popuLAtion of tHe $J=0,1,\doTs,15$ levels Of [$\mathrm{Hc_3n}$]{}, IncludinG coLliSioNs wITH [$\rM H_2$]{}(densities raNGIng fRoM $10^2$ to $10^6 \rm\; cm^{-3}$), A blAck-body PhoTon BatH at 2.7 k, iN the opticAlly thin ApPrOxImAtiOn, [@golDSmith72] : $$\begIn{AliGnEd}
\lAbel{eQ:Ss}
\frac{\TextmD{d}n_i | ].
Besides this strong $\ Delta J =2$ pr ope nsi ty rul e, o ne can see fro m tab le \[tab:rates2\] andfigur es \[fi g :r ateco mpare\] , \ [ f ig: ra te 12\ ]t ha t the ro d-likeinteractio n d ri ves large $\ D el ta J$ tran sfe rs. For inst anc e, for T $> 20$ K , r atesfor $\ D elta J > 6$ are g e nerall y larger t ha n ra tes for $\Delta J =1 $, and rates f or $\D el t aJ > 8 $ a re only on eorder of magn i tu d e bel o w those for $ \Delta J =2 $.This b eh avi o ur islikel yt o e mphasize th e ro le of col lision a l effec t s versu s radi ati veones . T hi s e ff e ct, of pu r ely geometr ic o rigin , ha s b e en p red icte d pre viously [@bos ana c80] and is o f eve n gr ea ter i mporta nce f or longer rods li ke $ \rm HC_5N $,$\ rmHC _7N$, $\rm H C_9 N$, see @s nell81[ @di ck i n s on 82].
We also obse rv e th at the r atio $ k _{ J_ 1 J'_1}(\t ex trm {He} ) / k_{J_ 1J'_ 1 }( \mbox{pa ra-H$_ 2 $} )$ is in a verage c los e t o $1/ 1 .4 \ sim 1/ \sqrt{2} $, th u s confirming t h e similarityo fH e a n d pa ra- [$\rm H_2$] {}as proj ecti l es , a s gene rally a s su m ed. But it is alsofa r from bein g a constant, as alread y o bservedforH $_ 2 $O [@phillips9 6] or CO [@wern l i06]. Ou r dat a showsthat the$ 1 /\sqrt{2 }$sca lin g r u l eresults in er r o rs u pto 50%.
P opulati oninv ers ion a nd critic al densi ti es
- -- --- ----- - -------- -- --- -- --- ----- - ------ -
Be caus eof the strong $\ D e ltaJ_ 1= 0,2, 4$pr opens ityr ule , popul ation inv ers i on c ou ld be str engthened ifLT E conditio ns ar e notm e t, evenneglecting hyperfine ef f ects[^1 ] [ @hunt 99]. In order to see t hed ensity condi tions g ivi n g rise t opop ul ation inve r s ion , weso lved the st eady-state equatio n s f or the popula tio n of t he $J = 0, 1 ,\d ot s ,15 $ levels of [$\ma thrm{HC_3N }$ ] {} , includin g co ll isionswith [$ \rm H _ 2$]{}(d ensitiesranging f ro m $1 0 ^ 2$to $10^6 \ rm\; cm^ {-3}$), a black - bo dy ph oto n bath a t 2 .7 K, in th e op tical ly thi napprox imati on , [@gold smith72] : $$\begin{ali gned}\labe l{e q:ss}
\f r ac{ \textmd{d }n_i | ].
Besides this_strong $\Delta_J = 2$ propensity_rule, one_can_see from_table \[tab:rates2\]_and figures \[fig:ratecompare\],_\[fig:rate12\] that the_rod-like interaction drives large_$\Delta J$ transfers._For_instance, for T $> 20$ K, rates for $\Delta J > 6$ are generally larger_than_rates for_$\Delta_J_= 1$, and rates for_$\Delta J > 8$ are_only one_order of magnitude below those for $\Delta J_=_2$. This behaviour_is likely to emphasize the role of collisional effects_versus radiative ones. This effect, of_purely geometric origin,_has_been_predicted previously [@bosanac80] and_is of even greater importance for_longer rods like $\rm HC_5N$, $\rm_HC_7N$, $\rm HC_9N$, see @snell81 [@dickinson82].
We also_observe that the ratio $k_{J_1J'_1}(\textrm{He})/k_{J_1J'_1}(\mbox{para-H$_2$}) $_is in average close to_$1/1.4 \sim_1/\sqrt{2}$, thus confirming the similarity_of He and_para-[$\rm H_2$]{}as_projectiles, as generally_assumed. But it is also far_from being a_constant, as already observed for H$_2$O_[@phillips96]_or CO [@wernli06]._Our_data_shows that_the $1/\sqrt{2}$ scaling_rule_results in_errors_up to 50%.
Population inversion and critical_densities
-------------------------------------------
Because_of the strong $\Delta J_1=0,2,4$ propensity rule,_population inversion could be_strengthened_if LTE conditions are_not met, even neglecting hyperfine_effects[^1] [@hunt99]. In order to see_the density_conditions giving_rise to population inversion, we solved the steady-state equations for the_population of the $J=0,1,\dots,15$ levels of_[$\mathrm{HC_3N}$]{}, including collisions with_[$\rm H_2$]{}(densities_ranging_from $10^2$ to_$10^6_\rm\; cm^{-3}$),_a black-body photon bath at 2.7 K, in_the optically_thin approximation, [@goldsmith72] : $$\begin{aligned}
\label{eq:ss}
_ \frac{\textmd{d}n_i |
ederna F. Olsson, A.S. Levine and H.B. Schioth, [*Analysis of the network of feeding neuroregulators using the Allen Brain Atlas*]{}, Neurosci. Biobehav. Rev. [**[32]{}**]{}, 945–956 (2008).
H.-W. Dong, *The Allen reference atlas: a digital brain atlas of the C57BL/6J male mouse*, Wiley, 2007.
P. Grange and P.P. Mitra, [*Computational neuroanatomy and gene expression: Optimal sets of marker genes for brain regions*]{}, IEEE, in CISS 2012, 46th annual conference on Information Science and Systems (Princeton),.
J.W. Bohland, H. Bokil, C.-K. Lee, L. Ng, C. Lau, C. Kuan, M. Hawrylycz and P.P. Mitra, [*Clustering of spatial gene expression patterns in the mouse brain and comparison with classical neuroanatomy*]{}, Methods, Volume [**[50]{}**]{}, Issue 2, February 2010, Pages 105-112.
C. Lau, L. Ng, C. Thompson, S. Pathak, L. Kuan, A. Jones and M. Hawrylycz, [*Exploration and visualization of gene expression with neuroanatomy in the adult mouse brain*]{}, BMC Bioinformatics 2008, [**[8]{}**]{}:153.
M. Hawrylycz, R.A. Baldock, A. Burger, T. Hashikawa, G.A. Johnson, M. Martone, L. Ng, C. Lau, S.D. Larsen, J. Nissanov, L. Puelles, S. Ruffins, F. Verbeek, I. Zaslavsky1 and J. Boline, [*Digital Atlasing and Standardization in the Mouse Brain*]{}, PLoS Computational Biology [**[7]{}**]{} (2) (2011).
The Allen Brain Atlas can be used online at.
The developmental atlas of the mouse brain is available from
I. Menashe, P. Grange, E.C. Larsen, S. Banerjee-Basu and P.P. Mitra, [*Co-expression profiling of autism genes in the mouse brain*]{}, SFN Abstracts 2012, and in preparation.
M. Hawrylycz [*et al | ederna F. Olsson, A.S. Levine and H.B. Schioth, [ * Analysis of the network of feeding neuroregulators using the Allen Brain Atlas * ] { }, Neurosci. Biobehav. Rev. [ * * [ 32 ] { } * * ] { }, 945–956 (2008).
H.-W. Dong, * The Allen reference point atlas: a digital mind atlas of the C57BL/6J male mouse *, Wiley, 2007.
P. Grange and P.P. Mitra, [ * Computational neuroanatomy and gene expression: Optimal sets of marker gene for brain regions * ] { }, IEEE, in CISS 2012, 46th annual league on Information Science and Systems (Princeton), .
J.W. Bohland, H. Bokil, C.-K. Lee, L. Ng, C. Lau, C. Kuan, M. Hawrylycz and P.P. Mitra, [ * bunch of spatial gene expression patterns in the mouse brain and comparison with classical neuroanatomy * ] { }, Methods, Volume [ * * [ 50 ] { } * * ] { }, offspring 2, February 2010, Pages 105 - 112.
C. Lau, L. Ng, C. Thompson, S. Pathak, L. Kuan, A. Jones and M. Hawrylycz, [ * Exploration and visualization of gene expression with neuroanatomy in the adult mouse genius * ] { }, BMC Bioinformatics 2008, [ * * [ 8]{}**]{}:153.
M. Hawrylycz, R.A. Baldock, A. Burger, T. Hashikawa, G.A. Johnson, M. Martone, L. Ng, C. Lau, S.D. Larsen, J. Nissanov, L. Puelles, S. Ruffins, F. Verbeek, I. Zaslavsky1 and J. Boline, [ * Digital Atlasing and Standardization in the Mouse Brain * ] { }, PLoS Computational Biology [ * * [ 7 ] { } * * ] { } (2) (2011).
The Allen Brain Atlas can be use online at.
The developmental atlas of the mouse mind is available from
I. Menashe, P. Grange, E.C. Larsen, S. Banerjee - Basu and P.P. Mitra, [ * Co - expression profiling of autism gene in the mouse brain * ] { }, SFN Abstracts 2012, and in preparation.
M. Hawrylycz [ * et al | edegna F. Olsson, A.S. Levine akd H.B. Schioth, [*Ancoysis mf the network of feeding neuroregulators nsint the Allen Brain Atlas*]{}, Neufosci. Biohehav. Rec. [**[32]{}**]{}, 945–956 (2008).
I.-W. Dong, *The Allei referekee atmws: a vigital brain ajlas of the W57BL/6J male mousa*, Dipey, 2007.
P. Grange and P.P. Mitra, [*Computatijnal neirlanatomy and ggne eqpwessjon: Optimal sets of marker genes fkr braii regions*]{}, IEEE, on CISS 2012, 46th annual conferejce ln Information Scifnce and Sywtemf (Princeton),.
J.W. Bohland, H. Bokil, C.-K. Leg, L. Ng, C. Lau, C. Kuan, M. Hawrylycz xnd P.'. Mitra, [*Cluwtwrijc of spatiao genv expression istternv in thr mouse brain snd conparison with classicel neuroanatomy*]{}, Methjds, Voluma [**[50]{}**]{}, Issue 2, February 2010, Pqges 105-112.
W. Lag, L. Vt, C. Thkm'soh, S. Pahhai, L. Kuan, A. Jones and N. Hawrylycz, [*Explorauiog and visualizztion jf gene expression with neuroanatomy in tve zdult mouse brain*]{}, BMC Buoinformatics 2008, [**[8]{}**]{}:153.
M. Hawrilycz, R.A. Bwldock, A. Burger, T. Hashikawa, G.A. Johnson, M. Martone, N. Ng, R. Uau, S.Q. Oagsen, J. Nissanov, L. Puelles, S. Ruffins, F. Verbeek, J. Easkavsky1 and J. Bjline, [*Digitsl Ayjasing and Stxndardnaafion in the Mouse Hrain*]{}, PJoS Cimputatiogal Niology [**[7]{}**]{} (2) (2011).
The Allen Brain Arlas can be lsed online at.
The devepopmental aclas og the mouse brain is availabue fdom
I. Menashf, P. Grangs, E.C. Larsen, S. Bandrjve-Bavu and P.K. Mitra, [*Co-expressijn profilmng oy autism gengs in tre mouse bgain*]{}, SFN Abstracts 2012, and ln prgparathon.
M. Hawrypycz [*et al | ederna F. Olsson, A.S. Levine and H.B. of network of neuroregulators using the Rev. 945–956 (2008). H.-W. *The Allen reference a digital brain atlas of the male mouse*, Wiley, 2007. P. Grange and P.P. Mitra, [*Computational neuroanatomy and gene Optimal sets of marker genes for brain regions*]{}, IEEE, in CISS 2012, 46th conference Information and (Princeton),. J.W. Bohland, H. Bokil, C.-K. Lee, L. Ng, C. Lau, C. Kuan, M. Hawrylycz and Mitra, [*Clustering of spatial gene expression patterns in mouse brain and comparison classical neuroanatomy*]{}, Methods, Volume [**[50]{}**]{}, 2, 2010, Pages C. L. C. Thompson, S. L. Kuan, A. Jones and M. Hawrylycz, [*Exploration and visualization of gene expression with neuroanatomy in the mouse brain*]{}, 2008, [**[8]{}**]{}:153. Hawrylycz, Baldock, Burger, T. Hashikawa, M. Martone, L. Ng, C. Lau, Nissanov, L. Puelles, S. Ruffins, F. Verbeek, I. and J. [*Digital Atlasing and Standardization in the Brain*]{}, PLoS Computational Biology [**[7]{}**]{} (2) (2011). The Brain Atlas can be used online at. The developmental atlas of the mouse brain is I. Menashe, P. Grange, Larsen, S. Banerjee-Basu P.P. [*Co-expression of genes in mouse brain*]{}, SFN Abstracts 2012, and in preparation. M. Hawrylycz [*et | ederna F. Olsson, A.S. Levine and H.b. Schioth, [*AnAlysiS of The NeTworK of fEeding neuroregULatoRs using the Allen Brain AtLas*]{}, NeUrOSci. BIObEhav. REv. [**[32]{}**]{}, 945–956 (2008).
H.-W. DonG, *thE aLleN rEfEreNcE AtLas: a dIgiTal braiN atlas of thE C57Bl/6J Male mouse*, WilEY, 2007.
P. grange and P.p. MiTra, [*ComputatiOnaL neuroAnAtoMY and gEne ExpreSsion: OPTimal sEts of markEr GEnes foR Brain reGIOnS*]{}, IEEe, in CISS 2012, 46th annual coNFeREnce on InformatIon SciEnCE aND sysTemS (Princeton),.
j.W. bohlaND, H. Bokil, c.-k. LEE, l. ng, C. lAu, C. Kuan, M. HawryLycz and P.P. MiTRa, [*CLusterInG of SPatial Gene eXpREssIon patterns In thE mouse braIn and cOMparisoN With claSsical NeuRoaNatoMY*]{}, MEtHodS, VOLumE [**[50]{}**]{}, isSue 2, fEbrUary 2010, PageS 105-112.
C. laU, L. Ng, C. thomPSON, s. PatHak, l. KuaN, A. JonEs and M. HawrylyCz, [*EXploRAtiOn and VisuaLizaTiOn of gEne expRessiOn With neuroanatomY in tHe adult moUse BrAin*]{}, bMc BioiNFormatIcs 2008, [**[8]{}**]{}:153.
m. HaWrylycz, r.A. BaldoCK, A. BUrGER, t. HAshikawa, G.A. Johnson, M. maRTOnE, L. Ng, C. Lau, s.D. LarsEN, J. niSSanov, L. PuElLes, s. RufFINs, F. VeRbeeK, i. ZAslavsky1 And J. BoLInE, [*DIgital ATlAsing aNd staNdaRdizaTIon iN the MoUse Brain*]{}, pLoS COMputational BioLOgy [**[7]{}**]{} (2) (2011).
The Allen BrAIn aTLaS Can bE usEd online at.
THe deVElopMentAL aTlaS Of the Mouse BrAIn IS available from
I. MenaShE, P. GranGe, E.C. LArsen, S. BanerjeE-Basu and P.P. mITRa, [*Co-exprEssiON pROfiling of autisM geneS in the mousE Brain*]{}, SFN abstrActs 2012, and iN preparatION.
M. HawrylYcz [*Et aL | ederna F. Olsson, A.S. Lev ine and H. B. Sc hio th, [ *Ana lysi s of the netwo r k of feeding neuroregulato rs us in g the Al len B rain At l as * ] {}, N eu ros ci . B iobeh av. Rev. [ **[32]{}** ]{} ,945–956 (200 8 ).
H.-W. Do ng, *The Allenref erence a tla s : a d igi tal b rain a t las of the C57B L/ 6 J male mouse*, W il ey,2007.
P. Grangea nd P.P. Mitra, [* Comput at i on a l ne uro anatomy an dgenee xpressi o n: O p tim a l sets of mar ker genes f o r b rain r eg ion s *]{},IEEE, i n CI SS 2012, 46 th a nnual con ferenc e on Inf o rmation Scien ceand Sys t em s(Pr in c eto n ), .
J .W. Bohland ,H. Boki l, C . - K . Lee , L . Ng , C.Lau, C. Kuan, M. Haw r yly cz an d P.P . Mi tr a, [* Cluste ringof spatial gene e xpre ssion pat ter ns in t he mo u se bra inand compar ison wi t h c la s s i ca l neuroanatomy*]{} ,M e th ods, Vol ume [* * [5 0] { }**]{},Is sue 2,F e bruar y 20 1 0, Pages 1 05-112 .
C. Lau, L .Ng, C. T hom pso n, S. Path ak, L. Kuan, A . Jon e s and M. Hawry l ycz, [*Explor a ti o n a n d vi sua lization of gen e exp ress i on wi t h neu roana to m yi n the adult mouse b ra in*]{} , BMC Bioinformati cs 2008, [ * * [ 8]{}**]{ }:15 3 .M. Hawrylycz,R.A.Baldock, A . Burger, T. H ashikawa , G.A. Jo h n son, M.Mar ton e,L.N g ,C. Lau, S.D.L a rsen ,J. Niss ano v, L. P uel les , S . R uf fins, F.Verbeek, I .Za sl avs ky1 a n d J. Bol in e,[* Dig italA tlasin g and Sta nd ar d iza tion in th e Mous eBr ain* ]{} ,PLoSComp u tat ional B iology [* *[7 ] {}** ]{ }(2) (20 11).
The All en Brain Atl as ca n be u s e d online at.
The developmental atlas o f t he mo usebrain isava ilable fr o m
I.Menash e, P. G ran g e , E.C . La rse n, S. Banerj e e -Ba su an dP.P. Mitra, [*Co-expression p r ofi ling of autis m g enes i nthe mo u sebr a in* ] { }, SFN Abstract s 2012, an di npreparatio n .
M. Hawryl ycz [*e t al | ederna F._Olsson, A.S._Levine and H.B. Schioth,_[*Analysis of_the_network of_feeding_neuroregulators using the_Allen Brain Atlas*]{},_Neurosci. Biobehav. Rev. [**[32]{}**]{},_945–956 (2008).
H.-W. Dong,_*The_Allen reference atlas: a digital brain atlas of the C57BL/6J male mouse*, Wiley, 2007.
P._Grange_and P.P._Mitra,_[*Computational_neuroanatomy and gene expression: Optimal_sets of marker genes for_brain regions*]{},_IEEE, in CISS 2012, 46th annual conference on_Information_Science and Systems_(Princeton),.
J.W. Bohland, H. Bokil, C.-K. Lee, L. Ng, C._Lau, C. Kuan, M. Hawrylycz and_P.P. Mitra, [*Clustering_of_spatial_gene expression patterns in_the mouse brain and comparison with_classical neuroanatomy*]{}, Methods, Volume [**[50]{}**]{}, Issue_2, February 2010, Pages 105-112.
C. Lau, L._Ng, C. Thompson, S. Pathak, L._Kuan, A. Jones and M._Hawrylycz, [*Exploration_and visualization of gene expression_with neuroanatomy in_the adult_mouse brain*]{}, BMC_Bioinformatics 2008, [**[8]{}**]{}:153.
M. Hawrylycz, R.A. Baldock,_A. Burger, T._Hashikawa, G.A. Johnson, M. Martone, L._Ng,_C. Lau, S.D._Larsen,_J._Nissanov, L._Puelles, S. Ruffins,_F._Verbeek, I._Zaslavsky1_and J. Boline, [*Digital Atlasing and_Standardization_in the Mouse Brain*]{}, PLoS Computational Biology_[**[7]{}**]{} (2) (2011).
The Allen_Brain_Atlas can be used_online at.
The developmental atlas of_the mouse brain is available from
I._Menashe, P._Grange, E.C._Larsen, S. Banerjee-Basu and P.P. Mitra, [*Co-expression profiling of autism genes_in the mouse brain*]{}, SFN Abstracts_2012, and in preparation.
M._Hawrylycz [*et_al |
Q$ such that $Qf(t_i,\cdot) :=Q_{t_i}f(t_i,\cdot)$ where for $1\leq i\leq n$ and $g:{\ensuremath{\mathbb{R}}}\to{\ensuremath{\mathbb{R}}}$, $Q_{t_i}g(x) := q_{t_i}(x)g(x)$.
\[thm1\] Assume that for all $1\leq i\leq j\leq k\leq n$ the following holds:
- [Right-invertibility]{}: ${\mathcal{W}}_{t_i,t_j}{\mathcal{W}}_{t_j,t_i}K_{t_i}=K_{t_i}$;
- [Semigroup property]{}: ${\mathcal{W}}_{t_i,t_j}{\mathcal{W}}_{t_j,t_k}={\mathcal{W}}_{t_i,t_k}$;
- [Reversibility relation]{}: ${\mathcal{W}}_{t_i,t_j}K_{t_j}=K_{t_i}{\mathcal{W}}_{t_i,t_j}$.
Then $$\det\!\big(I-QK^{\rm ext}\big)_{L^2(\{t_1,\dots,t_n\}\times {\ensuremath{\mathbb{R}}},\mu)}
=\det\!\big(I-K_{t_1}+{\overline{Q}}_{t_1}{\mathcal{W}}_{t_1,t_2}{\overline{Q}}_{t_2}\dotsm{\mathcal{W}}_{t_{n-1},t_n}{\overline{Q}}_{t_n}
{\mathcal{W}}_{t_n,t_1}K_{t_1}\big)_{L^2({\ensuremath{\mathbb{R}}})},$$ where $$K^{\rm ext}(t_i,x;t_j,y)=
\begin{dcases*}
{\mathcal{W}}_{t_i,t_j}K_{t_j}(x,y) & if $i\geq j$,\\
-{\mathcal{W}}_{t_i,t_j}(I-K_{t_j})(x,y) & if $i<j$,
\end{dcases*}$$ and ${\overline{Q}}_{t_ | Q$ such that $ Qf(t_i,\cdot): = Q_{t_i}f(t_i,\cdot)$ where for $ 1\leq i\leq n$ and $ g:{\ensuremath{\mathbb{R}}}\to{\ensuremath{\mathbb{R}}}$, $ Q_{t_i}g(x): = q_{t_i}(x)g(x)$.
\[thm1\ ] Assume that for all $ 1\leq i\leq j\leq k\leq n$ the following holds:
- [ Right - invertibility ] { }: $ { \mathcal{W}}_{t_i, t_j}{\mathcal{W}}_{t_j, t_i}K_{t_i}=K_{t_i}$;
- [ Semigroup place ] { }: $ { \mathcal{W}}_{t_i, t_j}{\mathcal{W}}_{t_j, t_k}={\mathcal{W}}_{t_i, t_k}$;
- [ Reversibility relation back ] { }: $ { \mathcal{W}}_{t_i, t_j}K_{t_j}=K_{t_i}{\mathcal{W}}_{t_i, t_j}$.
Then $ $ \det\!\big(I - QK^{\rm ext}\big)_{L^2(\{t_1,\dots, t_n\}\times { \ensuremath{\mathbb{R}}},\mu) }
= \det\!\big(I - K_{t_1}+{\overline{Q}}_{t_1}{\mathcal{W}}_{t_1,t_2}{\overline{Q}}_{t_2}\dotsm{\mathcal{W}}_{t_{n-1},t_n}{\overline{Q}}_{t_n }
{ \mathcal{W}}_{t_n, t_1}K_{t_1}\big)_{L^2({\ensuremath{\mathbb{R}}})},$$ where $ $ K^{\rm ext}(t_i, x;t_j, y)=
\begin{dcases * }
{ \mathcal{W}}_{t_i, t_j}K_{t_j}(x, y) & if $ i\geq j$,\\
-{\mathcal{W}}_{t_i, t_j}(I - K_{t_j})(x, y) & if $ i < j$,
\end{dcases*}$$ and $ { \overline{Q}}_{t _ | Q$ skch that $Qf(t_i,\cdot) :=Q_{t_i}f(t_i,\gdot)$ where for $1\lgq i\leq i$ and $g:{\snsuremagh{\mathbb{R}}}\to{\ensuremath{\mathbb{R}}}$, $Q_{r_i}g(x) := q_{t_i}(x)g(x)$.
\[thm1\] Assume that for all $1\peq i\leq j\lew k\leq n$ thx followlug homfs:
- [Cight-invertibillty]{}: ${\mathcal{F}}_{t_i,t_j}{\mathcal{W}}_{t_b,t_k}K_{c_i}=K_{t_i}$;
- [Semigroup property]{}: ${\mathcal{W}}_{t_y,t_j}{\mathvap{W}}_{t_j,t_k}={\mathcal{W}}_{j_i,t_k}$;
- [Revsgslbility relation]{}: ${\mathcal{W}}_{t_i,t_j}K_{t_j}=I_{t_i}{\mathbal{W}}_{t_i,t_j}$.
Then $$\det\!\bog(I-QK^{\rm ext}\big)_{L^2(\{t_1,\dots,t_n\}\timed {\enduremath{\mathbb{R}}},\mu)}
=\dft\!\big(I-K_{t_1}+{\ovetminq{W}}_{t_1}{\mathcal{W}}_{t_1,t_2}{\uverline{Q}}_{t_2}\dotsm{\mathcal{S}}_{t_{n-1},t_n}{\overline{Q}}_{t_n}
{\mathcal{W}}_{t_n,t_1}Y_{t_1}\big)_{K^2({\ensurematy{\mqthht{R}}})},$$ where $$K^{\rn ext}(n_i,x;t_j,y)=
\begin{dcases*}
{\mathcsl{W}}_{t_i,t_j}K_{t_j}(x,y) & lf $i\gxq j$,\\
-{\mathcal{W}}_{t_i,t_j}(I-K_{t_j})(x,b) & if $i<j$,
\end{dcases*}$$ wnd ${\overlhnz{Q}}_{t_ | Q$ such that $Qf(t_i,\cdot) :=Q_{t_i}f(t_i,\cdot)$ where for n$ $g:{\ensuremath{\mathbb{R}}}\to{\ensuremath{\mathbb{R}}}$, $Q_{t_i}g(x) q_{t_i}(x)g(x)$. \[thm1\] Assume j\leq n$ the following - [Right-invertibility]{}: ${\mathcal{W}}_{t_i,t_j}{\mathcal{W}}_{t_j,t_i}K_{t_i}=K_{t_i}$; [Semigroup property]{}: ${\mathcal{W}}_{t_i,t_j}{\mathcal{W}}_{t_j,t_k}={\mathcal{W}}_{t_i,t_k}$; - [Reversibility relation]{}: Then $$\det\!\big(I-QK^{\rm ext}\big)_{L^2(\{t_1,\dots,t_n\}\times {\ensuremath{\mathbb{R}}},\mu)} =\det\!\big(I-K_{t_1}+{\overline{Q}}_{t_1}{\mathcal{W}}_{t_1,t_2}{\overline{Q}}_{t_2}\dotsm{\mathcal{W}}_{t_{n-1},t_n}{\overline{Q}}_{t_n} {\mathcal{W}}_{t_n,t_1}K_{t_1}\big)_{L^2({\ensuremath{\mathbb{R}}})},$$ where $$K^{\rm ext}(t_i,x;t_j,y)= \begin{dcases*} {\mathcal{W}}_{t_i,t_j}K_{t_j}(x,y) & if j$,\\ -{\mathcal{W}}_{t_i,t_j}(I-K_{t_j})(x,y) & if $i<j$, \end{dcases*}$$ and ${\overline{Q}}_{t_ | Q$ such that $Qf(t_i,\cdot) :=Q_{t_i}f(t_i,\cdOt)$ where for $1\Leq i\lEq n$ And $G:{\eNsurEmatH{\mathbb{R}}}\to{\ensuREmatH{\mathbb{R}}}$, $Q_{t_i}g(x) := q_{t_i}(x)g(x)$.
\[thm1\] assumE tHAt foR AlL $1\leq i\Leq j\leq K\LeQ N$ The FoLlOwiNg HOlDs:
- [RigHt-iNvertibIlity]{}: ${\mathcAl{W}}_{T_i,T_j}{\mathcal{W}}_{t_j,T_I}K_{T_i}=K_{t_i}$;
- [SemigRouP property]{}: ${\matHcaL{W}}_{t_i,t_j}{\MaThcAL{W}}_{t_j,t_K}={\maThcal{w}}_{t_i,t_k}$;
- [REVersibIlity relaTiON]{}: ${\mathcAL{W}}_{t_i,t_j}K_{T_J}=k_{t_I}{\matHcal{W}}_{t_i,t_j}$.
Then $$\det\!\bIG(I-qk^{\rm ext}\big)_{L^2(\{t_1,\dotS,t_n\}\timEs {\ENsUREmaTh{\mAthbb{R}}},\mu)}
=\deT\!\bIg(I-K_{t_1}+{\OVerline{q}}_{T_1}{\mATHCal{w}}_{T_1,t_2}{\overline{Q}}_{t_2}\dOtsm{\mathcal{w}}_{T_{n-1},t_N}{\overlInE{Q}}_{t_N}
{\MathcaL{W}}_{t_n,t_1}k_{t_1}\BIg)_{L^2({\Ensuremath{\mAthbB{R}}})},$$ where $$K^{\rM ext}(t_i,X;T_j,y)=
\begiN{Dcases*}
{\mAthcal{w}}_{t_i,T_j}K_{T_j}(x,y) & IF $i\GeQ j$,\\
-{\mAtHCal{w}}_{T_i,T_j}(I-k_{T_j})(x,Y) & if $i<j$,
\end{DcAsEs*}$$ and ${\OverLINE{q}}_{t_ | Q$ such that $Qf(t_i,\cdot ) :=Q_{t_i }f(t_ i,\ cdo t) $ wh erefor $1\leq i\l e q n$ and $g:{\ensuremath{\ mathb b{ R }}}\ t o{ \ensu remath{ \ ma t h bb{ R} }} $,$Q _ {t _i}g( x):= q_{t _i}(x)g(x) $.
