On the locally analytic completed cohomology of modular curves

Lue Pan · 2026

Read On the locally analytic completed cohomology of modular curves by Lue Pan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-27937.txt.

Introduction

Pan surveys locally analytic vectors in modular curve cohomology. Completed cohomology of modular curves holds these vectors. His works map their structure across several papers. The survey gathers those results in one place. It serves readers tracking the p-adic Langlands area. Key constructions and outcomes are summarized. Links between papers are made explicit. The piece is a guide to his research line.

Catalog record

CREATORLue Pan
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38814
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.27937

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

On the locally analytic completed cohomology of modular curves Lue Pan Abstract . We survey our works on the locally analytic vectors of completed cohomology of modular curves. 1. Introduction Let p be a prime number. Emerton introduced the ( p -adically) completed cohomology in [ 17 , 19 ] which he described as “a suitable surrogate for a space of p -adic automorphic forms”. In his highly influential unpublished manuscript [ 18 ] , Emerton carefully studied the completed cohomology of modular curves and successfully applied it to attack the Fontaine-Mazur conjecture. We briefly recall its definition. Let K be an open subgroup \mathrm{GL}_{2}(\mathbb{A}_{f}) and consider the modular curve of level K Y_{K}(\mathbb{C})=\mathrm{GL}_{2}(\mathbb{Q})\backslash\left((\mathbb{C}\setminus\mathbb{R})\times\mathrm{GL}_{2}(\mathbb{A}_{f})/K\right). Throughout this paper fix an open compact subgroup K^{p} of \mathrm{GL}_{2}(\mathbb{A}^{p}_{f}) . The completed cohomology of tame level K^{p} is defined as \tilde{H}^{i}(K^{p}):=\varprojlim_{n}\varinjlim_{K_{p}\subset\mathrm{GL}_{2}(\mathbb{Q}_{p})}H^{i}(Y_{K^{p}K_{p}}(\mathbb{C}),\mathbb{Z}/p^{n}). It is p -adically complete and equipped with natural actions of \mathrm{GL}_{2}(\mathbb{Q}_{p}) and the Hecke algebra \mathbb{T} (generated by spherical Hecke operators). Moreover because modular curves are algebraic varieties canonically defined over \mathbb{Q} , \tilde{H}^{i}(K^{p}) also admits a natural action of G_{\mathbb{Q}}=\Gal(\bar{\mathbb{Q}}/\mathbb{Q}) . The key question here is to understand how the action of \mathrm{GL}_{2}(\mathbb{Q}_{p}) interacts with other two symmetries. Emerton’s approach is to show the interplay between the actions of \mathrm{GL}_{2}(\mathbb{Q}_{p}) and G_{\mathbb{Q}_{p}}=\Gal(\bar{\mathbb{Q}}_{p}/\mathbb{Q}_{p}) is compatible with the p -adic local Langlands correspondence for \mathrm{GL}_{2}(\mathbb{Q}_{p}) established by Breuil, Colmez, Berger, Kisin, Paškūnas… In our recent works [ 27 , 29 ] we found a new approach to understand the completed cohomology. Our method is based on the study of the p -adic geometry of the modular curve with infinite level at p introduced by Scholze in [ 33 ] . Roughly speaking, we adapt Sen’s method (reformulated by Berger-Colmez in [ 5 , 6 ] ) to study the locally analytic functions at the infinite level and show that the locally analytic vectors of the completed cohomology, as a \mathrm{GL}_{2}(\mathbb{Q}_{p}) -representation, can be “localized” (in the sense of Beilinson-Bernstein [ 3 ] ) naturally to certain D -modules on \mathbb{P}^{1} . A majority of our works is about understanding the structures of these D -modules. Our method has applications to the Fontaine-Mazur conjecture, study of overconvergent modular forms and the p -adic local Langlands correspondence. We will review our works below and discuss some recent development in the field. 