Descent for algebraic stacks

Olivier de Gaay Fortman · 2025

Read Descent for algebraic stacks by Olivier de Gaay Fortman (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-2504-13642.txt.

Introduction

Olivier de Gaay Fortman wrote this math paper in two thousand twenty-five as the sole author. It proves that algebraic stacks obey two-descent for fppf coverings.

Descent asks when an object built from local pieces, with glue data that agree, counts as a true global object. For schemes the Galois case is classic. Here the author widens Galois descent to stacks. The case in view is a finite Galois covering done in the fppf sense. Descent data then take the shape of a Galois group action on the stack. The result firms up a basic tool of modern algebraic geometry.

Catalog record

CREATOROlivier de Gaay Fortman
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE106 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-13642

Browse the full library · About the Tower of Babel

Full text

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

Descent for algebraic stacks Olivier de Gaay Fortman Abstract . We prove that algebraic stacks satisfy 2 2 -descent for fppf coverings. We generalize Galois descent for schemes to stacks, by considering the case where the fppf covering is a finite Galois covering, and reformulating 2 2 -descent data in terms of Galois group actions on the stack. August 24, 2026 1. Introduction † † August 24, 2026 Let S^{\prime}\to S be a morphism of affine schemes, faithfully flat and locally of finite presentation. By a theorem of Grothendieck, the functor X\mapsto X\times_{S}S^{\prime} induces an equivalence of categories between the category of S -schemes X and the category of pairs (X^{\prime},\phi) where X^{\prime} is an S^{\prime} -scheme and \phi a descent datum for X^{\prime} over S^{\prime} such that X^{\prime} admits an open covering by S^{\prime} -affine schemes which are stable under \phi (cf. [ 6 ] ). In the case when S=\textnormal{Spec}(k) , S^{\prime}=\textnormal{Spec}(k^{\prime}) , and the morphism S^{\prime}\to S corresponds to a finite Galois extension of fields k\subset k^{\prime} , this is known as Galois descent, and was proven by Weil (cf. [ 13 ] ). The aim of this paper is to present the most natural analogue of this result in the setting of algebraic stacks. To do this, we build on existing work in descent theory for stacks, especially results by Giraud [ 5 ] , Duskin [ 4 ] , and Breen [ 3 ] . Our main contributions are as follows: (1) We provide a reference for descent theory for stacks, explaining how the aforementioned results imply that stacks with descent data over a scheme S^{\prime} descend along fppf morphisms p\colon S^{\prime}\to S . (2) We show that algebraic stacks descend not just as stacks, but as algebraic stacks; we prove the same for Deligne–Mumford stacks. (3) We generalize Galois descent from schemes to stacks by considering the case where the fppf cover is a finite Galois cover, and by attaching 2-descent data to actions of the Galois group on the stack. 1.1. Main results In the case of stacks, the analogue of the aforementioned descent-theory for schemes is a notion called 2 -descent , which appears to have been introduced by Duskin [ 4 ] . As it turns out, 2 -descent data for algebraic stacks with respect to faithfully flat locally finitely presented morphisms of schemes are always effective. More precisely, we have the following result. For a scheme S , let ({\mathrm{Sch}}/S)_{fppf} be the big fppf site of S as in [ 12 , Tag 021S ] ; a stack over S is a stack in groupoids {\mathcal{X}}\to({\mathrm{Sch}}/S)_{fppf} , see [ 12 , Tag 0304 ] . Theorem 1.1 . Let S^{\prime}\to S be a faithfully flat locally finitely presented morphism of schemes, and let {\mathcal{X}}^{\prime} be a stack over S^{\prime} . Let (\phi,\psi) be a 2 -descent datum for the stack {\mathcal{X}}^{\prime} over S^{\prime} , see Definition 3.2 . Then the following holds. (1) The 2 -descent datum (\phi,\psi) is effective. That is, there exists a stack {\mathcal{X}} over S , an