\ [thm1\] Assu m ethat for a ll$1\leq i\leq j\ leq k\ le q n $ thefol lowin g hold s :
- [Right-i nv e rtibil i ty]{}:$ { \m athc al{W}}_{t_i,t_j}{ \ ma t hcal{W}}_{t_j, t_i}K_ {t _ i} = K _{t _i} $;
- [S em igrou p proper t y] { } : ${ \ mathcal{W}}_{ t_i,t_j}{\m a thc al{W}} _{ t_j , t_k}={ \math ca l {W} }_{t_i,t_k} $;
- [Reve rsibil i ty rela t ion]{}: ${\ma thc al{ W}}_ { t_ i, t_j }K _ {t_ j }= K_{ t _i} {\mathca l{ W} }_{t_ i,t_ j } $ .
Th en$$\d et\!\ big(I-QK^{\rm ex t}\b i g)_ {L^2( \{t_1 ,\do ts ,t_n\ }\time s {\e ns uremath{\mathbb {R}} },\mu)}
= \de t\ !\b ig (I-K_ { t_1}+{ \ov erl ine{Q}} _{t_1}{ \ mat hc a l { W} }_{t_1,t_2}{\overl in e { Q} }_{t_2}\ dotsm{ \ ma th c al{W}}_{ t_ {n- 1},t _ n }{\ov erli n e{ Q}}_{t_n }
{ \m at hcal{W} }_ {t_n,t _1 }K_ {t_ 1}\bi g )_{L ^2({\e nsuremat h{\ma t hbb{R}}})},$$w here $$K^{\rm ex t } (t _ i,x; t_j ,y)=
\beg in{d c ases *}
{\m a thcal {W}}_ {t _ i, t _j}K_{t_j}(x,y) & i f$i\geq j$,\ \
-{\math cal{W}}_{t _ i , t_j}(I-K _{t_ j }) ( x,y) & if $i<j $,
\end{dcase s *}$$ and ${\o verline{ Q}}_{t_ | Q$ such_that $Qf(t_i,\cdot)_:=Q_{t_i}f(t_i,\cdot)$ where for $1\leq_i\leq n$_and_$g:{\ensuremath{\mathbb{R}}}\to{\ensuremath{\mathbb{R}}}$, $Q_{t_i}g(x)_:=_q_{t_i}(x)g(x)$.
\[thm1\] Assume that_for all $1\leq_i\leq j\leq k\leq n$_the following holds:
-__ [Right-invertibility]{}: ${\mathcal{W}}_{t_i,t_j}{\mathcal{W}}_{t_j,t_i}K_{t_i}=K_{t_i}$;
- [Semigroup property]{}: ${\mathcal{W}}_{t_i,t_j}{\mathcal{W}}_{t_j,t_k}={\mathcal{W}}_{t_i,t_k}$;
- [Reversibility relation]{}: ${\mathcal{W}}_{t_i,t_j}K_{t_j}=K_{t_i}{\mathcal{W}}_{t_i,t_j}$.
Then $$\det\!\big(I-QK^{\rm_ext}\big)_{L^2(\{t_1,\dots,t_n\}\times_{\ensuremath{\mathbb{R}}},\mu)}
=\det\!\big(I-K_{t_1}+{\overline{Q}}_{t_1}{\mathcal{W}}_{t_1,t_2}{\overline{Q}}_{t_2}\dotsm{\mathcal{W}}_{t_{n-1},t_n}{\overline{Q}}_{t_n}
___{\mathcal{W}}_{t_n,t_1}K_{t_1}\big)_{L^2({\ensuremath{\mathbb{R}}})},$$ where $$K^{\rm ext}(t_i,x;t_j,y)=
_\begin{dcases*}
{\mathcal{W}}_{t_i,t_j}K_{t_j}(x,y)_& if_$i\geq j$,\\
-{\mathcal{W}}_{t_i,t_j}(I-K_{t_j})(x,y) & if_$i<j$,
_ \end{dcases*}$$ and_${\overline{Q}}_{t_ |
e_2$ in QD$_2$ are prepared in $|+\rangle_{e_1}$ and $|+\rangle_{e_2}$, respectively. Here $|\pm\rangle_e=\frac{1}{\sqrt{2}}(|\uparrow\rangle\pm|\downarrow\rangle)_e$.
After the two photons $A$ and $B$ in the hyperentangled Bell state pass through the quantum circuit shown in Fig. \[figure3.10\]a in sequence, the state of the system $ABe_1e_2$ evolves to $$\begin{aligned}
%eq.11 % Eq. 60
|\phi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\phi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\psi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\psi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2}.\;\;\;\;\;\;\;\;\end{aligned}$$ Then the excess electron spins $e_1$ and $e_2$ are measured in the orthogonal basis $\{|+\rangle_e, |-\rangle_e\}$. If the state of the excess electron spin $e_1$ is $|+\rangle_{e_1}$, the relative phase of the polarization state is $0$. If the state of the excess electron spin $e_1$ is $|-\rangle_{e_1}$, the relative phase of the | e_2 $ in QD$_2 $ are prepared in $ |+\rangle_{e_1}$ and $ |+\rangle_{e_2}$, respectively. Here $ |\pm\rangle_e=\frac{1}{\sqrt{2}}(|\uparrow\rangle\pm|\downarrow\rangle)_e$.
After the two photons $ A$ and $ B$ in the hyperentangled Bell state pass through the quantum racing circuit usher in Fig. \[figure3.10\]a in sequence, the state of the system $ ABe_1e_2 $ evolves to $ $ \begin{aligned }
% eq.11 % Eq. 60
|\phi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\phi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\psi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\psi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2}.\;\;\;\;\;\;\;\;\end{aligned}$$ Then the excess electron spin $ e_1 $ and $ e_2 $ are measured in the orthogonal footing $ \{|+\rangle_e, |-\rangle_e\}$. If the state of the excess electron spin $ e_1 $ is $ |+\rangle_{e_1}$, the relative phase of the polarization state of matter is $ 0$. If the state of the excess electron spin $ e_1 $ is $ |-\rangle_{e_1}$, the proportional phase of the | e_2$ ij QD$_2$ are prepared in $|+\rannle_{e_1}$ and $|+\rangle_{e_2}$, respecvively. Gere $|\pm\rxngle_e=\frac{1}{\sqrt{2}}(|\uparrow\rangle\pl|\diwnareow\rangle)_e$.
After the twu photons $A$ and $B$ in uhe hyperentanglev Bell sbcte pzds tkriugh the quantom circuit svown in Fig. \[figgrd3.10\]a in sequence, the state of the systei $ABe_1e_2$ rvllves to $$\begin{wligmqd}
%eq.11 % Es. 60
|\phi^\pm\gangle_P|\phi^\pm\ranglr_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\ohi^\pl\rangle_P|\phi^\pm\ranglf_S|\pm\rangle_{e_1}|\np\ragtle_{e_2},\nonumber\\
|\ohi^\pm\ranglt_P|\'si^\pm\rangle_A|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rxngle_'|\psi^\pm\ranglg_S|\lm\gdngle_{e_1}|\mp\ranjle_{e_2},\nogumber\\
|\psi^\pm\rangle_P|\phh^\pm\rangke_S|+\rangle_{e_1}|+\rangke_{e_2}&\cighrarrow&|\psi^\pm\rangle_P|\phi^\'m\rangle_S|\pm\rangle_{e_1}|\mp\tangle_{e_2},\nongmyer\\
|\psi^\pm\rangle_P|\psi^\pm\rqntle_S|+\rdngla_{e_1}|+\ravtle_{d_2}&\rifhvardow&|\psi^\om\rengle_P|\psi^\pm\dangle_S|\pm\rabgle_{e_1}|\mp\rangle_{e_2}.\;\;\;\;\;\;\;\;\end{aloggvc}$$ Then the escess qlqctron spins $e_1$ and $e_2$ are measured in tht ortgogonal basis $\{|+\rangle_e, |-\rqngle_e\}$. If the state ov the excqss electron spin $e_1$ is $|+\rangle_{e_1}$, the relative phase of tie pooavizaguoj state is $0$. If the state of the excess electrjh xppn $e_1$ is $|-\rangle_{e_1}$, tme relative phase pf tng | e_2$ in QD$_2$ are prepared in $|+\rangle_{e_1}$ respectively. $|\pm\rangle_e=\frac{1}{\sqrt{2}}(|\uparrow\rangle\pm|\downarrow\rangle)_e$. After two photons $A$ Bell pass through the circuit shown in \[figure3.10\]a in sequence, the state of system $ABe_1e_2$ evolves to $$\begin{aligned} %eq.11 % Eq. 60 |\phi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\ |\phi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\ |\psi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\ |\psi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2}.\;\;\;\;\;\;\;\;\end{aligned}$$ the excess electron spins $e_1$ and $e_2$ are measured in the orthogonal basis |-\rangle_e\}$. the of excess electron spin $e_1$ is $|+\rangle_{e_1}$, the relative phase of the polarization state is $0$. If state of the excess electron spin $e_1$ is the relative phase of | e_2$ in QD$_2$ are prepared in $|+\rangle_{e_1}$ And $|+\rangle_{e_2}$, RespeCtiVelY. HEre $|\pM\ranGle_e=\frac{1}{\sqrt{2}}(|\upARrow\Rangle\pm|\downarrow\ranglE)_e$.
AftEr THe twO PhOtons $a$ and $B$ in THe HYPerEnTaNglEd bElL statE paSs throuGh the quantUm cIrCuit shown in FIG. \[fIgure3.10\]a in seQueNce, the state oF thE systeM $Abe_1e_2$ EVolveS to $$\Begin{AligneD}
%Eq.11 % Eq. 60
|\phI^\pm\rangle_p|\pHI^\pm\ranGLe_S|+\rangLE_{E_1}|+\rAnglE_{e_2}&\rightarrow&|\phi^\pm\RAnGLe_P|\phi^\pm\rangle_s|\pm\ranGlE_{E_1}|\mP\RAngLe_{e_2},\Nonumber\\
|\phI^\pM\rangLE_P|\psi^\pm\RAnGLE_s|+\raNGle_{e_1}|+\rangle_{e_2}&\riGhtarrow&|\phi^\PM\raNgle_P|\pSi^\Pm\rANgle_S|\pM\rangLe_{E_1}|\Mp\rAngle_{e_2},\nonumBer\\
|\pSi^\pm\ranglE_P|\phi^\pM\Rangle_S|+\RAngle_{e_1}|+\rAngle_{e_2}&\RigHtaRrow&|\PSi^\Pm\RanGlE_p|\phI^\Pm\RanGLe_S|\Pm\rangle_{E_1}|\mP\rAngle_{E_2},\nonUMBER\\
|\psi^\Pm\rAnglE_P|\psi^\Pm\rangle_S|+\rangLe_{e_1}|+\RangLE_{e_2}&\rIghtaRrow&|\pSi^\pm\RaNgle_P|\Psi^\pm\rAngle_s|\pM\rangle_{e_1}|\mp\ranglE_{e_2}.\;\;\;\;\;\;\;\;\enD{aligned}$$ THen ThE exCeSs eleCTron spIns $E_1$ anD $e_2$ are meAsured iN The OrTHOGoNal basis $\{|+\rangle_e, |-\ranGlE_E\}$. if The state Of the eXCeSs ELectron sPiN $e_1$ iS $|+\ranGLE_{e_1}$, the RelaTIvE phase of The polARiZaTion staTe Is $0$. If thE sTatE of The exCEss eLectroN spin $e_1$ is $|-\RanglE_{E_1}$, the relative phASe of the | e_2$ in QD$_2$ are prepare d in $|+\r angle _{e _1} $and$|+\ rangle_{e_2}$, resp ectively. Here $|\pm\r angle _e = \fra c {1 }{\sq rt{2}}( | \u p a rro w\ ra ngl e\ p m| \down arr ow\rang le)_e$.
A fte rthe two phot o ns $A$ and $ B$in the hyper ent angled B ell state pa ss th rought he qua ntum circ ui t shown in Fig. \ [f igur e3.10\]a in seque n ce , the state ofthe sy st e m$ A Be_ 1e_ 2$ evolves t o $$\ b egin{al i gn e d }
% eq. 11 % Eq . 60
|\ phi^\ pm \ ran gle_P|\phi^ \pm\ rangle_S| +\rang l e_{e_1} | +\rangl e_{e_2 }&\ rig htar r ow &| \ph i^ \ pm\ r an gle _ P|\ phi^\pm\ ra ng le_S| \pm\ r a n g le_{ e_1 }|\m p\ran gle_{e_2},\no num ber\ \
|\ phi^\ pm\ra ngle _P |\psi ^\pm\r angle _S |+\rangle_{e_1} |+\r angle_{e_ 2}& \r igh ta rrow& | \phi^\ pm\ ran gle_P|\ psi^\pm \ ran gl e _ S |\ pm\rangle_{e_1}|\m p\ r a ng le_{e_2} ,\nonu m be r\ \
|\psi^\ pm \ra ngle _ P |\phi ^\pm \ ra ngle_S|+ \rangl e _{ e_ 1}|+\ra ng le_{e_ 2} &\r igh tarro w &|\p si^\pm \rangle_ P|\ph i ^\pm\rangle_S| \ pm\rangle_{e_ 1 }| \ m p\ r angl e_{ e_2},\nonum ber\ \
|\p si^\ p m\ ran g le_P| \psi^ \p m \r a ngle_S|+\rangle_{e_ 1} |+\ran gle_{ e_2}&\rightar row&|\psi^ \ p m \rangle_ P|\p s i^ \ pm\rangle_S|\p m\ran gle_{e_1}| \ mp\rangl e_{e_ 2}.\;\;\ ;\;\;\;\; \ ; \end{ali gne d}$ $ T hen t he excess elect r o n sp in s $e_1$ an d $e_2$ ar e m eas ure din the or thogonal b as is $ \{| +\ran g le_e, |- \r ang le _e\ }$. I f the s tateof t he e x ces s elect r on s pin$e _1 $ is $| +\ rangl e_{e _ 1}$ , the r elative p has e ofth epolariz ation state i s$0$. If th esta te oft h e excess electron spin $e_1$ is $|-\ran gle _{e_1 }$,the relat ive phase of the | e_2$ in_QD$_2$ are_prepared in $|+\rangle_{e_1}$ and_$|+\rangle_{e_2}$, respectively._Here_$|\pm\rangle_e=\frac{1}{\sqrt{2}}(|\uparrow\rangle\pm|\downarrow\rangle)_e$.
After the_two_photons $A$ and_$B$ in the_hyperentangled Bell state pass_through the quantum_circuit_shown in Fig. \[figure3.10\]a in sequence, the state of the system $ABe_1e_2$ evolves to $$\begin{aligned}
__ ___ _ _ _ __%eq.11 _ % Eq. 60
|\phi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\phi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\phi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\psi^\pm\rangle_P|\phi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\phi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2},\nonumber\\
|\psi^\pm\rangle_P|\psi^\pm\rangle_S|+\rangle_{e_1}|+\rangle_{e_2}&\rightarrow&|\psi^\pm\rangle_P|\psi^\pm\rangle_S|\pm\rangle_{e_1}|\mp\rangle_{e_2}.\;\;\;\;\;\;\;\;\end{aligned}$$ Then the excess electron_spins $e_1$ and $e_2$ are measured_in the orthogonal_basis_$\{|+\rangle_e,_|-\rangle_e\}$. If the state_of the excess electron spin $e_1$_is $|+\rangle_{e_1}$, the relative phase of_the polarization state is $0$. If the_state of the excess electron spin_$e_1$ is $|-\rangle_{e_1}$, the relative_phase of_the |
(\pi_\ast\Sigma_g, { {\textstyle {V_T\over M}} },
{ {\textstyle {1\over M}} } \pi_\ast B_T+{ {\textstyle {1\over M}} }\check{B}_M\right),$$ where $\check{B}_M\in\Htz\cap \left(\pi_\ast \Htzt^{\Z_M}\right)^\perp$ is a fixed primitive lattice vector with $\no{\check{B}_M}=-2M^2$.
The following application of Prop. \[wendland:ZMorb\] is a helpful exercise, see Sect. \[wendland:SI\]:
Let $a,\,b,\,c\in\Z$ such that $$Q_{a,b,c}:=\left(\begin{array}{cc} 8a&4b\\4b&8c \end{array}\right)$$ is positive definite. Then there is a toroidal SCFT $\TTT_{a,b,c}$ with $\Z_4$ orbifold $\TTT_{a,b,c}/\Z_4$ in $\MMM^{K3}_{SCFT}$ corresponding to a four-plane $x_{a,b,c}\subset\Her$ such that the following holds[^5]: $x_{a,b,c}=\Om^\ast\oplus\mho_{a,b,c}$, where $\Om^\ast,\,\mho_{a,b,c}$ are two-planes in $\Her$, such that $\Om^\ast\cap\Hez$, $\mho_{a,b,c}\cap\Hez$ have rank $2$ and quadratic forms $Q_{\Om^\ast}=\diag(2,2)$ and $Q_{\mho_{a,b,c}}=Q_{a,b,c}$, respectively.
Consider $T=\R^4/\Lambda=E_1\times E_2$ with orthogonal real two-tori $E_k$ at radii $R_k$, $k\in\{1,2\}$. That is, with respect to real Cartesian coordinates $x_1,\dots,x_4$ and the basis $dx_1,\dots,dx_4$ of $H^1(T,\R)$, the $\Z$-dual $\Lambda^\ast\cong H^1(T,\Z)$ of | ( \pi_\ast\Sigma_g, { { \textstyle { V_T\over M } } },
{ { \textstyle { 1\over M } } } \pi_\ast B_T+ { { \textstyle { 1\over M } } } \check{B}_M\right),$$ where $ \check{B}_M\in\Htz\cap \left(\pi_\ast \Htzt^{\Z_M}\right)^\perp$ is a fixed primitive lattice vector with $ \no{\check{B}_M}=-2M^2$.
The following lotion of Prop. \[wendland: ZMorb\ ] is a helpful use, visit Sect. \[wendland: SI\ ]:
Let $ a,\,b,\,c\in\Z$ such that $ $ Q_{a, b, c}:=\left(\begin{array}{cc } 8a&4b\\4b&8c \end{array}\right)$$ is incontrovertible definite. Then there be a toroidal SCFT $ \TTT_{a, b, c}$ with $ \Z_4 $ orbifold $ \TTT_{a, b, c}/\Z_4 $ in $ \MMM^{K3}_{SCFT}$ corresponding to a four - plane $ x_{a, b, c}\subset\Her$ such that the following holds[^5 ]: $ x_{a, b, c}=\Om^\ast\oplus\mho_{a, b, c}$, where $ \Om^\ast,\,\mho_{a, b, c}$ are two - plane in $ \Her$, such that $ \Om^\ast\cap\Hez$, $ \mho_{a, b, c}\cap\Hez$ have rate $ 2 $ and quadratic forms $ Q_{\Om^\ast}=\diag(2,2)$ and $ Q_{\mho_{a, b, c}}=Q_{a, b, c}$, respectively.
Consider $ T=\R^4/\Lambda = E_1\times E_2 $ with orthogonal real two - tori $ E_k$ at radius $ R_k$, $ k\in\{1,2\}$. That is, with respect to real Cartesian coordinate $ x_1,\dots, x_4 $ and the basis $ dx_1,\dots, dx_4 $ of $ H^1(T,\R)$, the $ \Z$-dual $ \Lambda^\ast\cong H^1(T,\Z)$ of | (\pi_\adt\Sigma_g, { {\textstyle {V_T\ovtr M}} },
{ {\textstyle {1\over M}} } \pi_\ast B_T+{ {\textrtyle {1\over M}} }\check{B}_M\right),$$ whxre $\xheck{V}_M\in\Htz\cap \left(\pi_\ast \Hgzt^{\Z_M}\righn)^\perp$ is q fieed primitive lavfice vegcor wjbh $\no{\ehxck{B}_M}=-2M^2$.
The folloeing appliwation of Prop. \[fevdpand:ZMorb\] is a helpful exercise, see Sect. \[wemdpand:SI\]:
Let $a,\,b,\,c\ig\Z$ slcr thzn $$W_{a,b,c}:=\left(\begin{array}{cc} 8a&4b\\4b&8c \end{zrray}\rijht)$$ is positive definite. Then there is a horoldal SCFT $\TTT_{a,b,c}$ wlth $\Z_4$ orbifild $\ERT_{a,b,c}/\Z_4$ in $\MMO^{K3}_{SCFT}$ corgzsponding tk a four-plane $x_{a,b,c}\subset\Her$ sucf thac the folloqibg jmlds[^5]: $x_{a,b,c}=\Om^\est\oplls\mho_{a,b,c}$, wherc $\Om^\ast,\,\kho_{a,b,c}$ sre two-planes ln $\Hec$, suxh that $\Om^\ast\cap\Hez$, $\mio_{a,b,c}\cap\Hez$ have ranh $2$ and quddxatic forms $Q_{\Om^\ast}=\diat(2,2)$ qnd $Q_{\kho_{a,t,c}}=Q_{a,c,x}$, rdsptctmvemy.
Consldec $T=\R^4/\Lambda=E_1\fimes E_2$ wity orthogonal real teo-eiri $E_k$ at radji $R_k$, $h\ig\{1,2\}$. That is, with respect to real Cartesiat ckordinates $x_1,\dots,x_4$ and tye basis $dx_1,\dots,dx_4$ of $J^1(T,\R)$, the $\Z$-qual $\Lambda^\ast\cong H^1(T,\Z)$ of | (\pi_\ast\Sigma_g, { {\textstyle {V_T\over M}} }, { M}} \pi_\ast B_T+{ {1\over M}} }\check{B}_M\right),$$ a primitive lattice vector $\no{\check{B}_M}=-2M^2$. The following of Prop. \[wendland:ZMorb\] is a helpful see Sect. \[wendland:SI\]: Let $a,\,b,\,c\in\Z$ such that $$Q_{a,b,c}:=\left(\begin{array}{cc} 8a&4b\\4b&8c \end{array}\right)$$ is positive definite. there is a toroidal SCFT $\TTT_{a,b,c}$ with $\Z_4$ orbifold $\TTT_{a,b,c}/\Z_4$ in $\MMM^{K3}_{SCFT}$ corresponding a $x_{a,b,c}\subset\Her$ that following holds[^5]: $x_{a,b,c}=\Om^\ast\oplus\mho_{a,b,c}$, where $\Om^\ast,\,\mho_{a,b,c}$ are two-planes in $\Her$, such that $\Om^\ast\cap\Hez$, $\mho_{a,b,c}\cap\Hez$ have rank $2$ quadratic forms $Q_{\Om^\ast}=\diag(2,2)$ and $Q_{\mho_{a,b,c}}=Q_{a,b,c}$, respectively. Consider $T=\R^4/\Lambda=E_1\times with orthogonal real two-tori at radii $R_k$, $k\in\{1,2\}$. That with to real coordinates and basis $dx_1,\dots,dx_4$ of the $\Z$-dual $\Lambda^\ast\cong H^1(T,\Z)$ of | (\pi_\ast\Sigma_g, { {\textstyle {V_T\oveR M}} },
{ {\textstylE {1\over m}} } \pi_\Ast b_T+{ {\TextStylE {1\over M}} }\check{B}_M\rIGht),$$ wHere $\check{B}_M\in\Htz\cap \lefT(\pi_\asT \HTZt^{\Z_M}\RIgHt)^\perP$ is a fixED pRIMitIvE lAttIcE VeCtor wIth $\No{\check{b}_M}=-2M^2$.
The follOwiNg Application oF prOp. \[wendland:zMoRb\] is a helpful ExeRcise, sEe secT. \[WendlAnd:sI\]:
Let $A,\,b,\,c\in\Z$ SUch thaT $$Q_{a,b,c}:=\left(\BeGIn{arraY}{Cc} 8a&4b\\4b&8c \eND{ArRay}\rIght)$$ is positive defINiTE. Then there is a tOroidaL ScfT $\ttt_{a,b,C}$ wiTh $\Z_4$ orbifolD $\TtT_{a,b,c}/\z_4$ In $\MMM^{K3}_{ScfT}$ CORResPOnding to a four-Plane $x_{a,b,c}\suBSet\her$ sucH tHat THe follOwing HoLDs[^5]: $x_{A,b,c}=\Om^\ast\oplUs\mhO_{a,b,c}$, where $\om^\ast,\,\mHO_{a,b,c}$ are TWo-planeS in $\Her$, SucH thAt $\Om^\ASt\CaP\HeZ$, $\mHO_{a,b,C}\CaP\HeZ$ HavE rank $2$ and QuAdRatic FormS $q_{\oM^\Ast}=\dIag(2,2)$ And $Q_{\Mho_{a,b,C}}=Q_{a,b,c}$, respectiVelY.
ConSIdeR $T=\R^4/\LaMbda=E_1\TimeS E_2$ With oRthogoNal reAl Two-tori $E_k$ at radiI $R_k$, $k\In\{1,2\}$. That is, wIth ReSpeCt To reaL cartesIan CooRdinateS $x_1,\dots,x_4$ ANd tHe BASIs $Dx_1,\dots,dx_4$ of $H^1(T,\R)$, the $\Z$-dUaL $\lAmBda^\ast\coNg H^1(T,\Z)$ oF | (\pi_\ast\Sigma_g, { {\tex tstyle {V_ T\ove r M }}},
{{\te xtstyle {1\ove r M}} } \pi_\ast B_T+{ {\te xtsty le {1\o v er M}}}\check { B} _ M \ri gh t) ,$$ w h er e $\c hec k{B}_M\ in\Htz\cap \l ef t(\pi_\ast \ H tz t^{\Z_M}\r igh t)^\perp$ is afixedpr imi t ive l att ice v ectorw ith $\ no{\check {B } _M}=-2 M ^2$.
T h e f ollo wing applicationo fP rop. \[wendlan d:ZMor b\ ] i s a h elp ful exerci se , see Sect. \ [ we n d l and : SI\]:
Let $a ,\,b,\,c\in \ Z$such t ha t $ $ Q_{a,b ,c}:= \l e ft( \begin{arra y}{c c} 8a&4b\ \4b&8c \end{ar r ay}\rig ht)$$ispos itiv e d ef ini te . Th e nthe r e i s a toro id al SCFT $\T T T _ { a,b, c}$ wit h $\Z _4$ orbifold$\T TT_{ a ,b, c}/\Z _4$ i n $\ MM M^{K3 }_{SCF T}$ c or responding to a fou r-plane $ x_{ a, b,c }\ subse t \Her$suc h t hat the follow i ngho l d s [^ 5]: $x_{a,b,c}=\Om ^\ a s t\ oplus\mh o_{a,b , c} $, where $\ Om ^\a st,\ , \ mho_{ a,b, c }$ are two -plane s i n$\Her$, s uch th at $\ Om^ \ast\ c ap\H ez$, $ \mho_{a, b,c}\ c ap\Hez$ have r a nk $2$ and qu a dr a t ic form s $ Q_{\Om^\ast }=\d i ag(2 ,2)$ an d $ Q _{\mh o_{a, b, c }} = Q_{a,b,c}$, respect iv ely.
Consi der $T=\R^4/\ Lambda=E_1 \ t i mes E_2$ wit h o r thogonal realtwo-t ori $E_k$a t radii$R_k$ , $k\in\ {1,2\}$.T h at is, w ith re spe ctt o r eal Cartesian c oord in ates $x _1, \dots,x _4$ an d t heba sis $dx_1 ,\dots,d x_ 4$ o f$H^ 1(T,\ R )$, the$\ Z$- du al$\Lam b da^\as t\con g H^ 1( T, \ Z)$ of | (\pi_\ast\Sigma_g, {_{\textstyle {V_T\over_M}} },
{ {\textstyle_{1\over M}}_}_\pi_\ast B_T+{_{\textstyle_{1\over M}} }\check{B}_M\right),$$_where $\check{B}_M\in\Htz\cap \left(\pi_\ast_\Htzt^{\Z_M}\right)^\perp$ is a fixed_primitive lattice vector_with_$\no{\check{B}_M}=-2M^2$.
The following application of Prop. \[wendland:ZMorb\] is a helpful exercise, see Sect. \[wendland:SI\]:
Let $a,\,b,\,c\in\Z$ such that_$$Q_{a,b,c}:=\left(\begin{array}{cc}_8a&4b\\4b&8c \end{array}\right)$$_is_positive_definite. Then there is a_toroidal SCFT $\TTT_{a,b,c}$ with $\Z_4$_orbifold $\TTT_{a,b,c}/\Z_4$_in $\MMM^{K3}_{SCFT}$ corresponding to a four-plane $x_{a,b,c}\subset\Her$ such_that_the following holds[^5]:_$x_{a,b,c}=\Om^\ast\oplus\mho_{a,b,c}$, where $\Om^\ast,\,\mho_{a,b,c}$ are two-planes in $\Her$, such that_$\Om^\ast\cap\Hez$, $\mho_{a,b,c}\cap\Hez$ have rank $2$ and_quadratic forms $Q_{\Om^\ast}=\diag(2,2)$_and_$Q_{\mho_{a,b,c}}=Q_{a,b,c}$,_respectively.
