stacks MOC

Last modified date: <%+ tp.file.last_modified_date() %>


stacks MOC

Idea: stacks are geometrically modeled on sites \(\mathsf{S}\), and e.g. \({\mathsf{Grpd}}\) is a stack modeled on \(\mathsf{S} = {\mathsf{Set}}\) with the discrete topology.

attachments/Pasted%20image%2020220420101720.png

attachments/Pasted%20image%2020220802130931.png attachments/Pasted%20image%2020220802131002.png

Write \(D\) for the dual numbers. attachments/Pasted%20image%2020220802131151.png attachments/Pasted%20image%2020220802131205.png attachments/Pasted%20image%2020220802131319.png attachments/Pasted%20image%2020220802131346.png

In moduli problems

attachments/Pasted%20image%2020220426013106.png attachments/Pasted%20image%2020220426013211.png

Definitions

attachments/Pasted%20image%2020220508201920.png

Informal definition: attachments/Pasted%20image%2020220421231305.png

In terms of a pseudofunctor:

attachments/Pasted%20image%2020220319213425.png

  • An algebraic space is a pair \((X, R)\) with \(R \subseteq X{ {}^{ \scriptscriptstyle\times^{2} } }\) an equivalence relation whose projections \(p_i: R\to X\) are etale morphisms. Idea: replace being locally isomorphic to affine space in the Zariski topology with the finer etale topology.

  • A prestack is a functor \({\mathsf{Aff}}{\mathsf{Sch}}_{/k}^{\operatorname{op}}\to {\mathsf{hoType}}\)

  • The prestack of quasicoherent sheaves over ${\mathsf{Sch}}_{/ {S}} $ is a stack wrt the fpqc topology.

  • A stack is a functor $M: \mathsf{C}\to {\mathsf{Sch}}_{/ {S}} $ that satisfies effective descent.

  • A 1-stack of groupoids on \(\mathsf{C}\) is a category fibered in groupoids \({\mathcal{X}}\to \mathsf{C}\) satisfying certain descent conditions. This form a category ${\mathsf{St}}^1 \leq { \underset{\infty}{ \mathsf{Cat}} }{} _{/ { \mathsf{C} }} $, so morphisms are cones (of functors) over \(\mathsf{C}\). For morphisms, \(f\simeq g\) means there is a natural transformation from \(f\) to \(g\) commuting with the projections to \(\mathsf{C}\), so one can form a homotopy category f 1-stacks.

  • A smooth proper stack is essentially a compact orbifold.

In terms of sheaves

See site.

attachments/Pasted%20image%2020220503235945.png attachments/Pasted%20image%2020220504000008.png

Artin Stack

attachments/Pasted%20image%2020220228092934.png

Algebraic Stacks

attachments/Pasted%20image%2020220323191326.png

attachments/Pasted%20image%2020220220034958.png

Geometric stacks

As in the case of the cotangent complex:

attachments/Pasted%20image%2020220319213617.png attachments/Pasted%20image%2020220319213700.png

Sheaves on Stacks

attachments/Pasted%20image%2020220228093141.png

Quotient stacks

attachments/Pasted%20image%2020220508201947.png

Examples

  • \({{\mathbf{B}}G}\): defined as \({\mathsf{G}{\hbox{-}}\mathsf{Torsors}} \leq G{\hbox{-}}{\mathsf{Set}}\) in terms of torsors. See BG, constructed as the quotient stack \([{\operatorname{pt}}/G]\).
  • For any \(X\in G{\hbox{-}}{\mathsf{Set}}\), \({\mathbf{B}}GX = [X/G]\) whose objects are ${\mathsf{G}{\hbox{-}}\mathsf{Torsors}}{/ {X}} \leq G{\hbox{-}}{\mathsf{Set}}{/ {X}} $.
  • \({\mathsf{Rep}}(G)\) can be interepreted as a category of sheaves on the stack \({{\mathbf{B}}G}\).

Quotient stacks and Picard stacks, in terms of torsors: attachments/Pasted%20image%2020220319213511.png

Exercises

  • Show that \([{\mathbf{A}}^1_{/ {k}} /{\mathbf{Z}}]\) for \(\operatorname{ch}k = 0\) is an algebraic space that is not quasiseparated.

Results

  • Working with a moduli stack instead of a moduli space allows pretending the space is smooth and admits a universal family.

  • Neat trick from algebraic geometry: for a stack \(X/G\) where $X \in{\mathsf{Var}}_{/ {{\mathbf{C}}}} $ and \(G \in {\mathsf{Fin}}{\mathsf{Grp}}\), \begin{align*} H^*(X/G; {\mathbf{Q}}) \cong H^*(X; {\mathbf{Q}})^G \end{align*} where the RHS denotes the taking the \(G{\hbox{-}}\)invariants. Seems to only work over \({\mathbf{Q}}\). The quotient is scheme-theoretic. The actual definition involves equivariant cohomology.

  • fpqc stack \(\implies\) fppf, etale stack

🗓️ Timeline
  • Prismatic cohomology

    Now has a better chance of being an algebraic stack instead of a formal stack. Bottom arrow kills the formal direction.

  • 2021-10-08

    Refs: stacks vector bundle Unsorted/descent

  • 2021-09-19

    So the condition of \({\mathcal{F}}\) being a sheaf seems to look like letting \({\mathcal{U}}\rightrightarrows X\) be an open cover, setting \(M = {\textstyle\coprod}U_i\), then applying a bar construction \begin{align*} M: M{ {}^{ \scriptscriptstyle \underset{\scriptscriptstyle {X} }{\times} ^{1} } } \leftarrow M{ {}^{ \scriptscriptstyle \underset{\scriptscriptstyle {X} }{\times} ^{2} } } \leftarrow\cdots .\end{align*} Then apply \({\mathcal{F}}\), and look at some kind of image sequence? And ask for exactness for \(n\) many levels to get a sheaf, Unsorted/stacks MOC, etc:

  • 2021-06-06
    stacks: presheaves of groupoids? See a stack is a category fibered in groupoids.
  • 2021-05-01

    As a general philosophy, one should expect that moduli space problems whose objects have nontrivial automorphisms are representable by Unsorted/stacks MOC, and those without nontrivial automorphisms are representable by scheme.

  • 2021-04-22

    schemes and stacks can be very singular.

  • 2021-03-24
    Unsorted/stacks MOC and algebraic stack :
Links to this page
#AG #MOC #higher-algebra/DAG #AG/moduli-spaces #resources/full-courses