isomorphism of stacks over S^{\prime} , \rho\colon{\mathcal{X}}\times_{S}S^{\prime}\xrightarrow{\sim}{\mathcal{X}}^{\prime}, and a 2 -isomorphism \chi\colon p_{2}^{\ast}\rho\circ\textnormal{can}\Rightarrow\phi\circ p_{1}^{\ast}\rho as in the following diagram: \textstyle{p_{1}^{\ast}({\mathcal{X}}\times_{S}S^{\prime})\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} can \scriptstyle{{\Swarrow}} \scriptstyle{p_{1}^{\ast}\rho} \textstyle{p_{2}^{\ast}({\mathcal{X}}\times_{S}S^{\prime})\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{p_{2}^{\ast}\rho} \textstyle{p_{1}^{\ast}{\mathcal{X}}^{\prime}\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\phi} \textstyle{p_{2}^{\ast}{\mathcal{X}}^{\prime},} such that the natural compatibility between \chi and \psi is satisfied. (2) The stack {\mathcal{X}}^{\prime} over S^{\prime} is an algebraic stack over S^{\prime} if and only if the stack {\mathcal{X}} over S is an algebraic stack over S . (3) The stack {\mathcal{X}}^{\prime} over S^{\prime} is a Deligne–Mumford stack over S^{\prime} if and only if the stack {\mathcal{X}} over S is a Deligne–Mumford stack over S . Note that even the case where {\mathcal{X}}^{\prime} is a scheme seems to yield a non-trivial result (cf. Corollary 3.5 ). Of course, in some sense these results are not surprising: the descended stack {\mathcal{X}} is obtained by defining {\mathcal{X}}(T) as the groupoid of objects of {\mathcal{X}}^{\prime}(T\times_{S}S^{\prime}) equipped with a descent datum relative to the 2 -descent datum of {\mathcal{X}}^{\prime} , for any scheme T over S . More precisely, the first assertion in the above theorem follows from: Theorem 1.2 (Breen, Giraud) . Consider the 2 -fibred category \underline{Stack}_{S}\to({\mathrm{Sch}}/S)_{fppf}, whose fibre over U\in({\mathrm{Sch}}/S)_{fppf} is the 2 -category \underline{Stack}(U) of stacks over U . Then \underline{Stack}_{S} is a 2 -stack over S . Proof. See [ 3 , Example 1.11.(1)] and [ 5 , Chapitre II, §2.1.5] . ∎ The other two assertions in Theorem 1.1 follow from the fact that the property of a stack of being algebraic (resp. Deligne–Mumford) is local on the base for the fppf (resp. étale) topology; see Lemma 3.4 for a precise statement. For details, we refer to Section 3 . Assume that S^{\prime}\to S is a finite surjective étale morphism of schemes which is a Galois covering with Galois group \Gamma , acting on the left on S^{\prime} . Then for a stack {\mathcal{X}}^{\prime} over S^{\prime} , one can reformulate the notion of 2 -descent datum for {\mathcal{X}}^{\prime} over S^{\prime} in terms of an action of \Gamma on {\mathcal{X}}^{\prime} over the action of \Gamma on S^{\prime} over S , as in the classical case. To explain this, for an element \sigma\in\Gamma , define {{}^{\sigma}{\mathcal{X}}^{\prime}} as the fibre product {\mathcal{X}}^{\prime}\times_{S^{\prime},\sigma}S^{\prime} , viewed as a stack over S^{\prime} whose structure morphism {{}^{\sigma}{\mathcal{X}}^{\prime}}\to S^{\prime} is the pull-back of the morphism {\mathcal{X}}^{\prime}\to S^{\prime} along \sigma\colon S^{\prime}\to S^{\prime} . Definition 1.3 . Let S^{\prime}\to S be a finite surjective étale morphism of schemes which is a Galois covering with Galois group \Gamma . Let {\mathcal{X}}^{\prime} be a stack over S^{\prime} . A Galois 2 -descent datum consists of: (1) a family of 1 -isomorphisms of S^{\prime} -stacks f_{\sigma}\colon{{}^{\sigma}{\mathcal{X}}^{\prime}}\xrightarrow{\sim}{\mathcal{X}}^{\prime} (\sigma\in\Gamma) ; (2) a family of 2 -isomorphisms \psi_{\tau,\sigma}\colon f_{\sigma}\circ{{}^{\sigma}f_{\tau}}\Rightarrow f_{\tau\sigma} ((\tau,\sigma)\in\Gamma\times\Gamma) ; such that for each (\gamma,\tau,\sigma)\in\Gamma\times\Gamma\times\Gamma , the diagram of 2 -morphisms (1.1) \displaystyle\begin{split}\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 30.94743pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\cr&\crcr}}}\ignorespaces{\hbox{\kern-30.94743pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise-2.5pt\hbox{$\textstyle{f_{\sigma}\circ{{}^{\sigma}f_{\tau}}\circ{{}^{\tau\sigma}f_{\gamma}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}\hbox{\kern-1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 