Consider $T=\R^4/\Lambda=E_1\times E_2$ with_orthogonal real two-tori $E_k$ at radii_$R_k$, $k\in\{1,2\}$. That is, with respect_to real Cartesian coordinates $x_1,\dots,x_4$ and the_basis $dx_1,\dots,dx_4$ of $H^1(T,\R)$, the $\Z$-dual_$\Lambda^\ast\cong H^1(T,\Z)$ of |
{\epsilon\over 2}+ \eta+ {\epsilon\over 2}=\epsilon+ \eta.$$ Clearly, the pervious relation for each $n\in\Bbb Z$ is true. Since $\epsilon>0$ is a small arbitrary number then for each $n\in\Bbb Z$ we have $r\Big(F_{\sigma_n}(x), F_{\sigma_n}(y)\Big)\leq \eta$. Whereof $\mathcal{F}$ is expansive relative to $\sigma$ with constant expansive $\eta$ hence we obtain $x= y$ and so $r(x, y)=0$. So $$\begin{aligned}
\begin{array}{ll}
r\Big(x_{N_i}, y_{N_i}\Big)& < r\Big(x_{N_i}, x\Big)+ r(x, y)+ r\Big(y, y_{N_i}\Big)\\
& = r\Big(x_{N_i}, x\Big)+ r\Big(y, y_{N_i}\Big).
\end{array}\end{aligned}$$ As $i\rightarrow \infty$, on the basis of convergency of the subsequences ${\Big\{x_{N_i}\Big\}}_{i=1}^{\infty}$ and ${\Big\{y_{N_i}\Big\}}_{i=1}^{\infty}$ to $x$ and $y$ respectively, for given $\epsilon>0(\epsilon<\mu)$ there exists $k_3\in\Bbb N$ such that for every $i\geq k_3$, $r\Big(x_{N_i}, x\Big)< {\epsilon\over 2}$ and $r\Big(y_{N_i}, y\Big)< {\epsilon\over 2}$. Thus, for sufficiently large $N_i$ when $i\rightarrow \infty$, the pervious relation will be as follows $$r\Big(x_{N_i}, y_{N_i}\Big)< {\epsilon\over 2}+ {\epsilon\over 2}= \epsilon< \mu$$ This contradicts absurd hypothesis and the assertion is proved.
We say that IFS $\mathcal{F}$ has [***shadowing uniqueness property relative to $\sigma$***]{} if there exists a constant number $\epsilon>0$ such that for every $\delta$-chain $\xi={\{x_k\}}_{k\ | { \epsilon\over 2}+ \eta+ { \epsilon\over 2}=\epsilon+ \eta.$$ Clearly, the pervious relation for each $ n\in\Bbb Z$ is true. Since $ \epsilon>0 $ is a small arbitrary numeral then for each $ n\in\Bbb Z$ we take $ r\Big(F_{\sigma_n}(x), F_{\sigma_n}(y)\Big)\leq \eta$. Whereof $ \mathcal{F}$ is expansive relative to $ \sigma$ with constant expansive $ \eta$ hence we obtain $ x= y$ and then $ r(x, y)=0$. So $ $ \begin{aligned }
\begin{array}{ll }
r\Big(x_{N_i }, y_{N_i}\Big) & < r\Big(x_{N_i }, x\Big)+ r(x, y)+ r\Big(y, y_{N_i}\Big)\\
& = r\Big(x_{N_i }, x\Big)+ r\Big(y, y_{N_i}\Big).
\end{array}\end{aligned}$$ As $ i\rightarrow \infty$, on the basis of convergence of the posteriority $ { \Big\{x_{N_i}\Big\}}_{i=1}^{\infty}$ and $ { \Big\{y_{N_i}\Big\}}_{i=1}^{\infty}$ to $ x$ and $ y$ respectively, for given $ \epsilon>0(\epsilon<\mu)$ there exist $ k_3\in\Bbb N$ such that for every $ i\geq k_3 $, $ r\Big(x_{N_i }, x\Big) < { \epsilon\over 2}$ and $ r\Big(y_{N_i }, y\Big) < { \epsilon\over 2}$. Thus, for sufficiently large $ N_i$ when $ i\rightarrow \infty$, the pervious relation back will be as follows $ $ r\Big(x_{N_i }, y_{N_i}\Big) < { \epsilon\over 2}+ { \epsilon\over 2}= \epsilon < \mu$$ This contradicts absurd guess and the assertion is proved.
We suppose that IFS $ \mathcal{F}$ has [ * * * shadowing singularity property relative to $ \sigma$ * * * ] { } if there exists a ceaseless number $ \epsilon>0 $ such that for every $ \delta$-chain $ \xi={\{x_k\}}_{k\ | {\epsllon\over 2}+ \eta+ {\epsilon\ovev 2}=\epsilon+ \eta.$$ Clgaely, thx pervikus relagion for each $n\in\Bbb Z$ is trne. Sunce $\tisilon>0$ is a small arbktrary nulber theb foc each $n\in\Bbb Z$ xs have $v\Yig(F_{\sjnma_n}(x), H_{\sigma_n}(y)\Big)\leq \gta$. Whereof $\kathcal{F}$ is ex[avsnve relative to $\sigma$ with constant qxpansife $\eta$ hence we jbtapn $x= y$ and so $r(x, y)=0$. So $$\begin{aligned}
\begin{adray}{ll}
r\Uig(x_{N_i}, y_{N_i}\Big)& < t\Big(x_{N_i}, x\Big)+ r(x, y)+ r\Big(y, y_{N_l}\Big)\\
& = t\Gig(v_{B_i}, x\Big)+ r\Big(h, y_{N_i}\Big).
\end{array}\end{alifned}$$ As $i\rightarrow \infty$, on thd basns of convetywncj of the subwequegces ${\Big\{x_{N_i}\Blb\}}_{i=1}^{\inftf}$ and ${\Bog\{y_{N_i}\Big\}}_{i=1}^{\infty}$ to $x$ qnd $y$ respectively, foc given $\epsilon>0(\epsiljn<\mu)$ thera zxists $k_3\in\Bbb N$ such rhqt fot evesy $i\ewq y_3$, $r\Gij(x_{N_j}, x\Big)< {\epailon\over 2}$ and $r\Big(y_{N_u}, y\Big)< {\epsilon\over 2}$. Trlx, for sufficjently lwrge $N_i$ when $i\rightarrow \infty$, the pervpous relation will be as foolows $$r\Big(x_{N_i}, y_{N_i}\Big)< {\gpsilon\ovew 2}+ {\epsilon\over 2}= \epsilon< \mu$$ This contradicts absurg hypkghewif qnf the assertion is proved.
We say that IFS $\mathszl{G}$ mas [***shadowing unlqueness property tepayyve to $\sigma$***]{} kf thexs sxists a constant jumber $\gpsilob>0$ such thwt fpr every $\delta$-chain $\xi={\{x_k\}}_{k\ | {\epsilon\over 2}+ \eta+ {\epsilon\over 2}=\epsilon+ \eta.$$ Clearly, relation each $n\in\Bbb is true. Since number for each $n\in\Bbb we have $r\Big(F_{\sigma_n}(x), \eta$. Whereof $\mathcal{F}$ is expansive relative $\sigma$ with constant expansive $\eta$ hence we obtain $x= y$ and so $r(x, So $$\begin{aligned} \begin{array}{ll} r\Big(x_{N_i}, y_{N_i}\Big)& < r\Big(x_{N_i}, x\Big)+ r(x, y)+ r\Big(y, y_{N_i}\Big)\\ & r\Big(x_{N_i}, r\Big(y, \end{array}\end{aligned}$$ $i\rightarrow \infty$, on the basis of convergency of the subsequences ${\Big\{x_{N_i}\Big\}}_{i=1}^{\infty}$ and ${\Big\{y_{N_i}\Big\}}_{i=1}^{\infty}$ to $x$ and respectively, for given $\epsilon>0(\epsilon<\mu)$ there exists $k_3\in\Bbb N$ that for every $i\geq $r\Big(x_{N_i}, x\Big)< {\epsilon\over 2}$ and y\Big)< 2}$. Thus, sufficiently $N_i$ $i\rightarrow \infty$, the relation will be as follows $$r\Big(x_{N_i}, y_{N_i}\Big)< {\epsilon\over 2}+ {\epsilon\over 2}= \epsilon< \mu$$ This contradicts absurd hypothesis the assertion We say IFS has uniqueness property relative if there exists a constant number for every $\delta$-chain $\xi={\{x_k\}}_{k\ | {\epsilon\over 2}+ \eta+ {\epsilon\over 2}=\Epsilon+ \eta.$$ clearLy, tHe pErViouS relAtion for each $n\iN\bbb Z$ Is true. Since $\epsilon>0$ is a sMall aRbITrarY NuMber tHen for eACh $N\IN\BbB Z$ We HavE $r\bIg(f_{\sigmA_n}(x), f_{\sigma_n}(Y)\Big)\leq \eta$. wheReOf $\mathcal{F}$ is EXpAnsive relaTivE to $\sigma$ with ConStant eXpAnsIVe $\eta$ HenCe we oBtain $x= Y$ And so $r(X, y)=0$. So $$\begin{AlIGned}
\beGIn{array}{LL}
R\BIg(x_{N_I}, y_{N_i}\Big)& < r\Big(x_{N_i}, x\BiG)+ R(x, Y)+ R\Big(y, y_{N_i}\Big)\\
& = r\BiG(x_{N_i}, x\BIg)+ R\biG(Y, Y_{N_i}\big).
\End{array}\enD{aLigneD}$$ as $i\righTArROW \InfTY$, on the basis of Convergency OF thE subseQuEncES ${\Big\{x_{N_I}\Big\}}_{i=1}^{\InFTy}$ aNd ${\Big\{y_{N_i}\Big\}}_{I=1}^{\infTy}$ to $x$ and $y$ RespecTIvely, foR Given $\epSilon>0(\ePsiLon<\Mu)$ thERe ExIstS $k_3\IN\BbB n$ sUch THat For every $I\gEq K_3$, $r\Big(X_{N_i}, x\bIG)< {\EPsilOn\oVer 2}$ aNd $r\BiG(y_{N_i}, y\Big)< {\epsilOn\oVer 2}$. THUs, fOr sufFicieNtly LaRge $N_i$ When $i\rIghtaRrOw \infty$, the perviOus rElation wiLl bE aS foLlOws $$r\BIG(x_{N_i}, y_{N_I}\BiG)< {\epSilon\ovEr 2}+ {\epsilON\ovEr 2}= \EPSIlOn< \mu$$ This contradictS aBSUrD hypotheSis and THe AsSErtion is PrOveD.
We sAY That IfS $\maTHcAl{F}$ has [***shAdowinG UnIqUeness pRoPerty rElAtiVe tO $\sigmA$***]{} If thEre exiSts a consTant nUMber $\epsilon>0$ sucH That for every $\dELtA$-CHaIN $\xi={\{x_K\}}_{k\ | {\epsilon\over 2}+ \eta+ { \epsilon\o ver 2 }=\ eps il on+\eta .$$ Clearly, t h e pe rvious relation for ea ch $n \i n \Bbb Z$ is t rue. Si n ce $ \ep si lo n>0 $i sa sma llarbitra ry numberthe nfor each $n\ i n\ Bbb Z$ wehav e $r\Big(F_{ \si gma_n} (x ),F _{\si gma _n}(y )\Big) \ leq \e ta$. Wher eo f $\mat h cal{F}$ i sexpa nsive relative to $\ s igma$ with con stantex p an s i ve$\e ta$ hencewe obta i n $x= y $ a n d so$ r(x, y)=0$. S o $$\begin{ a lig ned}
\ be gin { array} {ll}r\ B ig( x_{N_i}, y_ {N_i }\Big)& < r\Big ( x_{N_i} , x\Big) + r(x, y) + r \Big ( y, y _{N _i } \Bi g )\ \
& = r\B ig( x_{N _i},x\Big)+ r\Big (y, y_{ N _i} \Big) .
\en d{ar ra y}\en d{alig ned}$ $As $i\rightarro w \i nfty$, on th ebas is of c o nverge ncy of the su bsequen c es${ \ B i g\ {x_{N_i}\Big\}}_{i =1 } ^ {\ infty}$and ${ \ Bi g\ { y_{N_i}\ Bi g\} }_{i = 1 }^{\i nfty } $to $x$ a nd $y$ re sp ectivel y, for g iv en$\e psilo n >0(\ epsilo n<\mu)$there exists $k_3\in \ Bbb N$ such t h at f or ever y $ i\geq k_3$, $r\ B ig(x _{N_ i }, x\ B ig)<{\eps il o n\ o ver 2}$ and $r\Big( y_ {N_i}, y\Bi g)< {\epsilon \over 2}$. T h us, forsuff i ci e ntly large $N_ i$ wh en $i\righ t arrow \i nfty$ , the pe rvious re l a tion wil l b e a s f oll o w s$$r\Big(x_{N_ i } , y_ {N _i}\Big )<{\epsil on\ ove r 2 }+{\ epsilon\o ver 2}=\e ps il on < \ mu$$T his cont ra dic ts ab surdh ypothe sis a nd t he a s ser tion is pr o v ed.
W esaytha tIFS $ \mat h cal {F}$ ha s [***sha dow i ng u ni qu eness p roperty relat iv e to $\sig ma $** *]{} i f there ex ists a constant number$ \epsilo n>0 $ suc h th at for ev ery $\del ta$ - chain$\xi={ \{x_k \} }_{ k \ | {\epsilon\over 2}+_\eta+ {\epsilon\over_2}=\epsilon+ \eta.$$ Clearly, the_pervious relation_for_each $n\in\Bbb_Z$_is true. Since_$\epsilon>0$ is a_small arbitrary number then_for each $n\in\Bbb_Z$_we have $r\Big(F_{\sigma_n}(x), F_{\sigma_n}(y)\Big)\leq \eta$. Whereof $\mathcal{F}$ is expansive relative to $\sigma$ with constant_expansive_$\eta$ hence_we_obtain_$x= y$ and so $r(x,_y)=0$. So $$\begin{aligned}
\begin{array}{ll}
r\Big(x_{N_i}, y_{N_i}\Big)& <_r\Big(x_{N_i}, x\Big)+_r(x, y)+ r\Big(y, y_{N_i}\Big)\\
__ _ _ _ _&_=_r\Big(x_{N_i}, x\Big)+ r\Big(y, y_{N_i}\Big).
\end{array}\end{aligned}$$_As $i\rightarrow \infty$, on the basis_of convergency of the subsequences ${\Big\{x_{N_i}\Big\}}_{i=1}^{\infty}$_and ${\Big\{y_{N_i}\Big\}}_{i=1}^{\infty}$ to $x$ and $y$ respectively,_for given $\epsilon>0(\epsilon<\mu)$ there exists $k_3\in\Bbb_N$ such that for every_$i\geq k_3$,_$r\Big(x_{N_i}, x\Big)< {\epsilon\over 2}$ and_$r\Big(y_{N_i}, y\Big)< {\epsilon\over_2}$. Thus,_for sufficiently large_$N_i$ when $i\rightarrow \infty$, the pervious_relation will be_as follows $$r\Big(x_{N_i}, y_{N_i}\Big)< {\epsilon\over 2}+_{\epsilon\over_2}= \epsilon< \mu$$_This_contradicts_absurd hypothesis_and the assertion_is_proved.
We say_that_IFS $\mathcal{F}$ has [***shadowing uniqueness property_relative_to $\sigma$***]{} if there exists a constant_number $\epsilon>0$ such that_for_every $\delta$-chain $\xi={\{x_k\}}_{k\ |
j_0}}\Bigr) \mbox { as $$} \nonumber \\
&= \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\psi_o\psi_k\Bigr) \mbox { as $\psi_0=1$} \nonumber \\
&=0 \mbox{ due to orthogonality of the Fourier basis functions} \nonumber\end{aligned}$$ When n wild cards are present, the extension is straightforward and results in: $$\begin{aligned}
\sum_{x \in S}\psi_j(x)&= \exp \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lambda_n} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_l}k}{\lambda_l} \Bigr)\Bigr) \\
\sum_{x \in S}\psi_j(x)&= \exp \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lambda_n} \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\psi_0\psi_k\Bigr) \nonumber \\
%\Bigl(1+\sum_{k=m}^{d-1}\Bigl(\exp \frac{2{\pi}i{j_0}k}{\lambda_0} \Bigr)\Bigr) \nonumber \\
&=0 \mbox{ due to orthogonality of Fourier basis functions } \nonumber\end{aligned}$$ This optimization significantly increases processing speed, especially in high | j_0}}\Bigr) \mbox { as $ $ } \nonumber \\
& = \exp \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1 } \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1 } } { \lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0 - 1}\Bigl(\psi_o\psi_k\Bigr) \mbox { as $ \psi_0=1 $ } \nonumber \\
& = 0 \mbox { due to orthogonality of the Fourier basis functions } \nonumber\end{aligned}$$ When n wild card are present, the elongation is straightforward and results in: $ $ \begin{aligned }
\sum_{x \in S}\psi_j(x)&= \exp \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lambda_n } \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1 } } { \lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_l}k}{\lambda_l } \Bigr)\Bigr) \\
\sum_{x \in S}\psi_j(x)&= \exp \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lambda_n } \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1 } } { \lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\psi_0\psi_k\Bigr) \nonumber \\
% \Bigl(1+\sum_{k = m}^{d-1}\Bigl(\exp \frac{2{\pi}i{j_0}k}{\lambda_0 } \Bigr)\Bigr) \nonumber \\
& = 0 \mbox { due to orthogonality of Fourier footing functions } \nonumber\end{aligned}$$ This optimization significantly increases march speed, especially in eminent | j_0}}\Bihr) \mbox { as $$} \nonumber \\
&= \exp \Bigl(\frac{2{\pi}n{h_1}{x_1}}{\lambva_1} \Bigr) \times \caots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}e_{d-1}} {\lqmbda_{e-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\psi_u\psi_k\Bigr) \mbox { aw $\psm_0=1$} \nonumber \\
&=0 \mbox{ due to orthogkkalitv if the Fourier basis funwtions} \nonumbes\evd{cligned}$$ When n wild cards are presene, the ectfnsion is strayghtgjrwadd and results in: $$\begin{aligned}
\sum_{x \in S}\psp_j(x)&= \exp \Bigl(\frac{2{\po}i{j_{n}}{x_{n}}}{\lambda_n} \Bigr) \times \cfots \times \exp \Bigl(\fraf{2{\pi}i{j_{d-1}}x_{d-1}} {\lanbda_{q-1}}\Vigr) \prod_{l=0}^{n-1}\rum_{k=0}^{\lambda_l-1}\Bigl(\exp \Bigm(\frac{2{\pi}i{j_l}k}{\lambda_l} \Bigr)\Bigr) \\
\rum_{x \nn S}\psi_j(x)&= \ezp \Bihn(\frac{2{\pi}i{j_{n}}{x_{i}}}{\lambdw_n} \Bigr) \times \cdots \dimes \ecp \Bigl(\frac{2{\pi}i{m_{d-1}}x_{d-1}} {\namvda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lembda_l-1}\Bigl(\psi_0\psi_k\Bigt) \nonumber \\
%\Yigl(1+\sum_{k=m}^{d-1}\Bigl(\exp \frax{2{\pu}i{j_0}k}{\ldmbdd_0} \Biee)\Bier) \ionhmber \\
&=0 \mbkx{ due to krthogonaliry of Fourier basis ftbctions } \nonujber\enq{ajigned}$$ This optimization significantly pncrsases processing speed, wspecially in high | j_0}}\Bigr) \mbox { as $$} \nonumber \\ \Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1} \times \cdots \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\psi_o\psi_k\Bigr) \\ \mbox{ due to of the Fourier functions} \nonumber\end{aligned}$$ When n wild cards present, the extension is straightforward and results in: $$\begin{aligned} \sum_{x \in S}\psi_j(x)&= \exp \Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_l}k}{\lambda_l} \Bigr)\Bigr) \\ \sum_{x \in \exp \Bigr) \cdots \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\psi_0\psi_k\Bigr) \nonumber \\ %\Bigl(1+\sum_{k=m}^{d-1}\Bigl(\exp \frac{2{\pi}i{j_0}k}{\lambda_0} \Bigr)\Bigr) \nonumber \\ &=0 \mbox{ due to orthogonality Fourier basis functions } \nonumber\end{aligned}$$ This optimization significantly processing speed, especially in | j_0}}\Bigr) \mbox { as $$} \nonumber \\
&= \exp \BigL(\frac{2{\pi}i{j_1}{x_1}}{\LambdA_1} \BiGr) \tImEs \cdOts \tImes \exp \Bigl(\fraC{2{\Pi}i{j_{D-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambDa_0-1}\BigL(\pSI_o\psI_K\BIgr) \mbOx { as $\psi_0=1$} \NOnUMBer \\
&=0 \MbOx{ Due To ORtHogonAliTy of the fourier basIs fUnCtions} \nonumbER\eNd{aligned}$$ WHen N wild cards arE prEsent, tHe ExtENsion Is sTraigHtforwARd and rEsults in: $$\bEgIN{alignED}
\sum_{x \in s}\PSi_J(x)&= \exP \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lAMbDA_n} \Bigr) \times \cdoTs \timeS \eXP \BIGL(\frAc{2{\pI}i{j_{d-1}}x_{d-1}} {\lambDa_{D-1}}\Bigr) \PRod_{l=0}^{n-1}\suM_{K=0}^{\lAMBDa_l-1}\bIgl(\exp \Bigl(\fraC{2{\pi}i{j_l}k}{\lambDA_l} \BIgr)\BigR) \\
\sUm_{x \IN S}\psi_j(X)&= \exp \BIgL(\FraC{2{\pi}i{j_{n}}{x_{n}}}{\lamBda_n} \bigr) \times \Cdots \tIMes \exp \BIGl(\frac{2{\pI}i{j_{d-1}}x_{d-1}} {\LamBda_{D-1}}\BigR) \PrOd_{L=0}^{n-1}\sUm_{K=0}^{\LamBDa_L-1}\BiGL(\psI_0\psi_k\BigR) \nOnUmber \\
%\bigl(1+\SUM_{K=M}^{d-1}\BiGl(\eXp \frAc{2{\pi}i{J_0}k}{\lambda_0} \Bigr)\BIgr) \NonuMBer \\
&=0 \Mbox{ dUe to oRthoGoNalitY of FouRier bAsIs functions } \nonuMber\End{aligneD}$$ ThIs OptImIzatiON signiFicAntLy increAses proCEssInG SPEeD, especially in high | j_0}}\Bigr) \mbox { as $$ } \nonumbe r \\
&= \e xp \Bi gl(\ frac{2{\pi}i{j _ 1}{x _1}}{\lambda_1} \Bigr) \tim es \cdo t s\time s \exp\ Bi g l (\f ra c{ 2{\ pi } i{ j_{d- 1}} x_{d-1} } {\lambda _{d -1 }}\Bigr)\sum _ {k =0}^{\lamb da_ 0-1}\Bigl(\p si_ o\psi_ k\ Big r ) \mb ox{ as$\psi_ 0 =1$} \ nonumber\\ &=0 \m b ox{ due t oorth ogonality of theF ou r ier basis func tions} \ n on u m ber \en d{aligned} $$ When n wildc ar d s are present, theextension i s st raight fo rwa r d andresul ts in: $$\begin{a lign ed}
\sum_ {x \in S}\psi_ j (x)&= \ exp \B igl (\f rac{ 2 {\ pi }i{ j_ { n}} { x_ {n} } }{\ lambda_n }\B igr)\tim e s \ cdot s \ time s \ex p \Bigl(\frac {2{ \pi} i {j_ {d-1} }x_{d -1}} { \lamb da_{d- 1}}\B ig r) \prod_{l=0} ^{n- 1}\sum_{k =0} ^{ \la mb da_l- 1 }\Bigl (\e xp\Bigl(\ frac{2{ \ pi} i{ j _ l }k }{\lambda_l} \Bigr )\ B i gr ) \\
\sum_{ x \ in S}\psi_j (x )&= \ex p \Bigl (\fr a c{ 2{\pi}i{ j_{n}} { x_ {n }}}{\la mb da_n}\B igr ) \ times \cdo ts \ti mes \exp \Big l (\frac{2{\pi}i { j_{d-1}}x_{d- 1 }} { \l a mbda _{d -1}}\Bigr) \pr o d_{l =0}^ { n- 1}\ s um_{k =0}^{ \l a mb d a_l-1}\Bigl(\psi_0\ ps i_k\Bi gr) \ nonumber \\ %\Bigl(1+\ s u m _{k=m}^{ d-1} \ Bi g l(\exp \frac{2 {\pi} i{j_0}k}{\ l ambda_0} \Big r)\Bigr) \nonum b e r \\
&=0 \m box { d uet o o rthogonalityo f Fou ri er basi s f unction s } \n onu mbe r\ end{align ed}$$ Th is o pt im iza tions ignifica nt lyin cre asesp rocess ing s peed ,es p eci ally in hi g h | j_0}}\Bigr) _\mbox {_as $$} \nonumber \\_
&= \exp_\Bigl(\frac{2{\pi}i{j_1}{x_1}}{\lambda_1}_\Bigr) \times_\cdots_\times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}}_{\lambda_{d-1}}\Bigr)\sum_{k=0}^{\lambda_0-1}\Bigl(\psi_o\psi_k\Bigr) \mbox {_as $\psi_0=1$} \nonumber \\
&=0_\mbox{ due to_orthogonality_of the Fourier basis functions} \nonumber\end{aligned}$$ When n wild cards are present, the extension_is_straightforward and_results_in:_$$\begin{aligned}
\sum_{x \in S}\psi_j(x)&= \exp \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lambda_n}_\Bigr) \times \cdots \times \exp_\Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr)_ \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\exp \Bigl(\frac{2{\pi}i{j_l}k}{\lambda_l} \Bigr)\Bigr) \\
\sum_{x_\in_S}\psi_j(x)&= \exp \Bigl(\frac{2{\pi}i{j_{n}}{x_{n}}}{\lambda_n}_\Bigr) \times \cdots \times \exp \Bigl(\frac{2{\pi}i{j_{d-1}}x_{d-1}} {\lambda_{d-1}}\Bigr) \prod_{l=0}^{n-1}\sum_{k=0}^{\lambda_l-1}\Bigl(\psi_0\psi_k\Bigr)_\nonumber \\
%\Bigl(1+\sum_{k=m}^{d-1}\Bigl(\exp \frac{2{\pi}i{j_0}k}{\lambda_0} \Bigr)\Bigr) _ \nonumber \\
&=0_\mbox{_due_to orthogonality of Fourier_basis functions } \nonumber\end{aligned}$$ This optimization_significantly increases processing speed, especially in_high |
toward this X-ray filament, our qualitative analysis suggests that this filament is also matched by the galaxy distribution.
Finally, we consider the structure to the north. In contrast to the southeastern filament, here our spectroscopic coverage extends well to the north of the northern border of the X-ray field of view. Considering only the distribution of spectroscopically confirmed cluster galaxies (shown in the left panel of Figure \[fig:specs\_on\_x\]) without regard to their distribution of brightness, a linear feature can be seen extending to the northeast from the cluster center in the direction highlighted by @2014MNRAS.437.1909M. There are few galaxies on the northern filament identified by @2013HEAD...1340101V, although we note that we identified several galaxies in Figure \[fig:cmd\_spec\] that are along this filament and have the photometry of cluster members. In the top left panel of Figure \[fig:xray\_crosscor\], we show spectroscopic members and probable members brighter than $r = 19.5$ (our 50% completeness limit, as discussed in Appendix \[sec:spec\_comp\]). In this view, the northern galaxy distribution bends to the northwest, slightly above the filament detected by @2013HEAD...1340101V. From the optical data alone we can see that there is some structure to the north, but it is difficult to definitively constrain the path of that structure.
To provide a quantitative measure of how well the galaxy population aligns with the X-ray emission, we consider a cross-correlation between the X-ray flux and galaxy positions. If the galaxy distribution maps out the underlying distribution of X-ray emitting gas, we would expect that the cross-correlation would be peaked in the current alignment and would decrease as we move the galaxies. In particular, by rotating the galaxies around the cluster’s center, the amplitude of the cross-correlation should decrease (although a small bounce-back is expected as filaments in the galaxy distribution rotate onto the next X-ray filament). To test this prediction, we consider this formula for the galaxy X-ray cross-correlation: $${\cal C}_{g \times X} (\theta) = \frac{\sum\limits_{i}^n X_{x_i,y_i}(\theta) \times W_i}{\sum\limits_{i}^n W_i}.$$ Here, $X_{x_i,y_i}(\theta)$ is the X-ray flux at the position of galaxy $i$ | toward this X - ray filament, our qualitative analysis suggest that this fibril is also matched by the galax distribution.
Finally, we regard the structure to the union. In contrast to the southeastern filament, here our spectroscopic coverage extend well to the north of the northern molding of the X - ray field of scene. Considering merely the distribution of spectroscopically confirmed cluster galaxy (shown in the left panel of Figure \[fig: specs\_on\_x\ ]) without esteem to their distribution of brightness, a linear feature can be seen extend to the northeast from the cluster center in the direction highlighted by @2014MNRAS.437.1909M. There are few galaxy on the northern filament identified by @2013HEAD... 1340101V, although we note that we identified several galaxies in Figure \[fig: cmd\_spec\ ] that are along this filament and have the photometry of cluster members. In the top left panel of Figure \[fig: xray\_crosscor\ ], we show spectroscopic members and probable members brighter than $ roentgen = 19.5 $ (our 50% completeness limit, as discussed in Appendix \[sec: spec\_comp\ ]). In this opinion, the northerly galaxy distribution bends to the northwest, slenderly above the filament detect by @2013HEAD... 1340101V. From the optical data alone we can determine that there is some structure to the north, but it is difficult to definitively constrain the path of that structure.
