Tame proper base change for discretely ringed adic spaces
Katharina Hübner · 2025
Read Tame proper base change for discretely ringed adic spaces by Katharina Hübner (2025) 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-2503-13312.txt.
Introduction
Katharina Hübner proves a base change result for a tame topology on adic spaces in this short mathematics paper from twenty twenty-five. The setting involves a proper map between such spaces and a smaller subspace pulled back along it. For a certain kind of sheaf, one built from p-torsion data, two routes of moving the data agree exactly. The derived pushforward and the pullback commute, stated as an isomorphism. Results of this type let geometers compute global data from local pieces with confidence.
Catalog record
| CREATOR | Katharina Hübner |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 111 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-13312
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Tame proper base change for discretely ringed adic spaces Thanks: This research is partly supported by ERC Consolidator Grant 770922 - BirNonArchGeom. Moreover the author acknowledges support by Deutsche Forschungsgemeinschaft (DFG) through the Collaborative Research Centre TRR 326 "Geometry and Arithmetic of Uniformized Structures", project number 444845124. Katharina Hübner Abstract . We consider a proper morphism X → S X\to S and a locally closed immersion S ′ → S S^{\prime}\to S of discretely ringed adic spaces and prove proper base change for the tame topology in this setting. More precisely, we show for an abelian p p -torsion sheaf ( p = char + ( S ) p=\ch^{+}(S) ) on the tame site of X X that the base change homomorphism for the derived pushforward along X → S X\to S with the pullback along S ′ → S S^{\prime}\to S is an isomorphism. 1. Introduction We consider a cartesian diagram {\lx@inpgf@ignorespaces{\mathscr{X}}^{\prime}} {\lx@inpgf@ignorespaces{\mathscr{X}}} {\lx@inpgf@ignorespaces{\mathscr{S}}^{\prime}} {\lx@inpgf@ignorespaces{\mathscr{S}}} \scriptstyle{\lx@inpgf@ignorespaces\iota_{\mathscr{X}}} \scriptstyle{\lx@inpgf@ignorespaces f^{\prime}} \scriptstyle{\lx@inpgf@ignorespaces f} \scriptstyle{\lx@inpgf@ignorespaces\iota_{\mathscr{S}}} of adic spaces. By adjunction we always get a base change homomorphism \iota_{\mathscr{S}}^{*}Rf_{*}{\mathscr{F}}\longrightarrow Rf^{\prime}_{*}\iota_{\mathscr{X}}^{*}{\mathscr{F}} for any sheaf of torsion abelian groups {\mathscr{F}} on the étale site of {\mathscr{X}} . If f is proper, we expect this base change homomorphism to be an isomorphism. For analytic adic spaces with the usual finiteness assumptions this was shown in Huber’s book [ Hub96 ] , § 4.4, at least for torsion sheaves whose torsion is invertible on {\mathscr{X}} . Especially for p -torsion coefficients the theorem was largely extended, see [ Man22 ] , Theorem 1.2.4. We are intersested in a similar base change theorem for the tame site. More precisely, we are going to prove the following result. Theorem 1.1 . bc_p_torsion Let f\colon{\mathscr{X}}\to{\mathscr{S}} be a proper morphism of discretely ringed pseudoadic spaces of specialization characteristic \ch^{+}({\mathscr{S}})=p>0 . We consider the cartesian square {\lx@inpgf@ignorespaces{\mathscr{X}}^{\prime}} {\lx@inpgf@ignorespaces{\mathscr{X}}} {\lx@inpgf@ignorespaces{\mathscr{S}}^{\prime}} {\lx@inpgf@ignorespaces{\mathscr{S}}} \scriptstyle{\lx@inpgf@ignorespaces\iota_{\mathscr{X}}} \scriptstyle{\lx@inpgf@ignorespaces f^{\prime}} \scriptstyle{\lx@inpgf@ignorespaces f} \scriptstyle{\lx@inpgf@ignorespaces\iota_{\mathscr{S}}} for a locally closed immersion \iota_{\mathscr{S}} . Then for every p -torsion sheaf {\mathscr{F}} on the tame site {\mathscr{X}}_{t} , the base change homomorphism \iota_{\mathscr{S}}^{*}Rf_{*}{\mathscr{F}}\longrightarrow Rf^{\prime}_{*}\iota_{\mathscr{X}}^{*}{\mathscr{F}} is an isomorphism. Let us first explain why we cannot expect the ususal statement in full generality, whithout restricting to locally closed immersions. We consider the following example. Let (K,K^{+}) be a tamely closed valued field and (L/L^{+})/(K,K^{+}) a finite wildly ramified Galois extension with Galois group G . We consider the cartesian square {\lx@inpgf@ignorespaces\prod_{G}(L,L^{+})=\Spa(L\otimes_{K}L,L^{+}\otimes_{K^{+}}L^{+})} {\lx@inpgf@ignorespaces\Spa(L,L^{+})} {\lx@inpgf@ignorespaces\Spa(L,L^{+})} {\lx@inpgf@ignorespaces\Spa(K,K^{+}).