0.0pt\raise-20.55002pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.6389pt\hbox{$\scriptstyle{{f_{\sigma}}_{\ast}(^{\sigma}\psi_{\gamma,\tau})}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 0.0pt\raise-33.60004pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1.5}\lx@xy@tip{-1.5}}}}}}{\hbox{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}\hbox{\kern-1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}\hbox{\kern-1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\hbox{\kern 0.0pt\raise 1.0pt\hbox{\lx@xy@droprule}}\hbox{\kern 0.0pt\raise-1.0pt\hbox{\lx@xy@droprule}}}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 33.51112pt\raise 6.6582pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.68599pt\hbox{$\scriptstyle{(^{\tau\sigma}f_{\gamma})^{\ast}(\psi_{\tau,\sigma})}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 90.94743pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1.5}\lx@xy@tip{-1.5}}}}}}{\hbox{\hbox{\kern 0.0pt\raise 1.0pt\hbox{\lx@xy@droprule}}\hbox{\kern 0.0pt\raise-1.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\hbox{\kern 0.0pt\raise 1.0pt\hbox{\lx@xy@droprule}}\hbox{\kern 0.0pt\raise-1.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\kern 90.94743pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise-2.5pt\hbox{$\textstyle{f_{\tau\sigma}\circ{{}^{\sigma\tau}f_{\gamma}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}\hbox{\kern-1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 113.59177pt\raise-21.00835pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.44444pt\hbox{$\scriptstyle{\psi_{\gamma,\tau\sigma}}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 113.59177pt\raise-33.6556pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1.5}\lx@xy@tip{-1.5}}}}}}{\hbox{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}\hbox{\kern-1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}\hbox{\kern-1.0pt\raise 0.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\kern-20.8004pt\raise-41.10004pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise-2.5pt\hbox{$\textstyle{f_{\sigma}\circ{{}^{\sigma}f_{\gamma\tau}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\hbox{\kern 0.0pt\raise 1.0pt\hbox{\lx@xy@droprule}}\hbox{\kern 0.0pt\raise-1.0pt\hbox{\lx@xy@droprule}}}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 48.43542pt\raise-34.68338pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.44444pt\hbox{$\scriptstyle{\psi_{\gamma\tau,\sigma}}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 101.09448pt\raise-41.10004pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1.5}\lx@xy@tip{-1.5}}}}}}{\hbox{\hbox{\kern 0.0pt\raise 1.0pt\hbox{\lx@xy@droprule}}\hbox{\kern 0.0pt\raise-1.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\hbox{\kern 0.0pt\raise 1.0pt\hbox{\lx@xy@droprule}}\hbox{\kern 0.0pt\raise-1.0pt\hbox{\lx@xy@droprule}}}}{\hbox{\kern 101.09448pt\raise-41.10004pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise-2.5pt\hbox{$\textstyle{f_{\gamma\tau\sigma}}$}}}}}}}\ignorespaces}}}}\ignorespaces\end{split} is commutative. A group action (in the sense of [ 11 , Definition 1.3] ) of the finite group \Gamma on the stack {\mathcal{X}}^{\prime} over S , which is compatible with the action of \Gamma on S^{\prime} over S , gives rise to a Galois 2 -descent datum for {\mathcal{X}}^{\prime} , see Lemma 3.8 . Moreover, giving a Galois 2 -descent datum for {\mathcal{X}}^{\prime} is equivalent to giving 2 -descent datum for {\mathcal{X}}^{\prime} , see Lemma 3.6 . As a corollary of Theorem 1.1 , one therefore obtains: Theorem 1.4 . Let S^{\prime}\to S be a finite surjective étale morphism of schemes which is a Galois covering with Galois group \Gamma . Let {\mathcal{X}}^{\prime} be an algebraic stack over S^{\prime} , equipped with a Galois 2 -descent datum (f_{\sigma} (\sigma\in\Gamma) , \psi_{\tau,\sigma} (\tau,\sigma\in\Gamma)) . Then there exists

Read the complete text