To provide a quantitative measure of how well the galaxy population aligns with the X - re emission, we view a cross - correlation between the X - ray flux and galax positions. If the galaxy distribution maps out the underlying distribution of X - ray emitting gas, we would expect that the cross - correlation coefficient would be peaked in the current alignment and would decrease as we move the galaxies. In particular, by revolve the galaxies around the bunch ’s center, the amplitude of the cross - correlation should decrease (although a small leap - back is expected as filaments in the galax distribution rotate onto the next X - ray filament). To test this prediction, we consider this formula for the galaxy X - ray cross - correlation coefficient: $ $ { \cal C}_{g \times X } (\theta) = \frac{\sum\limits_{i}^n X_{x_i, y_i}(\theta) \times W_i}{\sum\limits_{i}^n W_i}.$$ Here, $ X_{x_i, y_i}(\theta)$ is the X - ray magnetic field at the position of galaxy $ i$ | toaard this X-ray filament, uur qualitative analysms suggssts thag this filament is also matcied vy tht galaxy distributiun.
Finally, we consuder rhe structnde to tmz norfm. In eoitrast to the sputheastert filament, hera uux spectroscopic coverage extends welj to thr jorth of the njrthtrn borsvr of the X-ray field of view. Cohsideriig only the disyribution of spectroscopicwlly confirmed cluster galaxies (syown un the left oanel of Fpyure \[fig:speds\_on\_x\]) without regard to their dkstriyution of btnthtjgss, a linear featlre can be secm extetding tp the northeasb frok tye cluster center in vhe direction highlidhted by @2014KNXAS.437.1909M. There are few gaoazies mn tve nuethdrn fmlajent ifenvified by @2013HSAD...1340101V, althouth we note that we odqbtified severzl galwxyes in Figure \[fig:cmd\_spec\] that are along thjs filament and have thw photometry of clustgr members. In the top left panel of Figure \[fig:xray\_crosscor\], fe shkd skegbrosziplc members and probable members brighter than $d = 19.5$ (pur 50% completencss limit, as discuxsfd yn Appendix \[sgc:spec\_comp\]). In this view, the jortherg galqxy distrybutoon bends to the northwest, wlightly abone tye filament detectzd by @2013HEAD...1340101V. Yrom tne opyical data alone we can see that there is some agructure to the vornh, bgt it is difficult to defigitively ronstxain the patn of trat structkre.
To provide a quantitahive ledsure of hlw well the galaxy population aligns with the X-say emissiou, we cpnsider a crofs-correlation yetween che X-rxy flux ans galaxb positions. Yf the galaxy fistribution maps oue thw uneerlyine distribution pf X-ray emitting gaw, we would expect bhat jhs cross-correlatnub would be peakrd kn ehv cnrrene alignment atd wuula decrdase as we nuve yhe galaxies. In parthculzr, by rotating the gwlaxies qround tre cluster’s crnter, the amplitudt of tie crors-cortelwtion should decrease (although a small houkce-back is ex[ectcd af filamentx in the galaxy distribution rotate onti the next X-ray fioament). To test this krediction, we consyder this formula for the galqxy X-ray cross-corvelation: $${\cal C}_{g \times S} (\thetd) = \frwc{\sum\limits_{i}^n X_{x_i,y_i}(\theta) \times W_i}{\sum\limits_{i}^n W_i}.$$ Here, $X_{x_i,y_i}(\theta)$ is the X-rqy flue wt the posifion of gdlcxy $i$ | toward this X-ray filament, our qualitative analysis this is also by the galaxy structure the north. In to the southeastern here our spectroscopic coverage extends well the north of the northern border of the X-ray field of view. Considering the distribution of spectroscopically confirmed cluster galaxies (shown in the left panel of \[fig:specs\_on\_x\]) regard their of brightness, a linear feature can be seen extending to the northeast from the cluster center the direction highlighted by @2014MNRAS.437.1909M. There are few on the northern filament by @2013HEAD...1340101V, although we note we several galaxies Figure that along this filament have the photometry of cluster members. In the top left panel of Figure \[fig:xray\_crosscor\], we show spectroscopic and probable than $r 19.5$ 50% limit, as discussed \[sec:spec\_comp\]). In this view, the northern to the northwest, slightly above the filament detected @2013HEAD...1340101V. From optical data alone we can see there is some structure to the north, but is difficult to definitively constrain the path of that structure. To provide a quantitative measure well the galaxy population with the X-ray we a between X-ray flux galaxy positions. If the galaxy distribution maps out the underlying distribution X-ray emitting gas, we would expect that the cross-correlation would in current alignment and decrease as we move galaxies. particular, by rotating the the center, the should (although a small bounce-back expected as filaments in the distribution rotate onto the this prediction, we consider this formula for the X-ray cross-correlation: $${\cal C}_{g \times X} (\theta) \frac{\sum\limits_{i}^n X_{x_i,y_i}(\theta) \times W_i}{\sum\limits_{i}^n W_i}.$$ Here, $X_{x_i,y_i}(\theta)$ is the X-ray flux at position of | toward this X-ray filament, our QualitativE analYsiS suGgEsts That This filament is ALso mAtched by the galaxy distrIbutiOn.
fInalLY, wE consIder the STrUCTurE tO tHe nOrTH. IN contRasT to the sOutheasterN fiLaMent, here our sPEcTroscopic cOveRage extends wEll To the nOrTh oF The noRthErn boRder of THe X-ray Field of viEw. cOnsideRIng only THE dIstrIbution of spectrosCOpICally confirmed ClusteR gALaXIEs (sHowN in the left PaNel of fIgure \[fiG:SpECS\_On\_x\]) WIthout regard tO their distrIButIon of bRiGhtNEss, a liNear fEaTUre Can be seen exTendIng to the nOrtheaST from thE Cluster Center In tHe dIrecTIoN hIghLiGHteD By @2014mNRas.437.1909M. THere are fEw GaLaxieS on tHE NORtheRn fIlamEnt idEntified by @2013HEAd...1340101V, aLthoUGh wE note That wE ideNtIfied SeveraL galaXiEs in Figure \[fig:cmD\_speC\] that are aLonG tHis FiLamenT And havE thE phOtometrY of clusTEr mEmBERS. IN the top left panel of fiGURe \[Fig:xray\_cRosscoR\], We ShOW spectroScOpiC memBERs and ProbABlE members BrightER tHaN $r = 19.5$ (our 50% coMpLeteneSs LimIt, aS discUSsed In AppeNdix \[sec:sPec\_coMP\]). In this view, the NOrthern galaxy DIsTRIbUTion BenDs to the nortHwesT, SligHtly ABoVe tHE filaMent dEtECtED by @2013HEAD...1340101V. From the optiCaL data aLone wE can see that thEre is some sTRUCture to tHe noRTh, BUt it is difficulT to deFinitively COnstrain The paTh of that Structure.
tO Provide a QuaNtiTatIve MEAsUre of how well tHE GalaXy PopulatIon Aligns wIth The x-raY emIsSion, we conSider a crOsS-cOrReLatIon beTWeen the X-RaY flUx And GalaxY PositiOns. If The gAlAxY DisTributiON mAPS out ThE uNderLyiNg DistrIbutIOn oF X-ray emItting gas, We wOUld eXpEcT that thE cross-correlaTiOn would be pEaKed In the cURRent aligNment and would decrease as WE move thE gaLaxieS. In pArticular, By rOtatinG thE GalaxiEs arouNd the ClUstER’S centER, ThE amPlItude of the CROss-CorreLaTion Should dEcrease (although a smALl bOunce-back is exPecTed aS FIlAmeNTs IN thE gALaxY DIstribution rotaTe onto the nExT x-rAy filament). tO teSt This preDiction, We conSIder thiS formula fOr the galaXy x-ray CROss-CorrelatioN: $${\cal C}_{g \tiMes X} (\theta) = \FRac{\suM\LiMits_{i}^N X_{x_I,y_i}(\theTa) \TimEs W_i}{\sUm\limiTS_{i}^n w_i}.$$ HerE, $X_{x_i,y_i}(\ThEta)$ is tHe X-raY fLux at the Position of galaxy $i$ | toward this X-ray filamen t, our qua litat ive an al ysis sug gests that thi s fil ament is also matchedby th eg alax y d istri bution.
F i n all y, w e c on s id er th e s tructur e to the n ort h. In contrast to the south eas tern filamen t,here o ur sp e ctros cop ic co verage extend s well to t h e nort h of the n or ther n border of the X - ra y field of view . Cons id e ri n g on lythe distri bu tiono f spect r os c o p ica l ly confirmedcluster gal a xie s (sho wn in the le ft pa ne l of Figure \[f ig:s pecs\_on\ _x\])w ithoutr egard t o thei r d ist ribu t io nofbr i ght n es s,a li near fea tu re canbe s e e n exte ndi ng t o the northeast fr omthec lus ter c enter inth e dir ection high li ghted by @2014M NRAS .437.1909 M.Th ere a re fe w galax ies on the no rthernf ila me n t id entified by @2013H EA D . .. 1340101V , alth o ug hw e note t ha t w e id e n tifie d se v er al galax ies in Fi gu re \[fi g: cmd\_s pe c\] th at ar e alo ng thi s filame nt an d have the phot o metry of clus t er m em b ers. In the top le ft p a nelof F i gu re\ [fig: xray\ _c r os s cor\], we show spec tr oscopi c mem bers and prob able membe r s brighter tha n $ r = 19.5$ (our50% c ompletenes s limit,as di scussedin Append i x \[sec:s pec \_c omp \]) . In this view, t h e nor th ern gal axy distri but ion be nds t o the nor thwest,sl ig ht ly ab ove t h e filame nt de te cte d by@ 2013HE AD... 1340 10 1V . Fr om theo pt i c al d at aalon e w ecan s ee t h atthere i s some st ruc t ureto t he nort h, but it isdi fficult to d efi nitive l y constra in the path of that str u cture.
To prov idea quantit ati ve mea sur e of ho w well thega lax y popul a t io n a li gns with t h e X- ray e mi ssio n, we c onsider a cross-co r rel ation between th e X- r a yflu x a n d g al a xyp o sitions. If the galaxy di st r ib ution maps out t he unde rlyingdistr i butionof X-rayemittingga s, w e wou ld expectthat the cross-co r relat i on woul d b e peak ed in thecurren t al ignme nt and w ould d ecrea se as we m ove the galaxies. In pa rticul ar, b y r otating t heg ala xies arou nd t he cluster ’scen ter,the ampli tude of th e cros s-co r relations ho uld d ec rease (alth o u g h a smal l b o unce-b ackis expected as fi l aments in thegala x y di str i buti on rotate onto t hene x t X-ray f il ament). Totest thi sp redic tion,we con sider t h i sf ormula for th e galaxyX-r ay cross-c or re l ation: $${ \c al C}_ {g \ti m es X } (\theta) = \frac {\sum \ l imits _ {i} ^n X_ {x _i,y_i} ( \the ta) \times W_i}{\sum\ limits _{i} ^n W_ i}.$$ H er e, $X_ {x_ i, y_i}(\thet a )$ is the X-ra y fluxat the po sition ofg a laxy$i$ | toward_this X-ray_filament, our qualitative analysis_suggests that_this_filament is_also_matched by the_galaxy distribution.
Finally, we_consider the structure to_the north. In_contrast_to the southeastern filament, here our spectroscopic coverage extends well to the north of_the_northern border_of_the_X-ray field of view. Considering_only the distribution of spectroscopically_confirmed cluster_galaxies (shown in the left panel of Figure_\[fig:specs\_on\_x\])_without regard to_their distribution of brightness, a linear feature can be_seen extending to the northeast from_the cluster center_in_the_direction highlighted by @2014MNRAS.437.1909M._There are few galaxies on the_northern filament identified by @2013HEAD...1340101V, although_we note that we identified several galaxies_in Figure \[fig:cmd\_spec\] that are along_this filament and have the_photometry of_cluster members. In the top_left panel of_Figure \[fig:xray\_crosscor\],_we show spectroscopic_members and probable members brighter than_$r = 19.5$_(our 50% completeness limit, as discussed_in_Appendix \[sec:spec\_comp\]). In_this_view,_the northern_galaxy distribution bends_to_the northwest,_slightly_above the filament detected by @2013HEAD...1340101V._From_the optical data alone we can see_that there is some_structure_to the north, but_it is difficult to definitively_constrain the path of that structure.
To_provide a_quantitative measure_of how well the galaxy population aligns with the X-ray emission,_we consider a cross-correlation between the_X-ray flux and galaxy_positions. If_the_galaxy distribution maps_out_the underlying_distribution of X-ray emitting gas, we would_expect that_the cross-correlation would be peaked in_the current alignment and_would_decrease as we move the galaxies._In particular, by rotating the galaxies_around the cluster’s center, the_amplitude_of_the cross-correlation should decrease (although_a small bounce-back is expected as_filaments in the_galaxy distribution rotate onto the next X-ray_filament)._To test this prediction, we consider_this_formula for the galaxy X-ray cross-correlation:_$${\cal_C}_{g_\times X} (\theta) = \frac{\sum\limits_{i}^n_X_{x_i,y_i}(\theta) \times W_i}{\sum\limits_{i}^n W_i}.$$ Here, $X_{x_i,y_i}(\theta)$_is the X-ray flux at the position of galaxy_$i$ |
n$-dimensional sphere and $\operatorname{B}$ is the beta function defined as $$\operatorname{B}{\left(}a,b{\right)}=\frac{\Gamma{\left(}a{\right)}\Gamma{\left(}b{\right)}}{\Gamma{\left(}a+b{\right)}}\quad\forall a,b>0.$$
We let $\operatorname{P}$ be the Schouten tensor defined as $$\operatorname{P}:=\frac{1}{n-2}{\left(}\operatorname{Ric}-\frac{\operatorname{S}}{2{\left(}n-1{\right)}}\,g{\right)}$$ and $\operatorname{B}$ be the Bach tensor whose coordinates are given by $$\operatorname{B}_{ij}:=g^{ab}g^{cd}\operatorname{P}_{ac}\operatorname{W}_{ibjd}+g^{ab}{\left(}\operatorname{P}_{ij;ab}-\operatorname{P}_{ia;jb}{\right)}$$ in Einstein’s summation notation, where $g^{ab}$, $\operatorname{P}_{ac}$, $\operatorname{P}_{ij;ab}$ and $\operatorname{W}_{ibjd}$ are the coordinates of $g^{-1}$, $\operatorname{P}$, $\nabla^2\operatorname{P}$ and $\operatorname{W}$, respectively. The first step in the proof of Proposition \[Pr\] is as follows:
\[Step1\] We have $$\begin{gathered}
\label{Step1Eq1}
P_{2k}=\Delta^k+k\Delta^{k-1}{\left(}f_1\cdot{\right)}+k{\left(}k-1{\right)}\Delta^{k-2}{\left(}f_2\cdot+{\left(}T_1,\nabla{\right)}+{\left(}T_2,\nabla^2{\right)}{\right)}\\
+k{\left(}k-1{\right)}{\left(}k-2{\right)}\Delta^{k-3}{\left(}{\left(}T_3,\nabla^2{\right)}+{\left(}T_4,\nabla^3{\right)}{\right)}\\
+k{\left(}k-1{\right)}{\left(}k-2{\right)}{\left(}k-3{\right)}\Delta^{k-4}{\left(}T_5,\nabla^4{\right)}+Z,\end{gathered}$$ where $Z$ is a | n$-dimensional sphere and $ \operatorname{B}$ is the beta function defined as $ $ \operatorname{B}{\left(}a, b{\right)}=\frac{\Gamma{\left(}a{\right)}\Gamma{\left(}b{\right)}}{\Gamma{\left(}a+b{\right)}}\quad\forall a, b>0.$$
We let $ \operatorname{P}$ be the Schouten tensor define as $ $ \operatorname{P}:=\frac{1}{n-2}{\left(}\operatorname{Ric}-\frac{\operatorname{S}}{2{\left(}n-1{\right)}}\,g{\right)}$$ and $ \operatorname{B}$ be the Bach tensor whose coordinate are given by $ $ \operatorname{B}_{ij}:=g^{ab}g^{cd}\operatorname{P}_{ac}\operatorname{W}_{ibjd}+g^{ab}{\left(}\operatorname{P}_{ij;ab}-\operatorname{P}_{ia;jb}{\right)}$$ in Einstein ’s summation note, where $ g^{ab}$, $ \operatorname{P}_{ac}$, $ \operatorname{P}_{ij;ab}$ and $ \operatorname{W}_{ibjd}$ are the coordinates of $ g^{-1}$, $ \operatorname{P}$, $ \nabla^2\operatorname{P}$ and $ \operatorname{W}$, respectively. The beginning step in the proof of suggestion \[Pr\ ] is as follows:
\[Step1\ ] We have $ $ \begin{gathered }
\label{Step1Eq1 }
P_{2k}=\Delta^k+k\Delta^{k-1}{\left(}f_1\cdot{\right)}+k{\left(}k-1{\right)}\Delta^{k-2}{\left(}f_2\cdot+{\left(}T_1,\nabla{\right)}+{\left(}T_2,\nabla^2{\right)}{\right)}\\
+ k{\left(}k-1{\right)}{\left(}k-2{\right)}\Delta^{k-3}{\left(}{\left(}T_3,\nabla^2{\right)}+{\left(}T_4,\nabla^3{\right)}{\right)}\\
+ k{\left(}k-1{\right)}{\left(}k-2{\right)}{\left(}k-3{\right)}\Delta^{k-4}{\left(}T_5,\nabla^4{\right)}+Z,\end{gathered}$$ where $ Z$ is a | n$-dilensional sphere and $\opevatorname{B}$ is thg veta fnnction defined as $$\operatorname{B}{\left(}a,b{\right)}=\hrac{\Tamma{\oeft(}a{\right)}\Gamma{\left(}b{\rieht)}}{\Gamma{\lvft(}a+b{\righr)}}\quav\forall a,b>0.$$
We let $\operatovuame{P}$ ne thz Wchouten tensot defined as $$\operatorname{P}:=\xrxc{1}{u-2}{\left(}\operatorname{Ric}-\frac{\operatorname{F}}{2{\left(}n-1{\rogjt)}}\,g{\right)}$$ and $\okeratpwnams{B}$ be the Bach tensor whose coordihates age given by $$\operayorname{B}_{ij}:=g^{ab}g^{cd}\operatornale{P}_{af}\operatorname{W}_{ibjd}+h^{ab}{\left(}\operqtorgqme{P}_{ij;ab}-\operxtorname{P}_{ia;jb}{\right)}$$ in Ginstein’s summation notation, whefe $g^{ay}$, $\operatornqmw{P}_{af}$, $\operatornane{P}_{ij;wb}$ and $\operabprname{F}_{ibjd}$ ate the coordinstev od $g^{-1}$, $\operatorname{P}$, $\nabna^2\operatorname{P}$ anq $\operatosncme{W}$, respectively. The furst vtep in gye orokf or Proplsivion \[Pr\] is aa follows:
\[Stwp1\] We have $$\begin{gatnewvc}
\label{Step1Eq1}
L_{2k}=\Deltw^k+h\Delta^{k-1}{\left(}f_1\cdot{\right)}+k{\left(}k-1{\right)}\Delta^{k-2}{\ltft(}f_2\csot+{\left(}T_1,\nabla{\right)}+{\left(}T_2,\babla^2{\right)}{\right)}\\
+k{\left(}k-1{\tight)}{\left(}k-2{\wight)}\Delta^{k-3}{\left(}{\left(}T_3,\nabla^2{\right)}+{\left(}T_4,\nabla^3{\right)}{\righd)}\\
+k{\lefv(}k-1{\figkb)}{\jddt(}n-2{\right)}{\left(}k-3{\right)}\Delta^{k-4}{\left(}T_5,\nabla^4{\right)}+Z,\end{gathqdec}$$ ehere $Z$ is a | n$-dimensional sphere and $\operatorname{B}$ is the beta as a,b>0.$$ We $\operatorname{P}$ be the and be the Bach whose coordinates are by $$\operatorname{B}_{ij}:=g^{ab}g^{cd}\operatorname{P}_{ac}\operatorname{W}_{ibjd}+g^{ab}{\left(}\operatorname{P}_{ij;ab}-\operatorname{P}_{ia;jb}{\right)}$$ in Einstein’s summation notation, $g^{ab}$, $\operatorname{P}_{ac}$, $\operatorname{P}_{ij;ab}$ and $\operatorname{W}_{ibjd}$ are the coordinates of $g^{-1}$, $\operatorname{P}$, $\nabla^2\operatorname{P}$ and respectively. The first step in the proof of Proposition \[Pr\] is as follows: We $$\begin{gathered} P_{2k}=\Delta^k+k\Delta^{k-1}{\left(}f_1\cdot{\right)}+k{\left(}k-1{\right)}\Delta^{k-2}{\left(}f_2\cdot+{\left(}T_1,\nabla{\right)}+{\left(}T_2,\nabla^2{\right)}{\right)}\\ +k{\left(}k-1{\right)}{\left(}k-2{\right)}{\left(}k-3{\right)}\Delta^{k-4}{\left(}T_5,\nabla^4{\right)}+Z,\end{gathered}$$ where $Z$ is a | n$-dimensional sphere and $\operAtorname{B}$ iS the bEta FunCtIon dEfinEd as $$\operatornaME{B}{\leFt(}a,b{\right)}=\frac{\Gamma{\left(}A{\righT)}\GAMma{\lEFt(}B{\righT)}}{\Gamma{\lEFt(}A+B{\RigHt)}}\QuAd\fOrALl A,b>0.$$
We lEt $\oPeratorName{P}$ be the schOuTen tensor defINeD as $$\operatoRnaMe{P}:=\frac{1}{n-2}{\left(}\OpeRatornAmE{RiC}-\Frac{\oPerAtornAme{S}}{2{\leFT(}n-1{\righT)}}\,g{\right)}$$ anD $\oPEratorNAme{B}$ be tHE baCh teNsor whose coordinaTEs ARe given by $$\operaTornamE{B}_{IJ}:=g^{AB}G^{cd}\OpeRatorname{P}_{Ac}\OperaTOrname{W}_{IBjD}+G^{AB}{\leFT(}\operatorname{p}_{ij;ab}-\operatORnaMe{P}_{ia;jB}{\rIghT)}$$ In EinsTein’s SuMMatIon notation, WherE $g^{ab}$, $\operaTornamE{p}_{ac}$, $\operATorname{p}_{ij;ab}$ aNd $\oPerAtorNAmE{W}_{IbjD}$ aRE thE CoOrdINatEs of $g^{-1}$, $\opeRaToRname{p}$, $\nabLA^2\OPEratOrnAme{P}$ And $\opEratorname{W}$, reSpeCtivELy. THe firSt steP in tHe Proof Of PropOsitiOn \[pr\] is as follows:
\[StEp1\] We Have $$\begin{GatHeRed}
\LaBel{StEP1Eq1}
P_{2k}=\DEltA^k+k\delta^{k-1}{\lEft(}f_1\cdoT{\RigHt)}+K{\LEFt(}K-1{\right)}\Delta^{k-2}{\left(}f_2\cDoT+{\LEfT(}T_1,\nabla{\rIght)}+{\leFT(}T_2,\NaBLa^2{\right)}{\rIgHt)}\\
+k{\Left(}K-1{\RIght)}{\lEft(}k-2{\RIgHt)}\Delta^{k-3}{\Left(}{\leFT(}T_3,\NaBla^2{\righT)}+{\lEft(}T_4,\naBlA^3{\riGht)}{\Right)}\\
+K{\Left(}K-1{\right)}{\Left(}k-2{\rigHt)}{\lefT(}K-3{\right)}\Delta^{k-4}{\leFT(}T_5,\nabla^4{\right)}+Z,\ENd{GAThERed}$$ wHerE $Z$ is a | n$-dimensional sphere and$\operator name{ B}$ is t he b etafunction defin e d as $$\operatorname{B}{\l eft(} a, b {\ri g ht )}=\f rac{\Ga m ma { \ lef t( }a {\r ig h t) }\Gam ma{ \left(} b{\right)} }{\ Ga mma{\left(}a + b{ \right)}}\ qua d\forall a,b >0. $$
We l et$ \oper ato rname {P}$ b e the S chouten t en s or def i ned as$ $ \o pera torname{P}:=\frac { 1} { n-2}{\left(}\o perato rn a me { R ic} -\f rac{\opera to rname { S}}{2{\ l ef t ( } n-1 { \right)}}\,g{ \right)}$$a nd$\oper at orn a me{B}$ be t he Bac h tensor wh osecoordinat es are given b y $$\ope ratorn ame {B} _{ij } := g^ {ab }g ^ {cd } \o per a tor name{P}_ {a c} \oper ator n a m e {W}_ {ib jd}+ g^{ab }{\left(}\ope rat orna m e{P }_{ij ;ab}- \ope ra torna me{P}_ {ia;j b} {\right)}$$ inEins tein’s su mma ti onno tatio n , wher e $ g^{ ab}$, $ \operat o rna me { P } _{ ac}$, $\operatorna me { P }_ {ij;ab}$ and $ \ op er a torname{ W} _{i bjd} $ are t he c o or dinatesof $g^ { -1 }$ , $\ope ra tornam e{ P}$ , $ \nabl a ^2\o perato rname{P} $ and $\operatorname { W}$, respecti v el y . T h e fi rst step in th e pr o of o f Pr o po sit i on \[ Pr\]is as follows:
\[Step1\] W e have $$\b egin{gathered }
\label{S t e p 1Eq1}
P_ {2k} = \D e lta^k+k\Delta^ {k-1} {\left(}f_ 1 \cdot{\r ight) }+k{\lef t(}k-1{\r i g ht)}\Del ta^ {k- 2}{ \le f t (} f_2\cdot+{\le f t (}T_ 1, \nabla{ \ri ght)}+{ \le ft( }T_ 2,\ na bla^2{\ri ght)}{\r ig ht )} \\
+k {\lef t (}k-1{\r ig ht) }{ \le ft(}k - 2{\rig ht)}\ Delt a^ {k - 3}{ \left(} { \l e f t(}T _3 ,\ nabl a^2 {\ right )}+{ \ lef t(}T_4, \nabla^3{ \ri g ht)} {\ ri ght)}\\
+k{\left(}k- 1{ \right)}{\ le ft( }k-2{\ r i ght)}{\l eft(}k-3{\right)}\Delta ^ {k-4}{\ lef t(}T_ 5,\n abla^4{\r igh t)}+Z, \en d {gathe red}$$ wher e$Z$ i s a | n$-dimensional sphere_and $\operatorname{B}$_is the beta function_defined as_$$\operatorname{B}{\left(}a,b{\right)}=\frac{\Gamma{\left(}a{\right)}\Gamma{\left(}b{\right)}}{\Gamma{\left(}a+b{\right)}}\quad\forall_a,b>0.$$
We let_$\operatorname{P}$_be the Schouten_tensor defined as_$$\operatorname{P}:=\frac{1}{n-2}{\left(}\operatorname{Ric}-\frac{\operatorname{S}}{2{\left(}n-1{\right)}}\,g{\right)}$$ and $\operatorname{B}$ be_the Bach tensor_whose_coordinates are given by $$\operatorname{B}_{ij}:=g^{ab}g^{cd}\operatorname{P}_{ac}\operatorname{W}_{ibjd}+g^{ab}{\left(}\operatorname{P}_{ij;ab}-\operatorname{P}_{ia;jb}{\right)}$$ in Einstein’s summation notation, where $g^{ab}$, $\operatorname{P}_{ac}$, $\operatorname{P}_{ij;ab}$ and_$\operatorname{W}_{ibjd}$_are the_coordinates_of_$g^{-1}$, $\operatorname{P}$, $\nabla^2\operatorname{P}$ and $\operatorname{W}$,_respectively. The first step in_the proof_of Proposition \[Pr\] is as follows:
\[Step1\] We have $$\begin{gathered}
\label{Step1Eq1}
P_{2k}=\Delta^k+k\Delta^{k-1}{\left(}f_1\cdot{\right)}+k{\left(}k-1{\right)}\Delta^{k-2}{\left(}f_2\cdot+{\left(}T_1,\nabla{\right)}+{\left(}T_2,\nabla^2{\right)}{\right)}\\
+k{\left(}k-1{\right)}{\left(}k-2{\right)}\Delta^{k-3}{\left(}{\left(}T_3,\nabla^2{\right)}+{\left(}T_4,\nabla^3{\right)}{\right)}\\
+k{\left(}k-1{\right)}{\left(}k-2{\right)}{\left(}k-3{\right)}\Delta^{k-4}{\left(}T_5,\nabla^4{\right)}+Z,\end{gathered}$$_where_$Z$ is a |
width="49.70000%"}
We also investigate the evolution for the Newtonian potential, $\Phi$, as a function of $a$ and this is shown in. We see that for a designer $\mathcal{F}(\mathcal{K})$ model which mimics a $\Lambda$CDM background the evolution is now sensitive to the scale, where $K_0= k/H_0$, unlike the case of a cosmological constant. The amplitude of $\Phi$ grows with respect to $\Lambda$CDM for large scales, while for smaller scales the amplitude is suppressed. For scales $K_0 \lesssim 1$, we see that $\Phi$ initially grows before reaching a maximum and then decays due to the increasing contribution from dark energy. A similar feature was also observed in [@Ferreira2] for their power law model of $\mathcal{F}$. We note that this is very similar to other models which introduce a new cosmological fluid with a negative squared sound speed, $c_\mathrm{s}^2=\delta P/\delta \rho$. We solve the differential equation governing the evolution of $\Phi$ [@PhiODE1; @PhiODE2] $$\label{Phi ODE}