2. Sen theory and localization of locally analytic completed cohomology 2.1. Relative/Geometric Sen theory In [ 27 , §3] , we explain how to adapt Sen’s construction of the Sen operator for p -adic Galois representations to define a canonical Higgs field over the pro-étale site of a smooth affinoid one-dimensional rigid analytic variety and call it “Relative Sen theory”. Later Rodriguez Camargo [ 31 ] systematically generalize the method to arbitrary dimension and call it “Geometric Sen theory”. We summarize some of the results here. Denote by C the completion of \bar{\mathbb{Q}}_{p} . Theorem 2.1 . Let X be a smooth rigid analytic variety over C of dimension d and V a pro-étale \mathbb{Z}_{p} -local system on X . One can canonically define a map on the pro-étale site X_{\mathrm{pro\acute{e}t}} \theta_{V}:V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{X}\to V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{X}\otimes_{\mathcal{O}_{X}}\Omega^{1}_{X}(-1), where \hat{\mathcal{O}}_{X} denotes the complete structural sheaf on X_{\mathrm{pro\acute{e}t}} and (-1) denotes the inverse of the Tate twist, satisfying the following (1) \theta_{V} is a Higgs field, i.e. \theta_{V} is \hat{\mathcal{O}}_{X} -linear and \theta_{V}\wedge\theta_{V}=0 . (2) \theta_{V} is functorial in V . Moreover if W is another \mathbb{Z}_{p} -local system, then \theta_{V\oplus W}=(\theta_{V},\theta_{W}) and \theta_{V\otimes W}=\theta_{V}\otimes 1+1\otimes\theta_{W} . (3) Let \nu:X_{\mathrm{pro\acute{e}t}}\to X_{\mathrm{\acute{e}t}} denote the projection of sites. Then \nu_{*}(V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{X})=\nu_{*}\ker\theta_{V} . (4) \theta_{V} can be explicitly described in local toric charts. Remark 2.2 . When X has a log structure defined by a normal crossing divisor, there is a logarithmic version of the theorem over the pro-Kummer-étale site. Remark 2.3 . V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{X} is a pro-étale vector bundle. [ 31 ] actually constructs Higgs fields for general pro-étale vector bundles and certain infinite-dimensional vector bundles [ 31 , Def. 1.0.1] , and shows that the Higgs field vanishes if and only the pro-étale vector bundle descends to the étale site. Remark 2.4 . For part (3), one can compute higher R^{i}\nu_{*}(V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{X}) as well using the Higgs complex [ 31 , Th.1.0.3 ] . Remark 2.5 . If G is a compact p -adic Lie group and \tilde{X} is a pro-étale G -torsor of X , one can take V to be the local system associated to the space of locally analytic functions on G (which is infinite-dimensional but the construction of \theta_{V} works here). Now H^{0}(\nu_{*}(V\widehat{\otimes}_{\mathbb{Q}_{p}}\hat{\mathcal{O}}_{X}))=\hat{\mathcal{O}}_{X}(\tilde{X})^{\mathrm{la}} computes the G -locally analytic functions of \hat{\mathcal{O}}_{X}(\tilde{X}) . Part (3) says that they satisfy the first-order differential equation defined by \theta_{V} . We call it the p -adic Cauchy-Riemann equation. This point of view is inspired by the interpretation of the Sen operator in [ 6 ] . Remark 2.6 . The explicit description in part (4) is as follows: suppose U=\Spa(A,A^{+})\subseteq X is an affinoid and admits a morphism to Y:=\Spa(C\langle T_{1}^{\pm 1},\cdots,T_{d}^{\pm 1}\rangle,O_{C}\langle T_{1}^{\pm 1},\cdots,T_{d}^{\pm 1}\rangle) which can be written as a composite of finite étale maps and rational localizations. Let U_{\infty} denote the pullback of the \mathbb{Z}_{p}(1)^{d} -cover \Spa(C\langle T_{1}^{\pm 1/p^{\infty}},\cdots,T_{d}^{\pm 1/p^{\infty}}\rangle,O_{C}\langle