} If the base change theorem held for this cartesian square, there would have to be an isomorphism {\mathscr{F}}(\Spa(L,L^{+}))\overset{\sim}{\longrightarrow}\bigoplus_{G}{\mathscr{F}}(\Spa(L,L^{+})) for any p -torsion sheaf on {\mathscr{X}}_{t} . This is clearly false. We could say that the reason for this failure is the cancellation of wild ramification. In the statement of the above theorem this cannot happen since {\mathscr{S}}^{\prime}\to{\mathscr{S}} is an immersion, which has trivial residue field extensions. It is conceivable to prove a slightly more general statement by imposing some tameness condition on {\mathscr{S}}^{\prime}\to{\mathscr{S}} (which a locally closed immersion should certainly satisfy). However, it seems that such a condition would be quite restrictive and already exclude examples like {\mathbb{A}}^{1}_{(K,K^{+})}\to\Spa(K,K^{+}) for some non-algebraically closed valued field (K,K^{+}) . Let us now explain the strategy for the proof of Theorem 1.1 . In fact we prove a base change theorem for the strongly étale topology on {\mathscr{X}} and in the end deduce the result for the tame site using that strongly étale and tame cohomology coincides for p -torsion sheaves. In order to prove the strongly étale base change theorem we study proper adic spaces over a base space of the form {\mathscr{S}}=\Spa(R,R^{+}) for a Huber pair (R,R^{+}) equipped with the discrete topology. We prove in Proposition 6.5 that proper adic spaces over {\mathscr{S}} are always of the form {\mathscr{X}}=\Spa(X,R^{+}) for a scheme X that is proper over R . The strongly étale cohomology of {\mathscr{X}} can be computed as a colimit over all X -modifications of \Spec R^{+} : For all sheaves {\mathscr{F}} on {\mathscr{X}}_{\textit{s\'{e}t}} , we have H^{i}({\mathscr{X}}_{\textit{s\'{e}t}},{\mathscr{F}})=\colim_{Y}H^{i}(Y_{\textit{\'{e}t}},\varphi_{Y,*}{\mathscr{F}}), where the colimit runs over all X -modifications of \Spec R^{+} and \varphi_{Y}\colon{\mathscr{X}}\to Y is the center map (see Corollary 5.7 ). The identification of the cohomology of {\mathscr{X}} as a colimit over all modifications is not so straightforward. This has to do with the fact that as a topological space, {\mathscr{X}} does not coincide with the limit over all X -modifications Y of \Spec R^{+} . Instead, the limit over all modifications is naturally a subspace of {\mathscr{X}} , the space of all Riemann Zariski points {\mathscr{X}}_{\mathrm{RZ}} (see [ Tem11 ] ). However, the subspace {\mathscr{X}}_{\mathrm{RZ}} can be thought of as a deformation retract of {\mathscr{X}} . At the level of strongly étale topoi this is reflected by the following diagram (see Section 4 ) {\lx@inpgf@ignorespaces{\mathscr{X}}_{\mathrm{RZ},\textit{s\'{e}t}}} {\lx@inpgf@ignorespaces{\mathscr{X}}_{\textit{s\'{e}t}}} {\lx@inpgf@ignorespaces{\mathscr{X}}_{\mathrm{RZ},\textit{s\'{e}t}}.} \scriptstyle{\lx@inpgf@ignorespaces\iota} \scriptstyle{\lx@inpgf@ignorespaces\mathrm{id}} \scriptstyle{\lx@inpgf@ignorespaces\pi} The projection {\mathscr{X}}_{\textit{s\'{e}t}}\to{\mathscr{X}}_{\mathrm{RZ},\textit{s\'{e}t}} is easy to define whereas the embedding \iota\colon{\mathscr{X}}_{\mathrm{RZ},\textit{s\'{e}t}}\to{\mathscr{X}}_{\textit{s\'{e}t}} requires some work. The main result of Section 4 is an isomorphism H^{i}({\mathscr{X}}_{\textit{s\'{e}t}},{\mathscr{F}})\cong H^{i}({\mathscr{X}}_{\mathrm{RZ},\textit{s\'{e}t}},\iota^{*}{\mathscr{F}}) and the right hand side naturally identifies with the colimit over all X -modifications Y of \Spec R^{+} of H^{i}(Y_{\textit{\'{e}t}},\varphi_{Y,*}\iota^{*}{\mathscr{F}}) (see Section 5 ). The idea now is to apply the proper base change theorem for the étale site of a scheme to these cohomology groups for all modifications Y . In the first four sections we work with arbitrary adic spaces. Only starting in Section 5 we restrict our attention to discretely ringed adic spaces. In Section 5 this is not yet crucial but makes the exposition easier. The structure of proper adic spaces described in Section 6 is much more complicated for arbitrary adic spaces as compared to discretely ringed ones. In the future we aim for a more general tame base change result for all adic spaces. This can build upon the first sections of the present article but needs more input in the study of proper morphisms. 2. Setup and notation In this article we will work with the tame and strongly étale site of pseudoadic spaces, mainly discretely ringed ones. To put everyone on the same page, we start by reviewing these concepts (assuming familiarity with adic spaces). Pseudoadic spaces have been introduced by Huber in order to study closed subsets of adic spaces, which do not necessarily carry the structure of an adic space. Following [ Hub96 ] , § 1.10, a pseudoadic space is a pair X=(\underline{X},\lvert X\rvert) , where \underline{X} is an adic space and \lvert X\rvert\subseteq\underline{X} is a convex proconstructible subset. A morphism of pseudoadic spaces from Y to X is a morphism of adic spaces f:\underline{Y}\to\underline{X} such that f(\lvert Y\rvert)\subseteq\lvert X\rvert . The morphism f is étale if it is étale as a morphism of adic spaces and \lvert Y\rvert is open in f^{-1}(\lvert X\rvert) . It is strongly étale if it is étale and for all points y\in\lvert Y\rvert mapping to x\in\lvert X\rvert the residue field extension k(y)/k(x) is unramified. Similarly, f is tame if it is étale and the above residue field extensions are tamely ramified. Defining coverings to be surjective families, we get the étale, strongly étale, and tame sites X_{\textit{\'{e}t}} , X_{\textit{s\'{e}t}} , and X_{t} of a pseudoadic space X . A point x of a (pseudo)adic space X has a residue field k(x) , which is a valued field with valuation ring k(x)^{+}