\frac{d^2\Phi}{da^2}+\left(\frac{1}{\mathcal{H}}\frac{d\mathcal{H}}{da}+\frac{4}{a}+3\frac{c_\mathrm{s}^2}{a} \right)\frac{d\Phi}{da} + \left[ \frac{2}{a\mathcal{H}} \frac{d\mathcal{H}}{da}+\frac{1}{a^2}(1+3c_\mathrm{s}^2)+\frac{c_\mathrm{s}^2k^2}{a^2\mathcal{H}^2}\right]\Phi = 0,$$ provided there is zero anisotropic stress and so $\Phi = \Psi$. In models where $c_\mathrm{s}^2 < 0 $ we observe the same behaviour for $\Phi$ rising to a maximum before decaying, as seen in. In these models, the initial growth is due to an imaginary $c_\mathrm{s}^2$ causing an unstable growth of perturbations. However, as dark energy begins to dominate $\Phi$ decays as in $\Lambda$CDM. This feature is enhanced for smaller scales until the effect of dark energy in unable to overcome the unstable growth of perturbations and $\Phi$ grows exponentially as seen for | width="49.70000% " }
We also investigate the evolution for the Newtonian potential, $ \Phi$, as a function of $ a$ and this is testify in. We examine that for a designer $ \mathcal{F}(\mathcal{K})$ model which mimics a $ \Lambda$CDM backdrop the evolution is now sensible to the scale, where $ K_0= k / H_0 $, unlike the case of a cosmological constant. The amplitude of $ \Phi$ develop with respect to $ \Lambda$CDM for large scale, while for smaller scales the amplitude is suppressed. For scales $ K_0 \lesssim 1 $, we see that $ \Phi$ initially grows before strive a maximum and then decays due to the increasing contribution from colored energy. A similar feature was also detect in [ @Ferreira2 ] for their power law model of $ \mathcal{F}$. We note that this is very similar to early models which introduce a new cosmological fluid with a negative squared sound focal ratio, $ c_\mathrm{s}^2=\delta P/\delta \rho$. We solve the differential equation governing the evolution of $ \Phi$ [ @PhiODE1; @PhiODE2 ] $ $ \label{Phi ODE }
\frac{d^2\Phi}{da^2}+\left(\frac{1}{\mathcal{H}}\frac{d\mathcal{H}}{da}+\frac{4}{a}+3\frac{c_\mathrm{s}^2}{a } \right)\frac{d\Phi}{da } + \left [ \frac{2}{a\mathcal{H } } \frac{d\mathcal{H}}{da}+\frac{1}{a^2}(1 + 3c_\mathrm{s}^2)+\frac{c_\mathrm{s}^2k^2}{a^2\mathcal{H}^2}\right]\Phi = 0,$$ provided there is zero anisotropic stress and so $ \Phi = \Psi$. In models where $ c_\mathrm{s}^2 < 0 $ we observe the like behavior for $ \Phi$ rising to a utmost before decaying, as see in. In these mannequin, the initial growth is due to an fanciful $ c_\mathrm{s}^2 $ causing an fluid growth of perturbations. However, as blue energy begins to dominate $ \Phi$ decays as in $ \Lambda$CDM. This feature is enhanced for little scales until the effect of dark energy in unable to overcome the precarious growth of perturbations and $ \Phi$ grows exponentially as seen for | widhh="49.70000%"}
We also investigate tht evolution for tkw Newtmnian lotentiau, $\Phi$, as a function of $a$ and tyis iw shown in. We see that for a dedigner $\mqthcel{F}(\mathcal{K})$ model which mimics w $\Lakuda$CDM backgroukd the evolgtion is now sanricive to the scale, where $K_0= k/H_0$, unlike ehe casr lf a cosmologisal bogstahn. Uhe amplitude of $\Phi$ grows with rsspect uo $\Lambda$CDM for lsrge scales, while for smalper dcales the amplitufe is supprgaseq. For scales $Y_0 \lesssim 1$, we see that $\Phi$ initially grows before reazhing a maximum abd hven decays vue to the increaslmg condributipn from dark ekergy. A wimilar feature was anso observed in [@Fetreira2] for tkeir power law model if $\mathwal{F}$. We vite thzt tgis is vecy similar fo other moeels which introduct a bew cosmologidal fltiq with a negative squared sound speed, $c_\katgrm{s}^2=\delta P/\delta \rho$. We solve the differentiwl equatijn governing the evolution of $\Phi$ [@PhiODE1; @PhiODE2] $$\ldbel{Pii ODT}
\fvwz{e^2\Pji}{da^2}+\left(\frac{1}{\mathcal{H}}\frac{d\mathcal{H}}{da}+\frac{4}{a}+3\frac{c_\iztnrk{s}^2}{a} \right)\frac{d\Ihi}{da} + \left[ \frac{2}{a\mstjcsj{H}} \frac{d\mathcxl{H}}{da}+\fxzc{1}{z^2}(1+3c_\mathrm{s}^2)+\frac{c_\mathgm{s}^2k^2}{a^2\majhcal{H}^2}\eight]\Phi = 0,$$ prpvided there is zero anisoteopic stress qnd so $\Phi = \Psi$. In models whexe $c_\majhrm{s}^2 < 0 $ we observe the same yehavikur for $\Phi$ rising tk a maximum beford dvcayhng, as stdn in. In these moqels, the mniticl growtf is due tj an imagijary $g_\kathrm{s}^2$ causing an unstcble crowth of oerturbations. However, as dark eixrgy begins tp gompnate $\Phi$ decans as in $\Lambda$SDM. This featute is enhcnced wor smalleg scales nntil the efsect of dark ajergy in unaule to ovqrcone tye unstxcle growth of lerturbatpous and $\Phu$ grows exponentiakly zs seen for | width="49.70000%"} We also investigate the evolution for potential, as a of $a$ and see for a designer model which mimics $\Lambda$CDM background the evolution is now to the scale, where $K_0= k/H_0$, unlike the case of a cosmological constant. amplitude of $\Phi$ grows with respect to $\Lambda$CDM for large scales, while for scales amplitude suppressed. scales $K_0 \lesssim 1$, we see that $\Phi$ initially grows before reaching a maximum and then due to the increasing contribution from dark energy. similar feature was also in [@Ferreira2] for their power model $\mathcal{F}$. We that is similar to other which introduce a new cosmological fluid with a negative squared sound speed, $c_\mathrm{s}^2=\delta P/\delta \rho$. We solve differential equation evolution of [@PhiODE1; $$\label{Phi \frac{d^2\Phi}{da^2}+\left(\frac{1}{\mathcal{H}}\frac{d\mathcal{H}}{da}+\frac{4}{a}+3\frac{c_\mathrm{s}^2}{a} \right)\frac{d\Phi}{da} + \frac{d\mathcal{H}}{da}+\frac{1}{a^2}(1+3c_\mathrm{s}^2)+\frac{c_\mathrm{s}^2k^2}{a^2\mathcal{H}^2}\right]\Phi = 0,$$ provided there is and so $\Phi = \Psi$. In models where < 0 we observe the same behaviour for rising to a maximum before decaying, as seen In these models, the initial growth is due to an imaginary $c_\mathrm{s}^2$ causing an unstable perturbations. However, as dark begins to dominate decays in This is enhanced smaller scales until the effect of dark energy in unable to the unstable growth of perturbations and $\Phi$ grows exponentially as | width="49.70000%"}
We also investigate the Evolution fOr the newTonIaN potEntiAl, $\Phi$, as a functiON of $a$ And this is shown in. We see tHat foR a DEsigNEr $\MathcAl{F}(\mathCAl{k})$ MOdeL wHiCh mImICs A $\LambDa$CdM backgRound the evOluTiOn is now sensiTIvE to the scalE, whEre $K_0= k/H_0$, unlike The Case of A cOsmOLogicAl cOnstaNt. The aMPlitudE of $\Phi$ groWs WIth resPEct to $\LaMBDa$cDM fOr large scales, whilE FoR Smaller scales tHe amplItUDe IS SupPreSsed. For scaLeS $K_0 \lesSSim 1$, we seE ThAT $\pHi$ iNItially grows bEfore reachiNG a mAximum AnD thEN decayS due tO tHE inCreasing conTribUtion from Dark enERgy. A simILar featUre was AlsO obServED iN [@FErrEiRA2] foR ThEir POweR law modeL oF $\mAthcaL{F}$. We NOTE That ThiS is vEry siMilar to other mOdeLs whICh iNtrodUce a nEw coSmOlogiCal fluId witH a Negative squared SounD speed, $c_\maThrM{s}^2=\DelTa p/\deltA \Rho$. We sOlvE thE differEntial eQUatIoN GOVeRning the evolution oF $\PHI$ [@phIODE1; @PhiOdE2] $$\labeL{phI Ode}
\frac{d^2\PhI}{dA^2}+\leFt(\frAC{1}{\MathcAl{H}}\fRAc{D\mathcal{h}}{da}+\fraC{4}{A}+3\fRaC{c_\mathrM{s}^2}{A} \right)\FrAc{d\phi}{Da} + \lefT[ \Frac{2}{A\mathcAl{H}} \frac{d\MathcAL{H}}{da}+\frac{1}{a^2}(1+3c_\mathRM{s}^2)+\frac{c_\mathrm{S}^2K^2}{a^2\MAThCAl{H}^2}\rIghT]\Phi = 0,$$ provideD theRE is zEro aNIsOtrOPic stRess aNd SO $\PHI = \Psi$. In models where $c_\mAtHrm{s}^2 < 0 $ we ObserVe the same behaViour for $\PhI$ RISing to a mAximUM bEFore decaying, as Seen iN. In these moDEls, the inItial Growth is Due to an imAGInary $c_\maThrM{s}^2$ cAusIng AN UnStable growth oF PErtuRbAtions. HOweVer, as daRk eNerGy bEgiNs To dominatE $\Phi$ decaYs As In $\laMbdA$CDM. THIs featurE iS enHaNceD for sMAller sCales UntiL tHe EFfeCt of darK EnERGy in UnAbLe to OveRcOme thE unsTAblE growth Of perturbAtiONs anD $\PHi$ Grows exPonentially as SeEn for | width="49.70000%"}
We als o investig ate t heevo lu tion for the Newtonian pote ntial, $\Phi$, as a fu nctio no f $a $ a nd th is is s h ow n in. W esee t h at fora d esigner $\mathcal {F} (\ mathcal{K})$ mo del whichmim ics a $\Lamb da$ CDM ba ck gro u nd th e e volut ion is now se nsitive t ot he sca l e, wher e $K _0=k/H_0$, unlike th e c a se of a cosmol ogical c o ns t a nt. Th e amplitud eof $\ P hi$ gro w sw i t h r e spect to $\La mbda$CDM fo r la rge sc al es, whilefor s ma l ler scales the amp litude is suppr e ssed. F o r scale s $K_0 \l ess sim1 $, w e s ee tha t $ \Ph i $ i nitially g ro ws be fore r e a chin g a max imumand then deca ysduet o t he in creas ingco ntrib utionfromda rk energy. A si mila r feature wa sals oobser v ed in[@F err eira2]for the i r p ow e r la w model of $\mathc al { F }$ . We not e that th is is verysi mil ar t o other mod e ls which i ntrodu c eanew cos mo logica lflu idwitha neg ativesquaredsound speed, $c_\mat h rm{s}^2=\delt a P / \ de l ta \ rho $. We solve the diff eren t ia l e q uatio n gov er n in g the evolution of $ \P hi$ [@ PhiOD E1; @PhiODE2] $$\label{ P h i ODE}
\f rac{ d ^2 \ Phi}{da^2}+\le ft(\f rac{1}{\ma t hcal{H}} \frac {d\mathc al{H}}{da } + \frac{4} {a} +3\ fra c{c _ \ ma thrm{s}^2}{a} \ righ t) \frac{d \Ph i}{da}+ \ lef t[\fr ac {2}{a\mat hcal{H}} \ fr ac {d \ma thcal { H}}{da}+ \f rac {1 }{a ^2}(1 + 3c_\ma thrm{ s}^2 )+ \f r ac{ c_\math r m{ s } ^2k^ 2} {a ^2\m ath ca l{H}^ 2}\r i ght ]\Phi = 0,$$ pro vid e d th er eis zero anisotropicst ress and s o$\P hi = \ P s i$. In m odels where $c_\mathrm{ s }^2 < 0 $we ob serv e the sam e b ehavio urf or $\P hi$ ri singto am a ximum b ef ore d ecaying, a s see n in. I n th ese mod els, the initial g r owt h is due to a n i magi n a ry $c _ \m a thr m{ s }^2 $ causing an unst able growt ho fperturbati o ns. H owever, as dar k ene r gy begi ns to dom inate $\P hi $ de c a ysas in $\La mbda$CDM . This fe a turei senhan ced for s ma lle r sca les un t ilthe e ffectof darkenerg yin unabl e to overcome the unsta ble gr owthofperturbat ion s an d $\Phi$grow s exponent ial lyas se enf or | width="49.70000%"}
We also_investigate the_evolution for the Newtonian_potential, $\Phi$,_as_a function_of_$a$ and this_is shown in._We see that for_a designer $\mathcal{F}(\mathcal{K})$_model_which mimics a $\Lambda$CDM background the evolution is now sensitive to the scale, where_$K_0=_k/H_0$, unlike_the_case_of a cosmological constant. The_amplitude of $\Phi$ grows with_respect to_$\Lambda$CDM for large scales, while for smaller scales_the_amplitude is suppressed._For scales $K_0 \lesssim 1$, we see that $\Phi$_initially grows before reaching a maximum_and then decays_due_to_the increasing contribution from_dark energy. A similar feature was_also observed in [@Ferreira2] for their_power law model of $\mathcal{F}$. We note_that this is very similar to_other models which introduce a_new cosmological_fluid with a negative squared_sound speed, $c_\mathrm{s}^2=\delta_P/\delta \rho$._We solve the_differential equation governing the evolution of_$\Phi$ [@PhiODE1; @PhiODE2]_$$\label{Phi ODE}
\frac{d^2\Phi}{da^2}+\left(\frac{1}{\mathcal{H}}\frac{d\mathcal{H}}{da}+\frac{4}{a}+3\frac{c_\mathrm{s}^2}{a} \right)\frac{d\Phi}{da} + \left[ \frac{2}{a\mathcal{H}}_\frac{d\mathcal{H}}{da}+\frac{1}{a^2}(1+3c_\mathrm{s}^2)+\frac{c_\mathrm{s}^2k^2}{a^2\mathcal{H}^2}\right]\Phi_= 0,$$ provided_there_is_zero anisotropic_stress and so_$\Phi_= \Psi$._In_models where $c_\mathrm{s}^2 < 0 $_we_observe the same behaviour for $\Phi$ rising_to a maximum before_decaying,_as seen in. In_these models, the initial growth_is due to an imaginary $c_\mathrm{s}^2$_causing an_unstable growth_of perturbations. However, as dark energy begins to dominate $\Phi$ decays_as in $\Lambda$CDM. This feature is_enhanced for smaller scales_until the_effect_of dark energy_in_unable to_overcome the unstable growth of perturbations and_$\Phi$ grows_exponentially as seen for |
^{0}$ are gauge equivalent to this first term.
In term of the induced representations, $P^{0}$ being rigid means that all the representations obtained from Maurer-Cartan elements of the form (\[eq:MC\_element\]) are equivalent as representations to the representation $$T_{g}^{P^{0}}\psi(x)=P^{0}(x)\psi(\varphi_{g}^{-1}(x)).$$
The following theorem gives cohomological conditions answering the questions mentioned above.
\[thm:ExistenceAndRigidity\]Let $P^{0}\in\pol_{d}^{\bullet}(0)$ be a Maurer-Cartan element and $P^{1}\in\pol_{d}^{1}(1)$ a one cocycle (i.e $d_{P^{0}}P^{1}=0$). If $$H^{2}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq2,$$ then there exists a Maurer-Cartan element $\omega$ in $\mathcal{P}_{\varphi}^{\bullet}$ such that $$\omega=P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2}).$$ Moreover if $$H^{1}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq1,$$ the Maurer-Cartan element $P^{0}$ is rigid.
The proof relies on Proposition A.3 and A.6 of the Appendix A of [@ACD]. Since $\gamma=P^{0}+\hbar P^{1}$ is a Maurer-Cartan element modulo $\mathcal{F}^{2}\mathcal{P}_{\varphi}^{\bullet}$, Proposition A.3 tells us that there exist a Maurer-Cartan element $\omega=P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2})$ provided $$H^{2}(\mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet},\, d_{\gamma})=0,\quad n\geq2,$$ where $d_{\gamma}$ is the operator $d_{\gamma}a=da+[\gamma,a]$, which becomes a differential on the quotient $\mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet}$. The first part of the theorem follows from (\[eq | ^{0}$ are gauge equivalent to this first term.
In term of the induce representation, $ P^{0}$ being rigid mean that all the representations receive from Maurer - Cartan elements of the phase (\[eq: MC\_element\ ]) are equivalent as representations to the representation $ $ T_{g}^{P^{0}}\psi(x)=P^{0}(x)\psi(\varphi_{g}^{-1}(x)).$$
The come theorem gives cohomological conditions answer the questions mentioned above.
\[thm: ExistenceAndRigidity\]Let $ P^{0}\in\pol_{d}^{\bullet}(0)$ be a Maurer - Cartan element and $ P^{1}\in\pol_{d}^{1}(1)$ a one cocycle (i.e $ d_{P^{0}}P^{1}=0 $). If $ $ H^{2}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq2,$$ then there exists a Maurer - Cartan element $ \omega$ in $ \mathcal{P}_{\varphi}^{\bullet}$ such that $ $ \omega = P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2}).$$ Moreover if $ $ H^{1}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq1,$$ the Maurer - Cartan component $ P^{0}$ is rigid.
The proof relies on Proposition A.3 and A.6 of the Appendix A of [ @ACD ]. Since $ \gamma = P^{0}+\hbar P^{1}$ is a Maurer - Cartan chemical element modulo $ \mathcal{F}^{2}\mathcal{P}_{\varphi}^{\bullet}$, Proposition A.3 tells us that there exist a Maurer - Cartan element $ \omega = P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2})$ provided $ $ H^{2}(\mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet},\, d_{\gamma})=0,\quad n\geq2,$$ where $ d_{\gamma}$ is the hustler $ d_{\gamma}a = da+[\gamma, a]$, which becomes a differential gear on the quotient $ \mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet}$. The first part of the theorem follows from (\[eq | ^{0}$ arf gauge equivalent to thls first term.
In jeem of vhe indhced repfesentations, $P^{0}$ being rigid mxans that all the representatiovs obtainvd from Mqurec-Cartan elements of the njrm (\[se:MC\_enxment\]) are equivslent as rapresentations tu che representation $$T_{g}^{P^{0}}\psi(x)=P^{0}(x)\psi(\varphy_{g}^{-1}(x)).$$
The goplowing theorei ginef cogomological conditions answering tge quesuions mentioned abpve.
\[thm:ExistenceAndRigidity\]Pet $O^{0}\in\pol_{d}^{\bullet}(0)$ be a Maurer-Cartqn ejwment and $P^{1}\iv\pol_{d}^{1}(1)$ a ont eocycle (i.e $s_{P^{0}}P^{1}=0$). If $$H^{2}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq2,$$ ghen chere existw q Mworer-Cartan eoemenn $\omega$ in $\mabncal{P}_{\vdrphi}^{\buklet}$ such that $$\omxga=P^{0}+\ybar P^{1}+\mathcal{O}(\hbar^{2}).$$ Moceover if $$H^{1}(\pol_{d}^{\bullej}(n),d_{P^{0}})=0,\quad n\ces1,$$ the Maurer-Cartab wlemett $P^{0}$ is fugia.
Tht pcoor relifs kn Proposifion A.3 and Q.6 of the Appendix A os [@ACD]. Since $\gajma=P^{0}+\hbwr P^{1}$ is a Maurer-Cartan element modulo $\matvcam{F}^{2}\mathcal{P}_{\varphi}^{\bullet}$, Proposition A.3 tells ud that thqre exist a Maurer-Cartan element $\omega=P^{0}+\hbar P^{1}+\mathwal{O}(\huaf^{2})$ pxividea $$H^{2}(\lathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{[}_{\batpmi}^{\bullet},\, d_{\gamma})=0,\qmad n\geq2,$$ where $d_{\gakmw}$ of the operatot $d_{\gammc}z=dz+[\gamma,a]$, which becoles a dyfferwntial on the quotient $\mathcal{F}^{n}\mathcal{P}_{\carphi}^{\bullet}/\iqthcal{F}^{n+1}\mathcal{P}_{\vaxphi}^{\bullet}$. Tke firxt patt of the theorem follocs froj (\[eq | ^{0}$ are gauge equivalent to this first term the induced $P^{0}$ being rigid obtained Maurer-Cartan elements of form (\[eq:MC\_element\]) are as representations to the representation $$T_{g}^{P^{0}}\psi(x)=P^{0}(x)\psi(\varphi_{g}^{-1}(x)).$$ following theorem gives cohomological conditions answering the questions mentioned above. \[thm:ExistenceAndRigidity\]Let $P^{0}\in\pol_{d}^{\bullet}(0)$ be Maurer-Cartan element and $P^{1}\in\pol_{d}^{1}(1)$ a one cocycle (i.e $d_{P^{0}}P^{1}=0$). If $$H^{2}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq2,$$ then exists Maurer-Cartan $\omega$ $\mathcal{P}_{\varphi}^{\bullet}$ such that $$\omega=P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2}).$$ Moreover if $$H^{1}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq1,$$ the Maurer-Cartan element $P^{0}$ is rigid. The relies on Proposition A.3 and A.6 of the A of [@ACD]. Since P^{1}$ is a Maurer-Cartan element $\mathcal{F}^{2}\mathcal{P}_{\varphi}^{\bullet}$, A.3 tells that exist Maurer-Cartan element $\omega=P^{0}+\hbar provided $$H^{2}(\mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet},\, d_{\gamma})=0,\quad n\geq2,$$ where $d_{\gamma}$ is the operator $d_{\gamma}a=da+[\gamma,a]$, which becomes a differential on the quotient The first the theorem from | ^{0}$ are gauge equivalent to this fIrst term.
In Term oF thE inDuCed rEpreSentations, $P^{0}$ beiNG rigId means that all the repreSentaTiONs obTAiNed frOm MaureR-caRTAn eLeMeNts Of THe Form (\[eQ:MC\_Element\]) Are equivalEnt As RepresentatiONs To the repreSenTation $$T_{g}^{P^{0}}\psi(X)=P^{0}(x)\Psi(\varPhI_{g}^{-1}(x)).$$
tHe folLowIng thEorem gIVes cohOmologicaL cONditioNS answerING tHe quEstions mentioned aBOvE.
\[Thm:ExistenceAnDRigidItY\]leT $p^{0}\In\pOl_{d}^{\Bullet}(0)$ be a MAuRer-CaRTan elemENt AND $p^{1}\in\POl_{d}^{1}(1)$ a one cocyclE (i.e $d_{P^{0}}P^{1}=0$). If $$H^{2}(\poL_{D}^{\buLlet}(n),d_{p^{0}})=0,\qUad N\Geq2,$$ theN therE eXIstS a Maurer-CarTan eLement $\omeGa$ in $\maTHcal{P}_{\vaRPhi}^{\bullEt}$ such ThaT $$\omEga=P^{0}+\HBaR P^{1}+\MatHcAL{O}(\hBAr^{2}).$$ morEOveR if $$H^{1}(\pol_{d}^{\BuLlEt}(n),d_{P^{0}})=0,\Quad N\GEQ1,$$ The MAurEr-CaRtan eLement $P^{0}$ is rigiD.
ThE proOF reLies oN PropOsitIoN A.3 and a.6 of the appenDiX A of [@ACD]. Since $\gamMa=P^{0}+\hBar P^{1}$ is a MaUreR-CArtAn ElemeNT modulO $\maThcAl{F}^{2}\mathCal{P}_{\varPHi}^{\bUlLET}$, prOposition A.3 tells us tHaT THeRe exist a maurer-cArTaN Element $\oMeGa=P^{0}+\Hbar p^{1}+\MAthcaL{O}(\hbAR^{2})$ pRovided $$H^{2}(\MathcaL{f}^{n}\MaThcal{P}_{\vArPhi}^{\bulLeT}/\maThcAl{F}^{n+1}\mAThcaL{P}_{\varpHi}^{\bullet},\, D_{\gammA})=0,\Quad n\geq2,$$ where $d_{\GAmma}$ is the operAToR $D_{\GaMMa}a=dA+[\gaMma,a]$, which beComeS A difFereNTiAl oN The quOtienT $\mAThCAl{F}^{n}\mathcal{P}_{\varphi}^{\bUlLet}/\matHcal{F}^{N+1}\mathcal{P}_{\varpHi}^{\bullet}$. ThE FIRst part oF the THeORem follows from (\[Eq | ^{0}$ are gauge equivalent to this f irstter m.
I n te rm o f the inducedr epre sentations, $P^{0}$ be ing r ig i d me a ns that all th e r e p res en ta tio ns ob taine d f rom Mau rer-Cartan el em ents of thef or m (\[eq:MC \_e lement\]) ar e e quival en t a s repr ese ntati ons to the re presentat io n $$T_{ g }^{P^{0 } } \p si(x )=P^{0}(x)\psi(\v a rp h i_{g}^{-1}(x)) .$$
T he fo l l owi ngtheorem gi ve s coh o mologic a lc o n dit i ons answering the questi o nsmentio ne d a b ove.
\[thm :E x ist enceAndRigi dity \]Let $P^ {0}\in \ pol_{d} ^ {\bulle t}(0)$ be aMaur e r- Ca rta ne lem e nt an d $P ^{1}\in\ po l_ {d}^{ 1}(1 ) $ a one co cycl e (i. e $d_{P^{0}}P ^{1 }=0$ ) . I f $$H ^{2}( \pol _{ d}^{\ bullet }(n), d_ {P^{0}})=0,\qua d n\ geq2,$$ t hen t her eexist s a Mau rer -Ca rtan el ement $ \ ome ga $ i n$\mathcal{P}_{\var ph i } ^{ \bullet} $ such th at $$\omega =P ^{0 }+\h b a r P^{ 1}+\ m at hcal{O}( \hbar^ { 2} ). $$ More ov er if$$ H^{ 1}( \pol_ { d}^{ \bulle t}(n),d_ {P^{0 } })=0,\quad n\g e q1,$$ the Mau r er - C ar t an e lem ent $P^{0}$ isr igid .
T h epro o f rel ies o nP ro p osition A.3 and A.6 o f theAppen dix A of [@AC D]. Since$ \ g amma=P^{ 0}+\ h ba r P^{1}$ is a M aurer -Cartan el e ment mod ulo $ \mathcal {F}^{2}\m a t hcal{P}_ {\v arp hi} ^{\ b u ll et}$, Proposi t i on A .3 tellsusthat th ere ex ist aMa urer-Cart an eleme nt $ \o me ga= P^{0} + \hbar P^ {1 }+\ ma thc al{O} ( \hbar^ {2})$ pro vi de d $$ H^{2}(\ m at h c al{F }^ {n }\ma thc al {P}_{ \var p hi} ^{\bull et}/\math cal { F}^{ n+ 1} \mathca l{P}_{\varphi }^ {\bullet}, \, d_ {\gamm a } )=0,\qua d n\geq2,$$ where $d_{\ g amma}$isthe o pera tor $d_{\ gam ma}a=d a+[ \ gamma, a]$, w hichbe com e s a di f f er ent ia l on the q u o tie nt $\ ma thca l{F}^{n }\mathcal{P}_{\var p hi} ^{\bullet}/\m ath cal{ F } ^{ n+1 } \m a thc al { P}_ { \ varphi}^{\bulle t}$. The f ir s tpart of th e th eo rem fol lows fr om (\ [ eq | ^{0}$ are_gauge equivalent_to this first term.
In_term of_the_induced representations,_$P^{0}$_being rigid means_that all the_representations obtained from Maurer-Cartan_elements of the_form_(\[eq:MC\_element\]) are equivalent as representations to the representation $$T_{g}^{P^{0}}\psi(x)=P^{0}(x)\psi(\varphi_{g}^{-1}(x)).$$
The following theorem gives cohomological conditions_answering_the questions_mentioned_above.
\[thm:ExistenceAndRigidity\]Let_$P^{0}\in\pol_{d}^{\bullet}(0)$ be a Maurer-Cartan element_and $P^{1}\in\pol_{d}^{1}(1)$ a one cocycle_(i.e $d_{P^{0}}P^{1}=0$)._If $$H^{2}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq2,$$ then there exists a Maurer-Cartan_element_$\omega$ in $\mathcal{P}_{\varphi}^{\bullet}$_such that $$\omega=P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2}).$$ Moreover if $$H^{1}(\pol_{d}^{\bullet}(n),d_{P^{0}})=0,\quad n\geq1,$$ the_Maurer-Cartan element $P^{0}$ is rigid.