T_{1}^{\pm 1/p^{\infty}},\cdots,T_{d}^{\pm 1/p^{\infty}}\rangle) of Y . If \theta_{V} is viewed as a \mathcal{O}_{X} -linear map \Omega^{1}_{X}(-1)^{\vee}\to\End(V\otimes\hat{\mathcal{O}}_{X}) , the action of the dual basis of \frac{dT_{1}}{T_{1}},\cdots,\frac{dT_{d}}{T_{d}} on the \mathbb{Z}_{p}(1)^{d} -locally analytic vectors of V\otimes\hat{\mathcal{O}}_{X}(U_{\infty}) agrees with the action of the standard basis of \Lie\mathbb{Z}_{p}(1)^{d} . When both X and V are defined over a finite extension of \mathbb{Q}_{p} and V is a de Rham local system, the Higgs field is closely related to the flat connection associated to V . We make this more precise in the case of modular curves. Fix a level K and let \mathcal{X}=\mathcal{X}_{K} denote the adic space associated to the compactification of Y_{K}\times_{\mathbb{Q}}C . Let f:E\to Y_{K} be the universal elliptic curve. The (canonical extension of) relative de Rham cohomology H^{1}_{\mathrm{dR}}(E/Y_{k}) defines a filtered vector bundle (D,\Fil^{\bullet}) on \mathcal{X} with a logarithm connection \nabla:D\to D\otimes_{\mathcal{O}_{\mathcal{X}}}\Omega^{1}_{\mathcal{X}}(\mathcal{C}) where \mathcal{C} denotes the cusps. On the other hand, R^{1}f_{\mathrm{\acute{e}t},*}\mathbb{Z}_{p} defines a pro-Kummer-étale local system V of rank 2 on \mathcal{X}_{\mathrm{prok\acute{e}t}} . We have the relative Hodge-Tate filtration on V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{\mathcal{X}} 0\to\gr^{0}D\otimes_{\mathcal{O}_{\mathcal{X}}}\hat{\mathcal{O}}_{\mathcal{X}}\to V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{\mathcal{X}}\to\gr^{1}D\otimes_{\mathcal{O}_{\mathcal{X}}}\hat{\mathcal{O}}_{\mathcal{X}}(-1)\to 0. The following result is essentially due to Faltings [ 20 , Th.5] [ 27 , Th.4.2.2, 4.2.4] . Theorem 2.7 . \theta_{V}|_{\gr^{0}D\otimes_{\mathcal{O}_{\mathcal{X}}}\hat{\mathcal{O}}_{\mathcal{X}}}=0 and \theta_{V}(V\otimes_{\mathbb{Z}_{p}}\hat{\mathcal{O}}_{\mathcal{X}})\subseteq\gr^{0}D\otimes_{\mathcal{O}_{\mathcal{X}}}\hat{\mathcal{O}}_{\mathcal{X}}\otimes_{\mathcal{O}_{\mathcal{X}}}\Omega^{1}_{\mathcal{X}}(\mathcal{C})(-1) . Moreover the induced map \gr^{1}D\otimes_{\mathcal{O}_{\mathcal{X}}}\hat{\mathcal{O}}_{\mathcal{X}}(-1)\to\gr^{0}D\otimes_{\mathcal{O}_{\mathcal{X}}}\hat{\mathcal{O}}_{\mathcal{X}}\otimes_{\mathcal{O}_{\mathcal{X}}}\Omega^{1}_{\mathcal{X}}(\mathcal{C})(-1) agrees with the tensor product of the Kodaira-Spencer map and \hat{\mathcal{O}}_{\mathcal{X}}(-1) up to a sign, hence it’s an isomorphism for modular curves. Remark 2.8 . Recall that the Kodaira-Spencer map is defined as the composite of \gr^{1}D=\Fil^{1}D\xrightarrow{\nabla}\Fil^{0}D\otimes\Omega^{1}_{\mathcal{X}}(\mathcal{C})\to\gr^{0}D\otimes\Omega^{1}_{\mathcal{X}}(\mathcal{C}). Remark 2.9 . The calculation of \theta_{V} has been generalized to arbitrary Shimura varieties by Rodriguez Camargo [ 32 , §5] using the main result of [ 14 ] . 2.2. Localization: the sheaf \mathcal{O}^{\mathrm{la}} In [ 33 ] , Scholze showed that \mathcal{X}_{K^{p}}\sim\varprojlim_{K_{p}}\mathcal{X}_{K^{p}K_{p}} exists as a perfectoid space (the modular curve with infinite level at p ). The local system V is canonically trivialized on \mathcal{X}_{K^{p}} and the relative Hodge-Tate filtration defines a \mathrm{GL}_{2}(\mathbb{Q}_{p}) -equivariant map called the Hodge-Tate period morphism: \pi_{\mathrm{HT}}:\mathcal{X}_{K^{p}}\to{\mathscr{F}\!\ell}=\mathbb{P}^{1}, where {\mathscr{F}\!\ell} denotes the adic space associated to the flag variety of \mathrm{GL}_{2}/C . Definition 2.10 . Define \mathcal{O}_{K^{p}}:=\pi_{\mathrm{HT},*}\mathcal{O}_{\mathcal{X}_{K^{p}}}

Read the complete text