The proof_relies on Proposition_A.3_and_A.6 of the Appendix_A of [@ACD]. Since $\gamma=P^{0}+\hbar P^{1}$_is a Maurer-Cartan element modulo $\mathcal{F}^{2}\mathcal{P}_{\varphi}^{\bullet}$,_Proposition A.3 tells us that there exist_a Maurer-Cartan element $\omega=P^{0}+\hbar P^{1}+\mathcal{O}(\hbar^{2})$ provided_$$H^{2}(\mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet},\, d_{\gamma})=0,\quad n\geq2,$$ where $d_{\gamma}$_is the_operator $d_{\gamma}a=da+[\gamma,a]$, which becomes a_differential on the_quotient $\mathcal{F}^{n}\mathcal{P}_{\varphi}^{\bullet}/\mathcal{F}^{n+1}\mathcal{P}_{\varphi}^{\bullet}$._The first part_of the theorem follows from (\[eq |
$K and continuous to grow up to $S_m =0.8 * (2RLn4)$ at T = 40K. Similar behavior is observed in the $x=1$ sample (i.e. Ce$_2$Ni$_2$Sn) but with $S_(T)$ increasing in a more monotonous manner. Compared with the $S_m(T)$ behavior of a sharp energy distribution of levels, provided by the results depicted in Fig. \[F5\]a, it becomes evident an overlap of the side levels of the ground and first excited CFE levels.
This is the key to understand the complex magnetic behavior of this Ni-rich alloys which show a decrease of the magnetic moment with a simultaneous increase of a weak $T_{N2}$. The former effect is driven by the increase of the $4f$ state hybridization whereas the latter is driven by the increase of the GS density of states due to the arising contribution of the low energy levels of the CFE excited state. Such an enhancement of the low energy density of states is supported by the increasing $C_m/T$ contribution to the specific heat above $T_{N2}$ deduced from Fig. \[F6\], which increases from 0.2J/molK$^2$ for x = 0.5 up to 0.45J/molK$^2$ for x=1 around $T\approx 15$K. Low effective moments and relatively high ordering temperatures are typical of itinerant magnetic systems [@DeLong], provided that the density of states is large enough. This scenario is favored by the reduction of the Ce-Ce spacings decrease driven by the reduction of $a$ and $b$ lattice parameters.
Within a simplified picture, such would be the expected way to undergo from a N = 2 degenerated GS to a N = 4 degenerated GS in a continuous way. In Fig. \[F5\]b we include a theoretical prediction [@Coqblin] for a N=4 degenerated state with a characteristic temperature $T_o=30$K. This comparison clearly indicates that, despite of the energy levels contribution of the first excited CFE state, the N=4 GS is not reached in this system. Similar situation was observed in other strongly anisotropic Ce systems such as CeTiGe [@Micha].
The low temperature magnetic properties are resumed in the magnetic phase diagram presented in Fig. \[F7\]. On the Pd-rich | $ K and continuous to grow up to $ S_m = 0.8 * (2RLn4)$ at T = 40K. Similar demeanor is observe in the $ x=1 $ sample (i.e. Ce$_2$Ni$_2$Sn) but with $ S_(T)$ increase in a more flat manner. Compared with the $ S_m(T)$ behavior of a astute department of energy distribution of levels, provided by the results depicted in Fig. \[F5\]a, it becomes discernible an overlap of the side horizontal surface of the flat coat and first excited CFE levels.
This is the key to understand the complex charismatic behavior of this Ni - rich alloys which testify a decrease of the magnetic moment with a coincident increase of a watery $ T_{N2}$. The former effect is driven by the increase of the $ 4f$ state hybridization whereas the latter is drive by the increase of the GS density of states due to the arising contribution of the low department of energy degree of the CFE excited state. Such an enhancement of the low energy density of states is supported by the increasing $ C_m / T$ contribution to the specific heat above $ T_{N2}$ deduce from Fig. \[F6\ ], which increases from 0.2J / molK$^2 $ for x = 0.5 up to 0.45J / molK$^2 $ for x=1 around $ T\approx 15$K. Low effective here and now and relatively gamey ordering temperatures are distinctive of itinerant magnetic system [ @DeLong ], provided that the density of country is large enough. This scenario is favored by the reduction of the Ce - Ce spacings decrease driven by the reduction of $ a$ and $ b$ wicket parameters.
Within a simplified picture, such would be the expected way to undergo from a N = 2 degenerated GS to a N = 4 degenerated GS in a continuous way. In Fig. \[F5\]b we admit a theoretical prediction [ @Coqblin ] for a N=4 degenerated department of state with a characteristic temperature $ T_o=30$K. This comparison clearly indicates that, despite of the energy levels contribution of the inaugural excited CFE state, the N=4 GS is not reached in this arrangement. Similar situation was observed in other powerfully anisotropic Ce systems such as CeTiGe [ @Micha ].
The low temperature magnetic properties are resumed in the charismatic phase diagram presented in Fig. \[F7\ ]. On the Pd - rich | $K ajd continuous to grow up to $S_m =0.8 * (2RLn4)$ at T = 40K. Smmilar gehavior is observed in the $x=1$ sample (i.w. Ce$_2$Nu$_2$Sn) but with $S_(T)$ increaring in a more mobotoious manner. Compeded witm the $D_m(T)$ yeiavior of a shatp energy divtribution of nexeps, provided by the results depicted in Fig. \[G5\]a, it becomes evydenu ag ovsglcp of the side levels of the grkund anv first excited CFE levels.
This is the key to knderstand the comolex magnetuc bqyavior of thks Ni-rich alloys which show a decrease of the magnetiz momznt with a winulhdneous incrxase os a weak $T_{N2}$. Bne forker effrct is driven ny thx inxrease of the $4f$ state hybridization wheteas the ldtcer is driven by the unxreasg of dhe EW ddnsjtb or statfs vue to the zrising conrribution of the loe qbergy levels kf the CSE excited state. Such an enhancement of ths low energy density of states is supported hy the insreasing $C_m/T$ contribution to the specific heat abmve $T_{I2}$ aedbgcd ffim Fig. \[F6\], which increases from 0.2J/molK$^2$ for x = 0.5 up tj 0.45J/kokK$^2$ for x=1 arounb $T\approx 15$K. Low rfvevjive moments avd relcfibely high ordering temperwturew are typycal of itinerant magnetic systwms [@DeLong], pgovieed that the densicy of states is karge enough. This scenario ir fabored by thf reductikv of the Ce-Ce spxcikgs decreast driven by the requction oh $a$ aud $b$ latgice paramqters.
Withij a slkplified picture, skch wlund be the fxpected way to undergo from a I = 2 degeneratgd CS no a N = 4 begenevated GS in a cjntinuous way. Nn Fig. \[F5\]b we ivclude a tgeoretiral predictijn [@Coqblin] fos a N=4 degenereted statq wirh a characgdristic tempersture $T_o=30$K. This compqrison clearly indlcater that, despite oy uhw energy levelx cuntwihuvion jx the first axciged VFE sgate, tht N=4 ES ix not reached in thiv syatem. Similar situayijn was ovserved yn other stromgly anisotropic Ct systxms surh as VeTyGe [@Micha].
The low temperature mzgnetic pgopcrties are refumee in the maguetic phase diagram presented in Fig. \[F7\]. On the Pd-rich | $K and continuous to grow up to * at T 40K. Similar behavior sample Ce$_2$Ni$_2$Sn) but with increasing in a monotonous manner. Compared with the $S_m(T)$ of a sharp energy distribution of levels, provided by the results depicted in \[F5\]a, it becomes evident an overlap of the side levels of the ground first CFE This the key to understand the complex magnetic behavior of this Ni-rich alloys which show a decrease the magnetic moment with a simultaneous increase of weak $T_{N2}$. The former is driven by the increase the state hybridization the is by the increase the GS density of states due to the arising contribution of the low energy levels of the excited state. enhancement of low density states is supported increasing $C_m/T$ contribution to the specific deduced from Fig. \[F6\], which increases from 0.2J/molK$^2$ x = up to 0.45J/molK$^2$ for x=1 around 15$K. Low effective moments and relatively high ordering are typical of itinerant magnetic systems [@DeLong], provided that the density of states is large scenario is favored by reduction of the spacings driven the of $a$ $b$ lattice parameters. Within a simplified picture, such would be the way to undergo from a N = 2 degenerated GS N 4 degenerated GS a continuous way. In \[F5\]b include a theoretical prediction a degenerated characteristic $T_o=30$K. comparison clearly indicates that, of the energy levels contribution the first excited CFE not reached in this system. Similar situation was in other strongly anisotropic Ce systems such CeTiGe [@Micha]. The low temperature magnetic properties are resumed in the magnetic diagram presented \[F7\]. On the Pd-rich | $K and continuous to grow up to $S_M =0.8 * (2RLn4)$ at T = 40K. SiMilar BehAviOr Is obServEd in the $x=1$ sample (I.E. Ce$_2$NI$_2$Sn) but with $S_(T)$ increasing In a moRe MOnotONoUs manNer. CompAReD WIth ThE $S_M(T)$ bEhAViOr of a ShaRp energY distributIon Of Levels, providED bY the resultS dePicted in Fig. \[F5\]A, it BecomeS eVidENt an oVerLap of The sidE Levels Of the grouNd ANd firsT Excited cfe lEvelS.
This is the key to unDErSTand the complex MagnetIc BEhAVIor Of tHis Ni-rich aLlOys whICh show a DEcREASe oF The magnetic moMent with a siMUltAneous InCreASe of a wEak $T_{N2}$. thE ForMer effect is DrivEn by the inCrease OF the $4f$ stATe hybriDizatiOn wHerEas tHE lAtTer Is DRivEN bY thE IncRease of tHe gS DensiTy of STATEs duE to The aRisinG contribution Of tHe loW EneRgy leVels oF the cFe exciTed staTe. SucH aN enhancement of tHe loW energy deNsiTy Of sTaTes is SUpportEd bY thE increaSing $C_m/T$ COntRiBUTIoN to the specific heat AbOVE $T_{n2}$ deduced From FiG. \[f6\], wHiCH increasEs FroM 0.2J/moLk$^2$ For x = 0.5 uP to 0.45J/MOlk$^2$ for x=1 aroUnd $T\apPRoX 15$K. low effeCtIve momEnTs aNd rElatiVEly hIgh ordEring temPeratURes are typical oF Itinerant magnETiC SYsTEms [@DELoNg], provided tHat tHE denSity OF sTatES is laRge enOuGH. THIs scenario is favored By The redUctioN of the Ce-Ce spaCings decreASE Driven by The rEDuCTion of $a$ and $b$ latTice pArameters.
WIThin a simPlifiEd picturE, such woulD BE the expeCteD waY to UndERGo From a N = 2 degenerATEd GS To A N = 4 degenEraTed GS in A coNtiNuoUs wAy. in Fig. \[F5\]b we Include a ThEoReTiCal PrediCTion [@CoqbLiN] foR a n=4 deGenerATed staTe witH a chArAcTEriStic temPErATUre $T_O=30$K. thIs coMpaRiSon clEarlY IndIcates tHat, despitE of THe enErGy Levels cOntribution of ThE first exciTeD CFe state, THE N=4 GS is noT reached in this system. SimILar situAtiOn was ObseRved in othEr sTronglY anISotropIc Ce syStems SuCh aS cETiGe [@mIChA].
ThE lOw temperatURE maGnetiC pRopeRties arE resumed in the magneTIc pHase diagram prEseNted IN fiG. \[F7\]. ON ThE pd-rIcH | $K and continuous to growup to $S_m =0.8 *(2R Ln 4)$at T = 40K. Simila r beh avior is observed in t he $x =1 $ sam p le (i.e . Ce$_2 $ Ni $ _ 2$S n) b utwi t h$S_(T )$increas ing in a m ore m onotonous ma n ne r. Compare d w ith the $S_m (T) $ beha vi oro f a s har p ene rgy di s tribut ion of le ve l s, pro v ided by t he res ults depicted inF ig . \[F5\]a, it b ecomes e v id e n t a n o verlap ofth e sid e levels of t h e g r ound and firs t excited C F E l evels.
Thi s is th e key t o un derstand th e co mplex mag neticb ehavior of this Ni-ri chall oysw hi ch sh ow a d e cr eas e of the mag ne ti c mom entw i t h a s imu ltan eousincrease of a we ak $ T _{N 2}$.The f orme reffec t is d riven b y the increaseof t he $4f$ s tat ehyb ri dizat i on whe rea s t he latt er is d r ive nb y th e increase of theGS d en sity ofstates du et o the ar is ing con t r ibuti on o f t he low e nergyl ev el s of th eCFE ex ci ted st ate.S uchan enh ancement of t h e low energy d e nsity of stat e si s s u ppor ted by the inc reas i ng $ C_m/ T $con t ribut ion t ot he specific heat above $ T_{N2} $ ded uced from Fig . \[F6\],w h i ch incre ases fr o m 0.2J/molK$^2 $ for x = 0.5 u p to 0.45 J/mol K$^2$ fo r x=1 aro u n d $T\app rox 15 $K. Lo w ef fective momen t s and r elative lyhigh or der ing te mpe ra tures are typical o fit in era nt ma g netic sy st ems [ @De Long] , provi ded t hatth ed ens ity ofs ta t e s is l ar ge e nou gh . Thi s sc e nar io is f avored by th e red uc ti on of t he Ce-Ce spac in gs decreas edri ven by t he reduc tion of $a$ and $b$ lat t ice par ame ters.
Wi thin a si mpl ifiedpic t ure, s uch wo uld b ethe e xpect e d w ayto undergo f r o m a N =2dege nerated GS to a N = 4 deg e ner ated GS in acon tinu o u sway . I n Fi g. \[F 5 \ ]b we include a theoretic al pr ediction [ @ Coq bl in] for a N=4degen e rated s tate with a charac te rist i c te mperature$T_o=30$ K. This c o mpari s on clea rly indic at esthat, despi t e o f the energ ylevels cont ri bution o f the first excited CFE state , the N= 4 GS is n otr eac hed in th is s ystem. Sim ila r s ituat ion was o bser v ed in other str o ngly anis o tr opi c Ce systems su c h asCeTiG e [ @ Micha] .
T he low temperatur e magnetic prop erti e s ar e r e sume din the magneti c p ha s e diagram p resented in Fig. \[ F7 \ ]. On the P d-rich | $K and_continuous to_grow up to $S_m_=0.8 *_(2RLn4)$_at T_=_40K. Similar behavior_is observed in_the $x=1$ sample (i.e._Ce$_2$Ni$_2$Sn) but with_$S_(T)$_increasing in a more monotonous manner. Compared with the $S_m(T)$ behavior of a sharp_energy_distribution of_levels,_provided_by the results depicted in_Fig. \[F5\]a, it becomes evident an_overlap of_the side levels of the ground and first_excited_CFE levels.
This is_the key to understand the complex magnetic behavior of_this Ni-rich alloys which show a_decrease of the_magnetic_moment_with a simultaneous increase_of a weak $T_{N2}$. The former_effect is driven by the increase_of the $4f$ state hybridization whereas the_latter is driven by the increase_of the GS density of_states due_to the arising contribution of_the low energy_levels of_the CFE excited_state. Such an enhancement of the_low energy density_of states is supported by the_increasing_$C_m/T$ contribution to_the_specific_heat above_$T_{N2}$ deduced from_Fig. \[F6\],_which increases_from_0.2J/molK$^2$ for x = 0.5 up_to_0.45J/molK$^2$ for x=1 around $T\approx 15$K. Low_effective moments and relatively_high_ordering temperatures are typical_of itinerant magnetic systems [@DeLong],_provided that the density of states_is large_enough. This_scenario is favored by the reduction of the Ce-Ce spacings decrease_driven by the reduction of $a$_and $b$ lattice parameters.
Within_a simplified_picture,_such would be_the_expected way_to undergo from a N = 2_degenerated GS_to a N = 4 degenerated_GS in a continuous_way._In Fig. \[F5\]b we include a theoretical_prediction [@Coqblin] for a N=4 degenerated_state with a characteristic temperature_$T_o=30$K._This_comparison clearly indicates that, despite_of the energy levels contribution of_the first excited_CFE state, the N=4 GS is not_reached_in this system. Similar situation was_observed_in other strongly anisotropic Ce systems_such_as_CeTiGe [@Micha].
The low temperature magnetic_properties are resumed in the magnetic_phase diagram presented in Fig. \[F7\]. On the Pd-rich |
\] (u22) edge\[normalEdge, <-\] (u21) ++(2, 0) node\[normalNode\] (u23) edge\[normalEdge, <-\] (u22); (s) ++(1.5, -1.5) node\[normalNode\] (u31) edge\[normalEdge, <-\] (s) ++(2, 0) node\[normalNode\] (u32) edge\[normalEdge, <-\] (u31) ++(2, 0) node\[normalNode\] (u33) edge\[normalEdge, <-\] (u32) ++(2, 1.5) node\[labeledNode\] (v) [$v$]{} edge\[normalEdge, bend left=15, <-\] (s) edge\[normalEdge, <-\] (u13) edge\[normalEdge, <-\] (u23) edge\[normalEdge, <-\] (u33) ++(2, 1.5) node\[normalNode\] (v11) edge\[normalEdge, <-\] (v) edge\[secondaryEdge, bend right=40, <->\] (u11) ++(2, 0) node\[normalNode\] (v12) edge\[normalEdge, <-\] (v11) edge\[secondaryEdge, <->, bend right=40\] (u21) ++(2, 0) node\[normalNode\] (v13) edge\[normalEdge, <-\] (v12) edge\[secondaryEdge, bend right=42, <->\] (u31); (v) ++(2, 0) node\[normalNode\] (v21) edge\[normalEdge, <-\] (v) edge\[secondaryEdge, bend right=45, <->\] (u12) ++(2, 0) node\[normalNode\] (v22) edge\[normalEdge, <-\] (v21) edge\[secondaryEdge, bend right=20, <->\] (u22) ++(2, 0) node\[normalNode\] (v23) edge\[normalEdge, <-\] (v22) edge\[secondaryEdge, | \ ] (u22) edge\[normalEdge, & lt;-\ ] (u21) + + (2, 0) node\[normalNode\ ] (u23) edge\[normalEdge, & lt;-\ ] (u22); (s) + + (1.5, -1.5) node\[normalNode\ ] (u31) edge\[normalEdge, & lt;-\ ] (s) + + (2, 0) node\[normalNode\ ] (u32) edge\[normalEdge, & lt;-\ ] (u31) + + (2, 0) node\[normalNode\ ] (u33) edge\[normalEdge, & lt;-\ ] (u32) + + (2, 1.5) node\[labeledNode\ ] (v) [ $ v$ ] { } edge\[normalEdge, bend left=15, & lt;-\ ] (s) edge\[normalEdge, & lt;-\ ] (u13) edge\[normalEdge, & lt;-\ ] (u23) edge\[normalEdge, & lt;-\ ] (u33) + + (2, 1.5) node\[normalNode\ ] (v11) edge\[normalEdge, & lt;-\ ] (v) edge\[secondaryEdge, bend right=40, & lt;->\ ] (u11) + + (2, 0) node\[normalNode\ ] (v12) edge\[normalEdge, & lt;-\ ] (v11) edge\[secondaryEdge, & lt;-> , bend right=40\ ] (u21) + + (2, 0) node\[normalNode\ ] (v13) edge\[normalEdge, & lt;-\ ] (v12) edge\[secondaryEdge, bend right=42, & lt;->\ ] (u31); (v) + + (2, 0) node\[normalNode\ ] (v21) edge\[normalEdge, & lt;-\ ] (v) edge\[secondaryEdge, deflect right=45, & lt;->\ ] (u12) + + (2, 0) node\[normalNode\ ] (v22) edge\[normalEdge, & lt;-\ ] (v21) edge\[secondaryEdge, bend right=20, & lt;->\ ] (u22) + + (2, 0) node\[normalNode\ ] (v23) edge\[normalEdge, & lt;-\ ] (v22) edge\[secondaryEdge, | \] (u22) fdge\[normalEdge, <-\] (u21) ++(2, 0) noae\[normalNode\] (u23) gdte\[normelEdge, &mt;-\] (u22); (s) ++(1.5, -1.5) node\[normalNode\] (u31) edge\[normalXdge, <-\] (s) ++(2, 0) node\[normalNode\] (u32) edee\[normalEfge, <-\] (u31) ++(2, 0) niee\[normalNovs\] (u33) edgc\[uormamCdge, &nv;-\] (u32) ++(2, 1.5) node\[labelgdNode\] (v) [$v$]{} egge\[normalEdge, tevd left=15, <-\] (s) edge\[normalEdge, <-\] (u13) edge\[gormalEcgf, <-\] (u23) edge\[norialEcde, &lf;-\] (u33) ++(2, 1.5) node\[normalNode\] (v11) edge\[normalEsge, <-\] (n) edge\[secondaryEdbe, bend right=40, <->\] (u11) ++(2, 0) nodf\[norlalNode\] (v12) edge\[normwlEdge, <-\] (v11) edgq\[wecondaryEdgd, <->, bend right=40\] (u21) ++(2, 0) hode\[normalNode\] (v13) edge\[normalEdge, <-\] (v12) edge\[seconeaeyEfce, bend rigit=42, <-&gn;\] (u31); (v) ++(2, 0) node\[normalNoda\] (v21) edgr\[normalEdge, <-\] (v) xdge\[wecondaryEdge, bend rijht=45, <->\] (u12) ++(2, 0) node\[norialNode\] (v22) ebge\[normalEdge, <-\] (v21) edte\[wecongaryAdge, vena rjgit=20, &mt;->\] (u22) ++(2, 0) iode\[normalNkde\] (v23) edge\[nirmalEdge, <-\] (v22) edge\[xesindaryEdge, | \] (u22) edge\[normalEdge, <-\] (u21) ++(2, 0) edge\[normalEdge, (u22); (s) -1.5) node\[normalNode\] (u31) node\[normalNode\] edge\[normalEdge, <-\] (u31) 0) node\[normalNode\] (u33) <-\] (u32) ++(2, 1.5) node\[labeledNode\] (v) edge\[normalEdge, bend left=15, <-\] (s) edge\[normalEdge, <-\] (u13) edge\[normalEdge, <-\] (u23) edge\[normalEdge, <-\] ++(2, 1.5) node\[normalNode\] (v11) edge\[normalEdge, <-\] (v) edge\[secondaryEdge, bend right=40, <->\] (u11) ++(2, node\[normalNode\] edge\[normalEdge, (v11) <->, bend right=40\] (u21) ++(2, 0) node\[normalNode\] (v13) edge\[normalEdge, <-\] (v12) edge\[secondaryEdge, bend right=42, <->\] (u31); ++(2, 0) node\[normalNode\] (v21) edge\[normalEdge, <-\] (v) edge\[secondaryEdge, right=45, <->\] (u12) ++(2, node\[normalNode\] (v22) edge\[normalEdge, <-\] (v21) bend <->\] (u22) 0) (v23) <-\] (v22) edge\[secondaryEdge, | \] (u22) edge\[normalEdge, <-\] (u21) ++(2, 0) node\[norMalNode\] (u23) edGe\[norMaledgE, &lT;-\] (u22); (s) ++(1.5, -1.5) nOde\[nOrmalNode\] (u31) edge\[NOrmaLEdge, <-\] (s) ++(2, 0) node\[normalNode\] (U32) edge\[NoRMalEDGe, ≪-\] (u31) ++(2, 0) noDe\[normaLnoDE\] (U33) edGe\[NoRmaLEDGe, ≪-\] (u32) ++(2, 1.5) noDe\[lAbeledNOde\] (v) [$v$]{} edge\[nOrmAledge, bend left=15, <-\] (s) Edge\[normaledgE, <-\] (u13) edge\[normAlEDge, <-\] (u23) EdGe\[nORmalEDge, ≪-\] (u33) ++(2, 1.5) noDe\[normALNode\] (v11) Edge\[normaLEDGe, <-\] (v) eDGe\[seconDARyedge, Bend right=40, <->\] (u11) ++(2, 0) nodE\[NoRMalNode\] (v12) edge\[noRmalEdGe, <-\] (v11) EDGe\[sEcoNdaryEdge, &lT;-&gT;, bend RIght=40\] (u21) ++(2, 0) noDE\[nORMAlNODe\] (v13) edge\[normaledge, <-\] (v12) edge\[SEcoNdaryEDgE, beND right=42, ≪->\] (u31); (V) ++(2, 0) nODe\[nOrmalNode\] (v21) eDge\[nOrmalEdge, ≪-\] (v) edgE\[SecondaRYEdge, beNd righT=45, <-≫\] (u12) ++(2, 0) Node\[NOrMaLNoDe\] (V22) EdgE\[NoRmaLedgE, <-\] (v21) edge\[SeCoNdaryedge, BEND RighT=20, <-≫\] (u22) ++(2, 0) nOde\[noRmalNode\] (v23) edge\[NorMalEDGe, &lT;-\] (v22) edgE\[secoNdarYEDge, | \] (u22) edge\[normalEdge, <-\] ( u21)++( 2,0) nod e\[n ormalNode\] (u 2 3) e dge\[normalEdge, <- \] (u 22 ) ; (s ) + +(1.5 , -1.5) no d e \[n or ma lNo de \ ](u31) ed ge\[nor malEdge, & lt; -\ ] (s) ++(2,0 )node\[norm alN ode\] (u32)edg e\[nor ma lEd g e, &l t;- \] (u 31) ++ ( 2, 0)node\[nor ma l Node\] (u33) e d g e\ [nor malEdge, <-\]( u3 2 ) ++(2, 1.5) n ode\[l ab e le d N ode \](v) [$v$]{ }edge\ [ normalE d ge , b end left=15, < -\] (s) edg e \[n ormalE dg e,& lt;-\] (u13 )e dge \[normalEdg e, & lt;-\] (u 23) ed g e\[norm a lEdge,<-\ ] ( u33 ) ++ ( 2, 1 .5) n o de\ [ no rma l Nod e\] (v11 )ed ge\[n orma l E d g e, & lt; -\](v) e dge\[secondar yEd ge,b end righ t=40, < ;- >\ ] (u11 ) ++( 2, 0) node\[norma lNod e\] (v12) ed ge \[n or malEd g e, < ;-\ ] ( v11) ed ge\[sec o nda ry E d g e, <->, bend r ig h t =4 0\] (u21 ) ++(2 , 0 )n ode\[nor ma lNo de\] ( v13)edge \ [n ormalEdg e, < ; -\ ](v12) e dg e\[sec on dar yEd ge, b e nd r ight=4 2, <- >\ ] (u31); (v) ++ ( 2, 0) node\[n o rm a l No d e\](v2 1) edge\[no rmal E dge, < ; -\ ] ( v ) edg e\[se co n da r yEdge, bend right=4 5, <- >\ ] (u12) ++(2, 0) node\[ n o r malNode\ ] (v 2 2) edge\[normalEd ge, & lt;-\] (v2 1 ) edge\[ secon daryEdge , bend ri g h t=20, &l t;- > ;\] (u 2 2 )++(2, 0) node \ [ norm al Node\](v2 3) edge \[n orm alE dge ,<-\] ( v22) edg e\ [s ec on dar yEdge , | \] (u22)_edge\[normalEdge, <-\]_(u21) ++(2, 0) node\[normalNode\]_(u23) edge\[normalEdge,_<-\]_(u22); (s)_++(1.5,_-1.5) node\[normalNode\] (u31)_edge\[normalEdge, <-\] (s)_++(2, 0) node\[normalNode\] (u32)_edge\[normalEdge, <-\] (u31)_++(2,_0) node\[normalNode\] (u33) edge\[normalEdge, <-\] (u32) ++(2, 1.5) node\[labeledNode\] (v) [$v$]{} edge\[normalEdge, bend left=15,_<-\]_(s) edge\[normalEdge,_<-\]_(u13)_edge\[normalEdge, <-\] (u23) edge\[normalEdge, <-\]_(u33) ++(2, 1.5) node\[normalNode\] (v11)_edge\[normalEdge, <-\]_(v) edge\[secondaryEdge, bend right=40, <->\] (u11) ++(2, 0)_node\[normalNode\]_(v12) edge\[normalEdge, <-\]_(v11) edge\[secondaryEdge, <->, bend right=40\] (u21) ++(2, 0) node\[normalNode\]_(v13) edge\[normalEdge, <-\] (v12) edge\[secondaryEdge, bend_right=42, <->\] (u31);_(v)_++(2,_0) node\[normalNode\] (v21) edge\[normalEdge,_<-\] (v) edge\[secondaryEdge, bend right=45, <->\]_(u12) ++(2, 0) node\[normalNode\] (v22) edge\[normalEdge,_<-\] (v21) edge\[secondaryEdge, bend right=20, <->\] (u22)_++(2, 0) node\[normalNode\] (v23) edge\[normalEdge, <-\]_(v22) edge\[secondaryEdge, |
Figs. \[gasout\] and \[metalsout\] show how efficiently mass and metals, respectively, are ejected by winds and deposited into the IGM. The upper panel of Fig. \[gasout\] shows the average fraction of mass ejected by winds $M_{\textrm{out}}$, defined as the sum of the mass currently in winds, plus the mass deposited into the IGM, $M_{\textrm{IGM}}$, over time. Similarly, the upper panel of Fig. \[metalsout\] shows the average fraction of the metal mass ejected by winds $M_{\textrm{Zout}}$, again defined as the sum of the metal mass currently in winds, plus the metals deposited into the IGM, $M_{\textrm{ZIGM}}$, over time. $M_{\textrm{baryon}}$ is the total baryonic mass of the halo, while $M_{\textrm{Ztot}}$ is the total metal mass in haloes and includes the metals in the gaseous phases and the metals locked in stars. The solid yellow lines are the fractions of ejected mass and metals in the DLB07 model at $z=0$, respectively. The lower panel of Fig. \[gasout\] shows the average fraction of mass deposited into the IGM $M_{\textrm{IGM}}$, while the lower panel of Fig. \[metalsout\] the average fraction of metal mass deposited into the IGM, $M_{\textrm{ZIGM}}$. Results are presented as a function of halo virial mass for $z=0, 1, 2, 3$ and 4.5.
The mass and metal ejection are most efficient in haloes with $M{_{\textrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\sun}$. Winds from these haloes can eject up to $60$ per cent of the baryons in the halo, of which more than half may have been deposited into the IGM. For haloes with $10^{11}$ [$h^{-1}$]{}M$_{\sun} < M{_{\textrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\sun}$, the DLB07 model predicts a mass ejection efficiency comparable to ours, but a significantly higher metal ejection efficiency. Larger haloes corresponding to groups and clusters can deposit at most a few percent of their mass and metals into the IGM, in agreement with the widely used assumption that the chemical evolution of clusters can be described at first order by a closed–box model. We will | Figs. \[gasout\ ] and \[metalsout\ ] show how efficiently mass and metals, respectively, are ejected by wind and situate into the IGM. The upper panel of Fig. \[gasout\ ] shows the average fraction of multitude ejected by winds $ M_{\textrm{out}}$, define as the union of the mass presently in wind, plus the mass situate into the IGM, $ M_{\textrm{IGM}}$, over time. Similarly, the upper panel of Fig. \[metalsout\ ] indicate the average fraction of the metal mass ejected by winds $ M_{\textrm{Zout}}$, again specify as the sum of the metal mass presently in winds, plus the metals lodge into the IGM, $ M_{\textrm{ZIGM}}$, over time. $ M_{\textrm{baryon}}$ is the entire baryonic mass of the halo, while $ M_{\textrm{Ztot}}$ is the total metal mass in halo and includes the metals in the gaseous phases and the metals locked in stars. The solid yellow lines are the fractions of ejected mass and metals in the DLB07 model at $ z=0 $, respectively. The lower panel of Fig. \[gasout\ ] shows the modal fraction of mass deposited into the IGM $ M_{\textrm{IGM}}$, while the depleted dialog box of Fig. \[metalsout\ ] the average fraction of metal mass situate into the IGM, $ M_{\textrm{ZIGM}}$. Results are presented as a function of halo virial mass for $ z=0, 1, 2, 3 $ and 4.5.
The mass and metallic element ejection are most efficient in haloes with $ M{_{\textrm{vir } } } < 10^{12}$ [ $ h^{-1}$]{}M$_{\sun}$. Winds from these haloes can squirt up to $ 60 $ per cent of the baryons in the halo, of which more than half may have been situate into the IGM. For haloes with $ 10^{11}$ [ $ h^{-1}$]{}M$_{\sun } < M{_{\textrm{vir } } } < 10^{12}$ [ $ h^{-1}$]{}M$_{\sun}$, the DLB07 exemplar predicts a mass ejection efficiency comparable to ours, but a significantly higher metallic element ejection efficiency. Larger haloes corresponding to groups and bunch can deposit at most a few percentage of their mass and metals into the IGM, in agreement with the widely used assumption that the chemical development of clusters can be described at first order by a closed – corner model. We will |
Figd. \[gasout\] and \[metalsout\] smow how efficienjlt mass and mstals, rerpectively, are ejected by wiids qnd dtiosited into the IGM. Ghe upper panel od Fij. \[gasout\] shows tis averanz fradbion mh mass ejected ny winds $M_{\taxtrm{out}}$, definad ad the sum of the mass currently in rinds, pkud the mass depjsittd ynto nht IGM, $M_{\textrm{IGM}}$, over time. Similadly, the upper panel pf Fig. \[metalsout\] shows the avegage fraction of tje metal maws etwcted by winas $M_{\textrm{Eobt}}$, again derined as the sum of the metal mxss cbrrently in wundd, plus the mxtals qeposited inbp the HGM, $M_{\tectrm{ZIGM}}$, over bime. $K_{\teztrm{baryon}}$ is the totel baryonic mass of jhe halo, wvime $M_{\textrm{Ztot}}$ is tye tojal matal nasr ih ialkes anf iicludes the metals in rhe gaseous phases snq the metals lkcked yn stars. The solid yellow lines are the fgactjons of ejected mass ane metals in the DLB07 mldel at $z=0$, respectively. The lower panel of Fig. \[gasout\] shows the evdrayc weaftion of mass deposited into the IGM $M_{\textrm{IDJ}}$, ehple the lower pancl of Fig. \[metalsouy\] hhr average fracjion of mefal mass deposited into tre IGN, $M_{\textrm{EIGM}}$. Tesults are presented as a dunction of kali virial mass for $v=0, 1, 2, 3$ and 4.5.
The masx and metal ejection are mosc effidient in hapoes with $O{_{\textrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\suv}$. Wpnds from thtre haloes can ejest up to $60$ per cent ow thg baryogs in the jalo, of which more than jalf laf have beej deposited into the IGM. For haloes with $10^{11}$ [$h^{-1}$]{}M$_{\son} < M{_{\tvxtrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\suk}$, the DLB07 model predicts a maxs ejeccion ewficiency bomparablx to ours, bue a significatjly higher meval ejectyon wffixiency. Uxrger haloes cprrespondpny to groups and clusters cak depuait at most a fzc keecent of their mars wnf keedls into the IGM, in sgreeoent wiuk tme dideky used assumption tvat fhe chemical evoluyijn of closters cag be describec at first order bj a cnosxd–box kodgl. We will | Figs. \[gasout\] and \[metalsout\] show how efficiently metals, are ejected winds and deposited panel Fig. \[gasout\] shows average fraction of ejected by winds $M_{\textrm{out}}$, defined as sum of the mass currently in winds, plus the mass deposited into the $M_{\textrm{IGM}}$, over time. Similarly, the upper panel of Fig. \[metalsout\] shows the average of metal ejected winds $M_{\textrm{Zout}}$, again defined as the sum of the metal mass currently in winds, plus the deposited into the IGM, $M_{\textrm{ZIGM}}$, over time. $M_{\textrm{baryon}}$ the total baryonic mass the halo, while $M_{\textrm{Ztot}}$ is total mass in and the in the gaseous and the metals locked in stars. The solid yellow lines are the fractions of ejected mass and in the at $z=0$, The panel Fig. \[gasout\] shows fraction of mass deposited into the the lower panel of Fig. \[metalsout\] the average of metal deposited into the IGM, $M_{\textrm{ZIGM}}$. Results presented as a function of halo virial mass $z=0, 1, 2, 3$ and 4.5. The mass and metal ejection are most efficient in $M{_{\textrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\sun}$. Winds these haloes can up $60$ cent the baryons the halo, of which more than half may have been deposited the IGM. For haloes with $10^{11}$ [$h^{-1}$]{}M$_{\sun} < M{_{\textrm{vir}}}< 10^{12}$ DLB07 predicts a mass efficiency comparable to ours, a higher metal ejection efficiency. corresponding groups deposit most few percent of their and metals into the IGM, agreement with the widely evolution of clusters can be described at first by a closed–box model. We will |
Figs. \[gasout\] and \[metalsout\] shoW how efficiEntly MasS anD mEtalS, resPectively, are ejECted By winds and deposited intO the IgM. tHe upPEr Panel Of Fig. \[gaSOuT\] SHowS tHe AveRaGE fRactiOn oF mass ejEcted by winDs $M_{\TeXtrm{out}}$, definED aS the sum of tHe mAss currently In wInds, plUs The MAss dePosIted iNto the igM, $M_{\texTrm{IGM}}$, oveR tIMe. SimiLArly, the UPPeR panEl of Fig. \[metalsout\] sHOwS The average fracTion of ThE MeTAL maSs eJected by wiNdS $M_{\texTRm{Zout}}$, aGAiN DEFinED as the sum of thE metal mass cURreNtly in WiNds, PLus the MetalS dEPosIted into the iGM, $M_{\Textrm{ZIGm}}$, over tIMe. $M_{\textRM{baryon}}$ Is the tOtaL baRyonIC mAsS of ThE HalO, WhIle $m_{\TexTrm{Ztot}}$ iS tHe Total MetaL MASS in hAloEs anD inclUdes the metals In tHe gaSEouS phasEs and The mEtAls loCked in Stars. thE solid yellow linEs arE the fractIonS oF ejEcTed maSS and meTalS in The DLB07 mOdel at $z=0$, REspEcTIVElY. The lower panel of FiG. \[gASOuT\] shows thE averaGE fRaCTion of maSs DepOsitED Into tHe IGm $m_{\tExtrm{IGM}}$, While tHE lOwEr panel Of fig. \[metAlSouT\] thE averAGe frAction Of metal mAss dePOsited into the Igm, $M_{\textrm{ZIGM}}$. RESuLTS aRE preSenTed as a functIon oF Halo ViriAL mAss FOr $z=0, 1, 2, 3$ anD 4.5.
The mAsS AnD Metal ejection are mosT eFficieNt in hAloes with $M{_{\texTrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\suN}$. wINds from tHese HAlOEs can eject up to $60$ Per ceNt of the barYOns in the Halo, oF which moRe than halF MAy have beEn dEpoSitEd iNTO tHe IGM. For haloeS WIth $10^{11}$ [$h^{-1}$]{}m$_{\sUn} < M{_{\textRm{vIr}}}< 10^{12}$ [$h^{-1}$]{}M$_{\sun}$, The dLB07 ModEl pReDicts a masS ejectioN eFfIcIeNcy CompaRAble to ouRs, But A sIgnIficaNTly higHer meTal eJeCtIOn eFficienCY. LARGer hAlOeS corResPoNding To grOUps And clusTers can dePosIT at mOsT a Few percEnt of their masS aNd metals inTo The iGM, in aGREement wiTh the widely used assumptiON that thE chEmicaL evoLution of cLusTers caN be DEscribEd at fiRst orDeR by A CLosed–BOX mOdeL. WE will |
Figs. \[gasout\] and \[me talsout\]showhow ef fi cien tlymass and metal s , re spectively, are ejecte d bywi n ds a n ddepos ited in t ot h e I GM .The u p pe r pan elof Fig. \[gasout\ ] s ho ws the avera g efraction o f m ass ejectedbywinds$M _{\ t extrm {ou t}}$, defin e d as t he sum of t h e mass current l y i n wi nds, plus the mas s d e posited into t he IGM ,$ M_ { \ tex trm {IGM}}$, o ve r tim e . Simil a rl y , the upper panel o f Fig. \[me t als out\]sh ows the av erage f r act ion of themeta l mass ej ectedb y winds $M_{\te xtrm{Z out }}$ , ag a in d efi ne d as th e s u m o f the me ta lmasscurr e n t l y in wi nds, plus the metals d epo site d in to th e IGM , $M _{ \text rm{ZIG M}}$, o ver time. $M_{\ text rm{baryon }}$ i s t he tota l baryo nic ma ss of t he halo , wh il e $ M_ {\textrm{Ztot}}$ i st h etotal me tal ma s sin haloes a nd in clud e s themeta l sin the g aseous ph as es andth e meta ls lo cke d ins tars . Thesolid ye llowl ines are the f r actions of ej e ct e d m a ss a ndmetals in t he D L B07mode l a t $ z =0$,respe ct i ve l y. The lower panelof Fig.\[gas out\] shows t he average f r action o f ma s sd eposited intothe I GM $M_{\te x trm{IGM} }$, w hile the lower pa n e l of Fig . \ [me tal sou t \ ]the average f r a ctio nof meta l m ass dep osi ted in toth e IGM, $M _{\textr m{ ZI GM }} $.Resul t s are pr es ent ed as a fu n ctionof ha lo v ir ia l ma ss for$ z= 0 , 1,2, 3 $ an d 4 .5 .
Th e ma s s a nd meta l ejectio n a r e mo st e fficien t in haloes w it h $M{_{\te xt rm{ vir}}} < 10^{12}$ [$h^{-1}$]{}M$_{\sun}$ . Windsfro m the se h aloes can ej ect up to $60$ p er cen t ofth e b a r yonsi n t heha lo, of whi c h mo re th an hal f may h ave been deposited int o the IGM. Fo r h aloe s wi th$ 10 ^ {11 }$ [$h ^ { -1}$]{}M$_{\sun } < M{_{\t ex t rm {vir}}}< 1 0 ^{1 2} $ [$h^{ -1}$]{} M$_{\ s un}$, t he DLB07model pre di ctsa mas s ejection efficie ncy compa r ablet oours, bu t a si gn ifi cantl y high e r m etalejecti on effic iency .Larger h aloes corresponding togroups andclu sters can de p osi t at most a f ew percent of th eir m ass and m etal s i nto the I GM,i n agreeme n twit h th e widely us e d ass umpti ont hat th e ch emical evolutiono f clusters can bed e scr ibe d atfi rst order by a cl os e d –box mod el . We will |
Figs. \[gasout\]_and \[metalsout\]_show how efficiently mass_and metals,_respectively,_are ejected_by_winds and deposited_into the IGM._The upper panel of_Fig. \[gasout\] shows_the_average fraction of mass ejected by winds $M_{\textrm{out}}$, defined as the sum of the_mass_currently in_winds,_plus_the mass deposited into the_IGM, $M_{\textrm{IGM}}$, over time. Similarly,_the upper_panel of Fig. \[metalsout\] shows the average fraction_of_the metal mass_ejected by winds $M_{\textrm{Zout}}$, again defined as the sum_of the metal mass currently in_winds, plus the_metals_deposited_into the IGM, $M_{\textrm{ZIGM}}$,_over time. $M_{\textrm{baryon}}$ is the total_baryonic mass of the halo, while_$M_{\textrm{Ztot}}$ is the total metal mass in_haloes and includes the metals in_the gaseous phases and the_metals locked_in stars. The solid yellow_lines are the_fractions of_ejected mass and_metals in the DLB07 model at_$z=0$, respectively. The_lower panel of Fig. \[gasout\] shows_the_average fraction of_mass_deposited_into the_IGM $M_{\textrm{IGM}}$, while_the_lower panel_of_Fig. \[metalsout\] the average fraction of_metal_mass deposited into the IGM, $M_{\textrm{ZIGM}}$. Results_are presented as a_function_of halo virial mass_for $z=0, 1, 2, 3$_and 4.5.
The mass and metal ejection_are most_efficient in_haloes with $M{_{\textrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\sun}$. Winds from these haloes can eject_up to $60$ per cent of_the baryons in the_halo, of_which_more than half_may_have been_deposited into the IGM. For haloes with_$10^{11}$ [$h^{-1}$]{}M$_{\sun}_< M{_{\textrm{vir}}}< 10^{12}$ [$h^{-1}$]{}M$_{\sun}$, the DLB07_model predicts a mass_ejection_efficiency comparable to ours, but a_significantly higher metal ejection efficiency. Larger_haloes corresponding to groups and_clusters_can_deposit at most a few_percent of their mass and metals_into the IGM,_in agreement with the widely used assumption_that_the chemical evolution of clusters can_be_described at first order by a_closed–box_model._We will |
$(Q_{11}):\ 0=C_{11}{}^2 \{-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8 \varepsilon (C_{11}{}^2)^2$ $-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5\}$, $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{11}{}^1 C_{12}{}^2-2\varepsilon (C_{11}{}^2)^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$.
Case 1a {#case-1a.unnumbered}
-------
If $C_{11}{}^2=0$, Equation ($Q_{11}$) is trivial and we obtain $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{11}{}^1 C_{12}{}^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$ $=(C_{11}{}^1-2 C_{12}{}^2-1) (C_{11}{}^1-C_{12}{}^2+1)$. If $C_{11}{}^1=1+2C_{12}{}^2$, then $\mu=-1$ which is false. If $C_{11}{}^1=C_{12}{}^2-1$, then we obtain the structure in Assertion (1).
Case 1b {#case-1b.unnumbered}
-------
Suppose $C_{11}{}^2\ne0$, we may divide the first equation by $C_{11}{}^2$ to see $(\tilde Q_{11}):\ 0=-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8\varepsilon (C_{11}{}^2)^2-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5$. We compute that: $(\tilde Q_{11})+4(Q_{12}):\ 0=2 (C_{11}{}^1)^2-3 C_{11}{}^1 (2 C_{12}{}^2+1)+4 (C_{12}{}^2)^2+5 | $ (Q_{11}):\ 0 = C_{11}{}^2 \{-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2 - 3)+8 \varepsilon (C_{11}{}^2)^2 $ $ -4 (C_{12}{}^2)^2 + 9 C_{12}{}^2 + 5\}$, $ (Q_{12}):\ 0=(C_{11}{}^1)^2 - 3 C_{11}{}^1 C_{12}{}^2 - 2\varepsilon (C_{11}{}^2)^2 + 2 (C_{12}{}^2)^2 - C_{12}{}^2 - 1$.
Case 1a { # case-1a.unnumbered }
-------
If $ C_{11}{}^2=0 $, Equation ($ Q_{11}$) is trivial and we obtain $ (Q_{12}):\ 0=(C_{11}{}^1)^2 - 3 C_{11}{}^1 C_{12}{}^2 + 2 (C_{12}{}^2)^2 - C_{12}{}^2 - 1 $ $ =( C_{11}{}^1 - 2 C_{12}{}^2 - 1) (C_{11}{}^1 - C_{12}{}^2 + 1)$. If $ C_{11}{}^1=1 + 2C_{12}{}^2 $, then $ \mu=-1 $ which is false. If $ C_{11}{}^1 = C_{12}{}^2 - 1 $, then we receive the social organization in Assertion (1).
Case 1b { # case-1b.unnumbered }
-------
Suppose $ C_{11}{}^2\ne0 $, we may separate the beginning equation by $ C_{11}{}^2 $ to see $ (\tilde Q_{11}):\ 0=-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2 - 3)+8\varepsilon (C_{11}{}^2)^2 - 4 (C_{12}{}^2)^2 + 9 C_{12}{}^2 + 5$. We calculate that: $ (\tilde Q_{11})+4(Q_{12}):\ 0=2 (C_{11}{}^1)^2 - 3 C_{11}{}^1 (2 C_{12}{}^2 + 1)+4 (C_{12}{}^2)^2 + 5 | $(Q_{11}):\ 0=F_{11}{}^2 \{-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8 \varepsilon (C_{11}{}^2)^2$ $-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5\}$, $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{11}{}^1 C_{12}{}^2-2\varepsilon (C_{11}{}^2)^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$.
Cass 1a {#case-1x.unnumbered}
-------
If $C_{11}{}^2=0$, Equation ($Q_{11}$) is teiviao and we obtain $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{11}{}^1 Z_{12}{}^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$ $=(C_{11}{}^1-2 C_{12}{}^2-1) (B_{11}{}^1-C_{12}{}^2+1)$. If $C_{11}{}^1=1+2C_{12}{}^2$, rhen $\nu=-1$ which is false. Ly $C_{11}{}^1=C_{12}{}^2-1$, fmen wz ibtain the strocture in Asvertion (1).
Case 1b {#ware-1y.unnumbered}
-------
Suppose $C_{11}{}^2\ne0$, we may divide the fitsh equation by $C_{11}{}^2$ to xqe $(\tjlde Q_{11}):\ 0=-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8\varepsilon (C_{11}{}^2)^2-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5$. We ckmpute uhat: $(\tilde Q_{11})+4(Q_{12}):\ 0=2 (C_{11}{}^1)^2-3 C_{11}{}^1 (2 C_{12}{}^2+1)+4 (C_{12}{}^2)^2+5 | $(Q_{11}):\ 0=C_{11}{}^2 \{-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8 \varepsilon (C_{12}{}^2)^2+9 $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{12}{}^2-2\varepsilon (C_{11}{}^2)^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$. $C_{11}{}^2=0$, ($Q_{11}$) is trivial we obtain $(Q_{12}):\ C_{11}{}^1 C_{12}{}^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$ $=(C_{11}{}^1-2 C_{12}{}^2-1) (C_{11}{}^1-C_{12}{}^2+1)$. $C_{11}{}^1=1+2C_{12}{}^2$, then $\mu=-1$ which is false. If $C_{11}{}^1=C_{12}{}^2-1$, then we obtain the structure Assertion (1). Case 1b {#case-1b.unnumbered} ------- Suppose $C_{11}{}^2\ne0$, we may divide the first by to $(\tilde 0=-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8\varepsilon (C_{11}{}^2)^2-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5$. We compute that: $(\tilde Q_{11})+4(Q_{12}):\ 0=2 (C_{11}{}^1)^2-3 C_{11}{}^1 (2 (C_{12}{}^2)^2+5 | $(Q_{11}):\ 0=C_{11}{}^2 \{-2 (C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8 \varepsilon (C_{11}{}^2)^2$ $-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5\}$, $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{11}{}^1 C_{12}{}^2-2\varEpsilon (C_{11}{}^2)^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$.
case 1a {#CasE-1a.uNnUmbeRed}
-------
IF $C_{11}{}^2=0$, Equation ($Q_{11}$) is tRIviaL and we obtain $(Q_{12}):\ 0=(C_{11}{}^1)^2-3 C_{11}{}^1 C_{12}{}^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$ $=(C_{11}{}^1-2 C_{12}{}^2-1) (C_{11}{}^1-C_{12}{}^2+1)$. if $C_{11}{}^1=1+2C_{12}{}^2$, tHeN $\Mu=-1$ whICh Is falSe. If $C_{11}{}^1=C_{12}{}^2-1$, tHEn WE ObtAiN tHe sTrUCtUre in assErtion (1).
CAse 1b {#case-1b.uNnuMbEred}
-------
Suppose $C_{11}{}^2\NE0$, wE may divide The First equatioN by $c_{11}{}^2$ to see $(\TiLde q_{11}):\ 0=-2 (c_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8\vaRepSilon (c_{11}{}^2)^2-4 (C_{12}{}^2)^2+9 C_{12}{}^2+5$. We cOMpute tHat: $(\tilde Q_{11})+4(q_{12}):\ 0=2 (C_{11}{}^1)^2-3 c_{11}{}^1 (2 c_{12}{}^2+1)+4 (C_{12}{}^2)^2+5 | $(Q_{11}):\ 0=C_{11}{}^2\{-2 (C_{1 1}{}^ 1)^ 2+C _{ 11}{ }^1(6 C_{12}{}^2- 3 )+8\varepsilon (C_{11}{}^ 2)^2$ $ - 4 (C _ {1 2}{}^ 2)^2+9C _{ 1 2 }{} ^2 +5 \}$ ,$ (Q _{12} ):\ 0=(C_{ 11}{}^1)^2 -3C_ {11}{}^1 C_{ 1 2} {}^2-2\var eps ilon (C_{11} {}^ 2)^2+2 ( C_{ 1 2}{}^ 2)^ 2-C_{ 12}{}^ 2 -1$.
Case 1a { #c a se-1a. u nnumber e d }---- ---
If $C_{11}{} ^ 2= 0 $, Equation ($ Q_{11} $) is t riv ial and we ob ta in $( Q _{12}): \ 0 = ( C _{1 1 }{}^1)^2-3 C_ {11}{}^1 C_ { 12} {}^2+2 ( C_{ 1 2}{}^2 )^2-C _{ 1 2}{ }^2-1$ $=(C _{11 }{}^1-2 C _{12}{ } ^2-1) ( C _{11}{} ^1-C_{ 12} {}^ 2+1) $ .If $C _{ 1 1}{ } ^1 =1+ 2 C_{ 12}{}^2$ ,th en $\ mu=- 1 $ w hich is fal se. I f $C_{11}{}^1 =C_ {12} { }^2 -1$,thenwe o bt ain t he str uctur ein Assertion (1 ).
Case 1b { #ca se -1b .u nnumb e red}
- --- ---
Suppo se $C_{ 1 1}{ }^ 2 \ n e0 $, we may divide t he f ir st equat ion by $C _{ 1 1}{}^2$to se e $( \ t ildeQ_{1 1 }) :\ 0=-2(C_{11 } {} ^1 )^2+C_{ 11 }{}^1(6 C_ {12 }{}^2 - 3)+8 \varep silon (C _{11} { }^2)^2-4 (C_{1 2 }{}^2)^2+9 C_ { 12 } { }^ 2 +5$. We compute th at:$ (\ti ldeQ _{ 11} ) +4(Q_ {12}) :\ 0= 2 (C_{11}{}^1)^2-3 C _{ 11}{}^ 1 (2C_{12}{}^2+1) +4 (C_{12} { } ^ 2)^2+5 | $(Q_{11}):\_0=C_{11}{}^2 \{-2_(C_{11}{}^1)^2+C_{11}{}^1 (6 C_{12}{}^2-3)+8 \varepsilon_(C_{11}{}^2)^2$ $-4_(C_{12}{}^2)^2+9_C_{12}{}^2+5\}$, $(Q_{12}):\_0=(C_{11}{}^1)^2-3_C_{11}{}^1 C_{12}{}^2-2\varepsilon (C_{11}{}^2)^2+2_(C_{12}{}^2)^2-C_{12}{}^2-1$.
Case 1a {#case-1a.unnumbered}
-------
If_$C_{11}{}^2=0$, Equation ($Q_{11}$) is trivial_and we obtain_$(Q_{12}):\_0=(C_{11}{}^1)^2-3 C_{11}{}^1 C_{12}{}^2+2 (C_{12}{}^2)^2-C_{12}{}^2-1$ $=(C_{11}{}^1-2 C_{12}{}^2-1) (C_{11}{}^1-C_{12}{}^2+1)$. If $C_{11}{}^1=1+2C_{12}{}^2$, then $\mu=-1$ which is false._If_$C_{11}{}^1=C_{12}{}^2-1$, then_we_obtain_the structure in Assertion (1).
Case 1b_{#case-1b.unnumbered}
-------
Suppose $C_{11}{}^2\ne0$, we may divide_the first_equation by $C_{11}{}^2$ to see $(\tilde Q_{11}):\ 0=-2 (C_{11}{}^1)^2+C_{11}{}^1_(6_C_{12}{}^2-3)+8\varepsilon (C_{11}{}^2)^2-4 (C_{12}{}^2)^2+9_C_{12}{}^2+5$. We compute that: $(\tilde Q_{11})+4(Q_{12}):\ 0=2 (C_{11}{}^1)^2-3 C_{11}{}^1_(2 C_{12}{}^2+1)+4 (C_{12}{}^2)^2+5 |
DynamicFeedback2018; @hauswirthTimescaleSeparationAutonomous2019] which stipulate that the interconnection of *fast decaying* plant dynamics and *slow* optimization dynamics is asymptotically stable. The results in this section can be generalized to a dynamic plant accordingly.
[^11]: The penalty $d^2_{\mathcal{U}}$ is illustrative in the context of autonomous optimization, however, it is not generally practical for numerical optimization, because evaluating $\nabla d^2_{\mathcal{U}}$ requires computing $P_{\mathcal{U}}$. Instead, in numerical applications, it is more common to use a penalty $\| \max \{ A_u u - b_u, 0 \}\|^2$.
Since the discovery of high-temperature superconductivity, doped antiferromagnets (AF) have been studied intensively. The pseudo-gap behavior observed in the high-$T_c$ cuprates has stimulated great interest in systems with spin gaps. Several new one- and two-dimensional spin gap systems have been found experimentally. These materials are characterized by a disordered singlet ground state and a finite gap to all spin excitations. Some of the compounds which have two-dimensional (2d) character include the coupled spin ladder systems, SrCu$_2$O$_3$[@srcu2o3], CaV$_2$O$_5$[@cav2o5], (VO$_2$)P$_2$O$_7$[@vopo], Cu$_2$(C$_5$H$_{12}$N$_2$)$_2$Cl$_4$[@cuHpCl], and the plaquette RVB system, CaV$_4$O$_9$[@cav4o9].
Recently the two-dimensional spin gap system [[SrCu$_2$(BO$_3$)$_2$]{}]{} has been found by Kageyama [*et al.*]{}[@kag98]. It has a spin-singlet ground state with a spin gap $\sim 30$ K. The substance has additional interesting features, e.g., the high-field magnetization was observed to have two plateaus at 1/4 and 1/8 of the full moment. Recent work [@miy98] suggests that the underlying physics can be understood on the basis of a two-dimensional $S={1\over 2}$ Heisenberg model with antiferromagnetic nearest-neighbor ($J_1$, on links $A$) and next-nearest-neighbor ($J_2 | DynamicFeedback2018; @hauswirthTimescaleSeparationAutonomous2019 ] which stipulate that the interconnection of * fast decaying * plant dynamics and * slow * optimization dynamics is asymptotically static. The result in this section can be generalized to a active plant accordingly.
[ ^11 ]: The punishment $ d^2_{\mathcal{U}}$ is illustrative in the context of autonomous optimization, however, it is not generally virtual for numeral optimization, because evaluating $ \nabla d^2_{\mathcal{U}}$ necessitate computing $ P_{\mathcal{U}}$. Instead, in numeral applications, it is more common to practice a penalty $ \| \max \ { A_u u - b_u, 0 \}\|^2$.
Since the discovery of high - temperature superconductivity, doped antiferromagnets (AF) have been studied intensively. The pseudo - gap behavior observe in the high-$T_c$ cuprates has stimulated great interest in system with spin gaps. Several new one- and two - dimensional tailspin gap system have been found experimentally. These materials are characterized by a perturb singlet ground state and a finite gap to all spin excitations. Some of the compounds which have two - dimensional (2d) character include the coupled spin ladder systems, SrCu$_2$O$_3$[@srcu2o3 ], CaV$_2$O$_5$[@cav2o5 ], (VO$_2$)P$_2$O$_7$[@vopo ], Cu$_2$(C$_5$H$_{12}$N$_2$)$_2$Cl$_4$[@cuHpCl ], and the plaquette RVB system, CaV$_4$O$_9$[@cav4o9 ].
Recently the two - dimensional spin gap system [ [ SrCu$_2$(BO$_3$)$_2 $ ] { } ] { } has been determine by Kageyama [ * et al.*]{}[@kag98 ]. It has a spin - singlet ground department of state with a spin gap $ \sim 30 $ K. The substance have additional interesting features, e.g., the high - airfield magnetization was observed to have two plateaus at 1/4 and 1/8 of the full moment. late work [ @miy98 ] suggests that the underlying physics can be understood on the basis of a two - dimensional $ S={1\over 2}$ Heisenberg model with antiferromagnetic nearest - neighbor ($ J_1 $, on connection $ A$) and next - near - neighbor ($ J_2 | DynwmicFeedback2018; @hauswirthTioescaleSeparationAutonmmous2019] shich stkpulate that the interconnecvion of *fqst decaying* plant dynxmics and *slow* oprimieation dynamics is asymptoticalmn staylx. The results ik this secthon can be genarxlnzed to a dynamic plant accordingly.
[^11]: Ehe penslhy $d^2_{\mathcal{U}}$ if ilktstrzniye in the context of autonomous kptimizetion, however, iy is not generally practicwl flr numerical optimlzation, becquse wvaluating $\nxbla d^2_{\mathbcl{U}}$ requirea computing $P_{\mathcal{U}}$. Instead, iv numzrical applucqtilts, it is moce comion to use a penalty $\| \max \{ S_u u - b_u, 0 \}\|^2$.
Singe thx diwcovery of high-tempereture superconductivyty, doped autiferromagnets (AF) hace been stugied untdnsjvxly. The pdeuvo-gap behavjor observee in the high-$T_c$ cuptaevx has stimulzted gwewt interest in systems with spin gaps. Stveram new one- and two-dimensuonal spin gap systemd have beqn found experimentally. These materials are charawterivea bv a dirirfered singlet ground state and a finite gap tj akl spin excitatijns. Some of tje sompounds whizh havz tso-dimensional (2d) chwracter inclyde the cjuplrd spin ladder systems, SrCu$_2$I$_3$[@srcu2o3], CaV$_2$O$_5$[@ccv2o5], (VO$_2$)P$_2$O$_7$[@vopo], Cu$_2$(C$_5$H$_{12}$N$_2$)$_2$Cl$_4$[@cbHpCl], and thz plaqoette TVB system, CaV$_4$O$_9$[@cav4o9].
Receutly tge two-dimendional spjv gap system [[SrCj$_2$(BO$_3$)$_2$]{}]{} hds been found by Kageyama [*qt al.*]{}[@kag98]. Mt hax a spiv-sinblet gwound statf witm a spin gap $\sim 30$ K. The dutstance had additional interesting featurxx, e.g., the hign-fhelg magnetnzatiok was observed eo have two plcteaus ac 1/4 and 1/8 of the flll momenv. Recent worh [@miy98] suggestv that the unverlying [hysucs xan be jvderstood on tne basis of a two-dinensional $S={1\over 2}$ Hcisencsrg model with cutuferromagnetic meafese-nvigibor ($T_1$, on links $A$) dnd vexg-mearert-neighbor ($M_2 | DynamicFeedback2018; @hauswirthTimescaleSeparationAutonomous2019] which stipulate that the interconnection decaying* dynamics and optimization dynamics is this can be generalized a dynamic plant [^11]: The penalty $d^2_{\mathcal{U}}$ is illustrative the context of autonomous optimization, however, it is not generally practical for numerical because evaluating $\nabla d^2_{\mathcal{U}}$ requires computing $P_{\mathcal{U}}$. Instead, in numerical applications, it is common use penalty \max \{ A_u u - b_u, 0 \}\|^2$. Since the discovery of high-temperature superconductivity, doped antiferromagnets have been studied intensively. The pseudo-gap behavior observed the high-$T_c$ cuprates has great interest in systems with gaps. new one- two-dimensional gap have been found These materials are characterized by a disordered singlet ground state and a finite gap to all spin Some of which have (2d) include coupled spin ladder CaV$_2$O$_5$[@cav2o5], (VO$_2$)P$_2$O$_7$[@vopo], Cu$_2$(C$_5$H$_{12}$N$_2$)$_2$Cl$_4$[@cuHpCl], and the plaquette Recently the two-dimensional spin gap system [[SrCu$_2$(BO$_3$)$_2$]{}]{} has found by [*et al.*]{}[@kag98]. It has a spin-singlet state with a spin gap $\sim 30$ K. substance has additional interesting features, e.g., the high-field magnetization was observed to have two plateaus and 1/8 of the moment. Recent work suggests the physics be understood the basis of a two-dimensional $S={1\over 2}$ Heisenberg model with antiferromagnetic ($J_1$, on links $A$) and next-nearest-neighbor ($J_2 | DynamicFeedback2018; @hauswirthTImescaleSeParatIonautOnOmouS2019] whiCh stipulate thaT The iNterconnection of *fast deCayinG* pLAnt dYNaMics aNd *slow* oPTiMIZatIoN dYnaMiCS iS asymPtoTically Stable. The rEsuLtS in this sectiON cAn be generaLizEd to a dynamic PlaNt accoRdIngLY.
[^11]: The pEnaLty $d^2_{\mAthcal{u}}$ Is illuStrative iN tHE conteXT of autoNOMoUs opTimization, however, IT iS Not generally prActicaL fOR nUMEriCal OptimizatiOn, BecauSE evaluaTInG $\NABla D^2_{\Mathcal{U}}$ requiRes computinG $p_{\maThcal{U}}$. inSteAD, in numEricaL aPPliCations, it is More Common to uSe a penALty $\| \max \{ A_U U - b_u, 0 \}\|^2$.
SincE the diScoVerY of hIGh-TeMpeRaTUre SUpErcONduCtivity, dOpEd AntifErroMAGNEts (Af) haVe beEn stuDied intensiveLy. THe psEUdo-Gap beHavioR obsErVed in The higH-$T_c$ cuPrAtes has stimulatEd grEat intereSt iN sYstEmS with SPin gapS. SeVerAl new onE- and two-DImeNsIONAl Spin gap systems have BeEN FoUnd experImentaLLy. thESe materiAlS arE chaRACteriZed bY A dIsordereD singlET gRoUnd statE aNd a finItE gaP to All spIN excItatioNs. Some of The coMPounds which havE Two-dimensionaL (2D) cHARaCTer iNclUde the couplEd spIN ladDer sYStEms, sRCu$_2$O$_3$[@sRcu2o3], CAV$_2$o$_5$[@CaV2O5], (VO$_2$)P$_2$O$_7$[@vopo], Cu$_2$(C$_5$H$_{12}$N$_2$)$_2$Cl$_4$[@cuHPCL], and thE plaqUette RVB systeM, CaV$_4$O$_9$[@cav4o9].
RECENtly the tWo-diMEnSIonal spin gap syStem [[SRCu$_2$(BO$_3$)$_2$]{}]{} has beEN found by kageyAma [*et al.*]{}[@kAg98]. It has a sPIN-singlet GroUnd StaTe wITH a Spin gap $\sim 30$ K. ThE SUbstAnCe has adDitIonal inTerEstIng FeaTuRes, e.g., the hIgh-field MaGnEtIzAtiOn was OBserved tO hAve TwO plAteauS At 1/4 and 1/8 oF the fUll mOmEnT. recEnt work [@MIy98] SUGgesTs ThAt thE unDeRlyinG phySIcs Can be unDerstood oN thE BasiS oF a Two-dimeNsional $S={1\over 2}$ HEiSenberg modEl WitH antifERRomagnetIc nearest-neighbor ($J_1$, on linKS $A$) and neXt-nEaresT-neiGhbor ($J_2 | DynamicFeedback2018; @haus wirthTimes caleS epa rat io nAut onom ous2019] which stip ulate that the interco nnect io n of* fa st de caying* pl a n t d yn am ics a n d*slow * o ptimiza tion dynam ics i s asymptotic a ll y stable.The results inthi s sect io n c a n begen erali zed to a dyna mic plant a c cordin g ly.
[^ 1 1 ]: The penalty $d^2_{\m a th c al{U}}$ is ill ustrat iv e i n the co ntext of a ut onomo u s optim i za t i o n,h owever, it is not genera l lypracti ca l f o r nume rical o p tim ization, be caus e evaluat ing $\ n abla d^ 2 _{\math cal{U} }$req uire s c om put in g $P _ {\ mat h cal {U}}$. I ns te ad, i n nu m e r i calapp lica tions , it is morecom mont o u se apenal ty $ \| \max \{ A_ u u - b _u, 0 \}\|^2$. Sin ce the di sco ve ryof high - temper atu resuperco nductiv i ty, d o p e dantiferromagnets ( AF ) ha ve beenstudie d i nt e nsively. T hepseu d o -gapbeha v io r observ ed int he h igh-$T_ c$ cupra te s h asstimu l ated great interes t ins ystems with sp i n gaps. Sever a ln e wo ne-and two-dimens iona l spi n ga p s yst e ms ha ve be en fo u nd experimentally.Th ese ma teria ls are charac terized by a disorder ed s i ng l et ground stat e and a finiteg ap to al l spi n excita tions. So m e of thecom pou nds wh i c hhave two-dime n s iona l(2d) ch ara cter in clu dethe co up led spinladder s ys te ms ,SrC u$_2$ O $_3$[@sr cu 2o3 ], Ca V$_2$ O $_5$[@ cav2o 5],(V O$ _ 2$) P$_2$O$ _ 7$ [ @ vopo ], C u$_2 $(C $_ 5$H$_ {12} $ N$_ 2$)$_2$ Cl$_4$[@c uHp C l],an dthe pla quette RVB sy st em, CaV$_4 $O $_9 $[@cav 4 o 9].
Rec ently the two-dimension a l spingap syst em [ [SrCu$_2$ (BO $_3$)$ _2$ ] {}]{}has be en fo un d b y Kagey a m a[*e tal.*]{}[@k a g 98] . Itha s aspin-si nglet ground state wit h a spin gap$\s im 3 0 $ K . T h es ubs ta n ceh a s additional in terestingfe a tu res, e.g., the h igh-fie ld magn etiza t ion was observed to havetw o pl a t eau s at 1/4 a nd 1/8 o f the ful l mome n t. Rece ntwork [ @m iy9 8] su ggests tha t the under ly ing ph ysics c an be un derstood on the basis o f a tw o-dim ens ional $S= {1\ o ver 2}$ Heis enbe rg model w ith an tifer rom a gneti c ne a re st- n eighb or ( $ J_1$, onl in ks$ A $) and next-n e a r est -neig hbo r ($J_2 | DynamicFeedback2018; @hauswirthTimescaleSeparationAutonomous2019]_which stipulate_that the interconnection of_*fast decaying*_plant_dynamics and_*slow*_optimization dynamics is_asymptotically stable. The_results in this section_can be generalized_to_a dynamic plant accordingly.
[^11]: The penalty $d^2_{\mathcal{U}}$ is illustrative in the context of autonomous_optimization,_however, it_is_not_generally practical for numerical optimization,_because evaluating $\nabla d^2_{\mathcal{U}}$ requires_computing $P_{\mathcal{U}}$._Instead, in numerical applications, it is more common_to_use a penalty_$\| \max \{ A_u u - b_u, 0 \}\|^2$.
_Since the discovery of high-temperature superconductivity,_doped antiferromagnets (AF)_have_been_studied intensively. The pseudo-gap_behavior observed in the high-$T_c$ cuprates_has stimulated great interest in systems_with spin gaps. Several new one- and_two-dimensional spin gap systems have been_found experimentally. These materials are_characterized by_a disordered singlet ground state_and a finite_gap to_all spin excitations._Some of the compounds which have_two-dimensional (2d) character_include the coupled spin ladder systems,_SrCu$_2$O$_3$[@srcu2o3],_CaV$_2$O$_5$[@cav2o5], (VO$_2$)P$_2$O$_7$[@vopo], Cu$_2$(C$_5$H$_{12}$N$_2$)$_2$Cl$_4$[@cuHpCl],_and_the_plaquette RVB_system, CaV$_4$O$_9$[@cav4o9].
Recently the_two-dimensional_spin gap_system_[[SrCu$_2$(BO$_3$)$_2$]{}]{} has been found by Kageyama_[*et_al.*]{}[@kag98]. It has a spin-singlet ground state_with a spin gap_$\sim_30$ K. The substance_has additional interesting features, e.g.,_the high-field magnetization was observed to_have two_plateaus at_1/4 and 1/8 of the full moment. Recent work [@miy98] suggests_that the underlying physics can be_understood on the basis_of a_two-dimensional_$S={1\over 2}$ Heisenberg_model_with antiferromagnetic_nearest-neighbor ($J_1$, on links $A$) and next-nearest-neighbor_($J_2 |
\delta \Gamma = \frac{\langle || \delta \mathbf{u}^{\text{NA}} ||^2 \rangle}{\ell^2 \delta \gamma^2},$$ in which $\ell$ is the typical bond length of the network, and $\delta \mathbf{u}^{\text{NA}} = \mathbf{u} - \mathbf{u}^{\text{affine}}$ is the nonaffine displacement of a node that is caused by applying an infinitesimal shear strain $\delta \gamma$. To better illustrate this parameter, we show the nonaffine displacement vectors of nodes for a diluted triangular network before, at and after the critical strain in Fig. \[fig:A5\] in the Appendix [@sharma_strain-driven_2016]. The differential nonaffinity $\delta \Gamma$ diverges at the critical strain for central force networks, with a susceptibility-like exponent $\lambda = \phi - f$, i.e., $\delta \Gamma \sim |\Delta \gamma|^{-\lambda}$ [@broedersz_mechanics_2011; @sharma_strain-driven_2016; @shivers_scaling_2019]. Moreover, as the system approaches the critical strain, the correlation length diverges as $\xi \sim |\Delta \gamma|^{-\nu}$. When the correlation length is smaller than the system size $W$, i.e., $|\Delta \gamma| \times W^{1/\nu}>1$, we should find $\delta \Gamma \sim |\Delta \gamma|^{-\lambda}$. Near the critical strain, however, the finite-size effects result in $\delta \Gamma \sim |\Delta \gamma|^{\lambda/\nu}$. Therefore, the following scaling form must capture the behavior of fluctuations [@sharma_strain-driven_2016] $$\label{FluctuationsScaling}
\delta \Gamma = W^{\lambda/\nu} \mathcal{H}(\Delta \gamma W^{1/\nu}),$$ where the scaling function $\mathcal{H}(x)$ is constant for $|x|<1$ and $|x|^{-\lambda}$ otherwise. The differential nonaffinity is shown for different system sizes of a diluted triangular network in Fig. \[fig:A5\] in the Appendix. Based on the above scaling form, we perform a finite-size scaling analysis as shown in Fig. \[fig:4\]. The correlation length exponent $\nu$ is computed from the hyperscaling relation $f=d\nu - 2$ obtained for | \delta \Gamma = \frac{\langle || \delta \mathbf{u}^{\text{NA } } ||^2 \rangle}{\ell^2 \delta \gamma^2},$$ in which $ \ell$ is the typical bond length of the network, and $ \delta \mathbf{u}^{\text{NA } } = \mathbf{u } - \mathbf{u}^{\text{affine}}$ is the nonaffine translation of a lymph node that is caused by applying an infinitesimal shear sift $ \delta \gamma$. To well illustrate this parameter, we express the nonaffine shift vectors of nodes for a diluted triangular network before, at and after the critical form in Fig. \[fig: A5\ ] in the Appendix [ @sharma_strain - driven_2016 ]. The differential nonaffinity $ \delta \Gamma$ diverges at the critical strain for central power networks, with a susceptibility - like exponent $ \lambda = \phi - f$, i.e., $ \delta \Gamma \sim |\Delta \gamma|^{-\lambda}$ [ @broedersz_mechanics_2011; @sharma_strain - driven_2016; @shivers_scaling_2019 ]. Moreover, as the system approaches the critical stress, the correlation length diverges as $ \xi \sim |\Delta \gamma|^{-\nu}$. When the correlation distance is smaller than the system size $ W$, i.e., $ |\Delta \gamma| \times W^{1/\nu}>1 $, we should find $ \delta \Gamma \sim |\Delta \gamma|^{-\lambda}$. Near the critical strain, however, the finite - size consequence result in $ \delta \Gamma \sim |\Delta \gamma|^{\lambda/\nu}$. Therefore, the following scaling form must capture the demeanor of fluctuations [ @sharma_strain - driven_2016 ] $ $ \label{FluctuationsScaling }
\delta \Gamma = W^{\lambda/\nu } \mathcal{H}(\Delta \gamma W^{1/\nu}),$$ where the scaling function $ \mathcal{H}(x)$ is constant for $ |x|<1 $ and $ |x|^{-\lambda}$ otherwise. The differential nonaffinity is shown for different arrangement sizes of a diluted triangular network in Fig. \[fig: A5\ ] in the Appendix. Based on the above scaling form, we perform a finite - size scaling analysis as shown in Fig. \[fig:4\ ]. The correlation length exponent $ \nu$ is calculate from the hyperscaling sexual intercourse $ f = d\nu - 2 $ obtained for | \delha \Gamma = \frac{\langle || \deuta \mathbf{u}^{\text{UQ}} ||^2 \ranjle}{\ell^2 \selta \gaoma^2},$$ in which $\ell$ is the typiral vond oength of the network, xnd $\delta \mathbf{u}^{\rext{IA}} = \mathbf{u} - \matigf{u}^{\text{affine}}$ ls thz ionaffine displscement of a node that iv zabsed by applying an infinitesimal shqar strsij $\delta \gamma$. Jo beutew ilmlsurate this parameter, we show the honaffiie displacement vectors of nodes for a diputef triangular netwogk before, aj anq after the cfitical stgcin in Fig. \[fjg:A5\] in the Appendix [@sharma_straiv-drivzn_2016]. The diffgxwntldl nonaffinmty $\dejta \Gamma$ diyvrges ad the ctitical strain foc cebtral force networks, xith a susceptibiliti-like expoteut $\lambda = \phi - f$, i.e., $\eeota \Gdmma \sim |\Eelga \fakmz|^{-\lambdw}$ [@bcoedersz_mecganics_2011; @sharna_strain-driven_2016; @shivtrs_fbsling_2019]. Moreovsr, as ehq system approaches the critical strain, ths correlation length dicerges as $\xi \sim |\Deltw \gamma|^{-\nu}$. When the correlation length is smaller than the vystej siee $W$, i.e., $|\Eepta \gamma| \times W^{1/\nu}>1$, we should find $\delta \Gammw \som |\Delta \gamma|^{-\laibda}$. Near tne ctytical strain, howevzd, fhe finite-size efffcts refult un $\delta \Damms \sim |\Delta \gamma|^{\lambda/\nu}$. Tyerefore, the dollowing scaling yorm must ca'ture jhe benavior of fluctuations [@rharja_strain-drigen_2016] $$\label{RuuctuationsScalivg}
\dvlta \Gamma = W^{\lambda/\nu} \mathcal{H}(\Qelta \gamna W^{1/\uu}),$$ where the scaligg functioj $\matmwal{H}(x)$ is constant vor $|x|<1$ atd $|x|^{-\lambda}$ otherwise. The differential nonehfinity is shpwt fmr diffexent snstem sizes of w diluted triaugular nztwork in Fig. \[fig:Z5\] in thx Appendix. Bwsed on the atlve scaling horm, we pqrfoem a finite-rkze scaling anslysis as shown in Dig. \[fig:4\]. The correlabion usngth exponent $\ub$ us computed frok tfe ryierxcwning relatiot $f=d\vu - 2$ obtakned for | \delta \Gamma = \frac{\langle || \delta \mathbf{u}^{\text{NA}} \delta in which is the typical and \mathbf{u}^{\text{NA}} = \mathbf{u} \mathbf{u}^{\text{affine}}$ is the displacement of a node that is by applying an infinitesimal shear strain $\delta \gamma$. To better illustrate this parameter, show the nonaffine displacement vectors of nodes for a diluted triangular network before, and the strain Fig. \[fig:A5\] in the Appendix [@sharma_strain-driven_2016]. The differential nonaffinity $\delta \Gamma$ diverges at the critical strain central force networks, with a susceptibility-like exponent $\lambda \phi - f$, i.e., \Gamma \sim |\Delta \gamma|^{-\lambda}$ [@broedersz_mechanics_2011; @shivers_scaling_2019]. as the approaches critical the correlation length as $\xi \sim |\Delta \gamma|^{-\nu}$. When the correlation length is smaller than the system size $W$, i.e., \gamma| \times should find \Gamma |\Delta Near the critical the finite-size effects result in $\delta \gamma|^{\lambda/\nu}$. Therefore, the following scaling form must capture behavior of [@sharma_strain-driven_2016] $$\label{FluctuationsScaling} \delta \Gamma = W^{\lambda/\nu} \gamma W^{1/\nu}),$$ where the scaling function $\mathcal{H}(x)$ is for $|x|<1$ and $|x|^{-\lambda}$ otherwise. The differential nonaffinity is shown for different system sizes of triangular network in Fig. in the Appendix. on above form, perform a scaling analysis as shown in Fig. \[fig:4\]. The correlation length exponent is computed from the hyperscaling relation $f=d\nu - 2$ obtained | \delta \Gamma = \frac{\langle || \delta \Mathbf{u}^{\texT{NA}} ||^2 \raNglE}{\elL^2 \dElta \GammA^2},$$ in which $\ell$ is tHE typIcal bond length of the netWork, aNd $\DElta \MAtHbf{u}^{\tExt{NA}} = \maTHbF{U} - \MatHbF{u}^{\TexT{aFFiNe}}$ is tHe nOnaffinE displacemEnt Of A node that is cAUsEd by applyiNg aN infinitesimAl sHear stRaIn $\dELta \gaMma$. to betTer illUStrate This paramEtER, we shoW The nonaFFInE disPlacement vectors oF NoDEs for a diluted tRianguLaR NeTWOrk BefOre, at and afTeR the cRItical sTRaIN IN FiG. \[Fig:A5\] in the AppeNdix [@sharma_sTRaiN-driveN_2016]. THe dIFferenTial nOnAFfiNity $\delta \GaMma$ dIverges at The criTIcal strAIn for ceNtral fOrcE neTworKS, wItH a sUsCEptIBiLitY-LikE exponenT $\lAmBda = \phI - f$, i.e., $\DELTA \GamMa \sIm |\DeLta \gaMma|^{-\lambda}$ [@broeDerSz_meCHanIcs_2011; @shArma_sTraiN-dRiven_2016; @ShiverS_scalInG_2019]. Moreover, as the sYsteM approachEs tHe CriTiCal stRAin, the CorRelAtion leNgth divERgeS aS $\XI \SiM |\Delta \gamma|^{-\nu}$. When tHe CORrElation lEngth iS SmAlLEr than thE sYstEm siZE $w$, i.e., $|\DeLta \gAMmA| \times W^{1/\nU}>1$, we shoULd FiNd $\delta \gaMma \sim |\deLta \GamMa|^{-\lamBDa}$. NeAr the cRitical sTrain, HOwever, the finitE-Size effects reSUlT IN $\dELta \GAmmA \sim |\Delta \gaMma|^{\lAMbda/\Nu}$. ThEReForE, The foLlowiNg SCaLIng form must capture tHe BehaviOr of fLuctuations [@shArma_strain-DRIVen_2016] $$\label{flucTUaTIonsScaling}
\delTa \GamMa = W^{\lambda/\nU} \Mathcal{H}(\delta \Gamma W^{1/\nu}),$$ Where the sCALing funcTioN $\maThcAl{H}(X)$ IS cOnstant for $|x|<1$ anD $|X|^{-\LambDa}$ OtherwiSe. THe diffeRenTiaL noNafFiNity is shoWn for difFeReNt SySteM sizeS Of a dilutEd TriAnGulAr netWOrk in FIg. \[fig:a5\] in tHe apPEndIx. Based ON tHE AbovE sCaLing ForM, wE perfOrm a FIniTe-size sCaling anaLysIS as sHoWn In Fig. \[fiG:4\]. The correlatiOn Length expoNeNt $\nU$ is comPUTed from tHe hyperscaling relation $f=D\Nu - 2$ obtaiNed For | \delta \Gamma = \frac{\lan gle || \de lta \ mat hbf {u }^{\ text {NA}} ||^2 \ra n gle} {\ell^2 \delta \gamma^ 2},$$ i n whi c h$\ell $ is th e t y p ica lbo ndle n gt h ofthe networ k, and $\d elt a\mathbf{u}^{ \ te xt{NA}} =\ma thbf{u} - \m ath bf{u}^ {\ tex t {affi ne} }$ is the n o naffin e displac em e nt ofa node t h a tis c aused by applying an infinitesimalshearst r ai n $\d elt a \gamma$. T o bet t er illu s tr a t e th i s parameter,we show the non affine d isp l acemen t vec to r s o f nodes for a d iluted tr iangul a r netwo r k befor e, atand af tert he c rit ic a l s t ra ini n F ig. \[fi g: A5 \] in the A p p endi x [ @sha rma_s train-driven_ 201 6].T hediffe renti al n on affin ity $\ delta \ Gamma$ diverges atthe criti cal s tra in forc entral fo rce networ ks, wit h asu s c e pt ibility-like expon en t $\ lambda = \phi- f $, i.e., $\ de lta \Ga m m a \si m |\ D el ta \gamm a|^{-\ l am bd a}$ [@b ro edersz _m ech ani cs_20 1 1; @ sharma _strain- drive n _2016; @shiver s _scaling_2019 ] .M o re o ver, as the system app r oach es t h ecri t icalstrai n, th e correlation length d iverge s as$\xi \sim |\D elta \gamm a | ^ {-\nu}$. Whe n t h e correlationlengt h is small e r than t he sy stem siz e $W$, i. e . , $|\Del ta\ga mma | \ t i me s W^{1/\nu}>1 $ , wesh ould fi nd$\delta \G amm a \ sim | \Delta \g amma|^{- \l am bd a} $.Neart he criti ca l s tr ain , how e ver, t he fi nite -s iz e ef fects r e su l t in$\ de lta\Ga mm a \si m |\ D elt a \gamm a|^{\lamb da/ \ nu}$ .Th erefore , the followi ng scaling f or m m ust ca p t ure thebehavior of fluctuation s [@shar ma_ strai n-dr iven_2016 ] $ $\labe l{F l uctuat ionsSc aling }\de l t a \Ga m m a= W ^{ \lambda/\n u } \m athca l{ H}(\ Delta \ gamma W^{1/\nu}),$ $ wh ere the scali ngfunc t i on $\ m at h cal {H } (x) $ is constant for $|x|<1$ a nd $| x|^{-\lamb d a}$ o therwis e. Thediffe r entialnonaffini ty is sho wn for d iff erent syst em sizes of a dil u ted t r ia ngula r n etwork i n F ig. \ [fig:A 5 \]in th e Appe nd ix. Ba sed o nthe abov e scaling form, we perf orm afinit e-s ize scali nga nal ysis as s hown in Fig. \ [fi g:4 \]. T hec orrel atio n l eng t h exp onen t $\nu$ is co mpu t e dfrom the hy p e r sca lingrel a tion $ f=d\ nu - 2$ obtainedf or | \delta \Gamma_= \frac{\langle_|| \delta \mathbf{u}^{\text{NA}} ||^2_\rangle}{\ell^2 \delta_\gamma^2},$$_in which_$\ell$_is the typical_bond length of_the network, and $\delta_\mathbf{u}^{\text{NA}} = \mathbf{u}_-_\mathbf{u}^{\text{affine}}$ is the nonaffine displacement of a node that is caused by applying an_infinitesimal_shear strain_$\delta_\gamma$._To better illustrate this parameter,_we show the nonaffine displacement_vectors of_nodes for a diluted triangular network before, at_and_after the critical_strain in Fig. \[fig:A5\] in the Appendix [@sharma_strain-driven_2016]. The differential_nonaffinity $\delta \Gamma$ diverges at the_critical strain for_central_force_networks, with a susceptibility-like_exponent $\lambda = \phi - f$,_i.e., $\delta \Gamma \sim |\Delta \gamma|^{-\lambda}$_[@broedersz_mechanics_2011; @sharma_strain-driven_2016; @shivers_scaling_2019]. Moreover, as the system_approaches the critical strain, the correlation_length diverges as $\xi \sim_|\Delta \gamma|^{-\nu}$._When the correlation length is_smaller than the_system size_$W$, i.e., $|\Delta_\gamma| \times W^{1/\nu}>1$, we should find_$\delta \Gamma \sim_|\Delta \gamma|^{-\lambda}$. Near the critical strain,_however,_the finite-size effects_result_in_$\delta \Gamma_\sim |\Delta \gamma|^{\lambda/\nu}$._Therefore,_the following_scaling_form must capture the behavior of_fluctuations_[@sharma_strain-driven_2016] $$\label{FluctuationsScaling}
\delta \Gamma = W^{\lambda/\nu} \mathcal{H}(\Delta \gamma_W^{1/\nu}),$$ where the scaling_function_$\mathcal{H}(x)$ is constant for_$|x|<1$ and $|x|^{-\lambda}$ otherwise. The_differential nonaffinity is shown for different_system sizes_of a_diluted triangular network in Fig. \[fig:A5\] in the Appendix. Based on the_above scaling form, we perform a_finite-size scaling analysis as_shown in_Fig. \[fig:4\]._The correlation length_exponent_$\nu$ is_computed from the hyperscaling relation $f=d\nu -_2$ obtained_for |