9
$\begingroup$

It seems to me that many classical statements about Cech cohomology should without much effort follow from Lurie's HTT, but I am struggling to fill out the details. I want to show that, given an ordinary sheaf $A \in \operatorname{Sh}(X, \operatorname{Ch}(\mathbb{Z}))$ of chain complexes on a topological space $X$,

  • The functor that sends an open subset $U$ to the derived sections $\operatorname{R \Gamma}(U,A) \in D(\mathbb{Z})$ in the derived $\infty$-category is an $\infty$-sheaf.
  • If $A = F[n]$ is concentrated in degree $n >0$, for $F$ an ordinary sheaf of abelian groups, then (under the Dold-Kan correspondence) $A$ corresponds to the Eilenberg-Maclane object $B^n F$ defined in HTT, as suggested in this n-lab article.

From the first statement (if it is true), one can show Theorem 3.8 on the correspondence between Cech and sheaf cohomology in the n-lab article on Cech cohomology, by using the fact that the Cech complex calculates the homotopy limit that describes descent for $\infty$-sheaves. I also want to prove Theorem 3.7:

  • On a paracompact Hausdorff space $X$, Cech cohomology and sheaf cohomology agree: $$ \operatorname{\check{H}}^*(X,A) \cong \operatorname{H}^*(X,A) $$ Here, the Cech complex is defined by taking the colimit over all covers of $X$.

The mentioned nlab articles of course give references on the classical proofs of these statements (as well as, in the first case, a proof via simplicial cosheaves); but I am interested in whether, and how, one can abstractly deduce them from results in Lurie's books. In particular, Chapter 7 in HTT develops a lot of theory for paracompact Hausdorff spaces, and this would be an interesting application. I would be very greatful for some hints or references on how this works.

$\endgroup$
10
  • 1
    $\begingroup$ The first statement is basically a tautology, depending on your definitions. The point is that every covering sieve becomes a colimit diagram in the topos, so taking sections yields a sheaf. $\endgroup$
    – Zhen Lin
    Commented Jan 15, 2023 at 11:45
  • 1
    $\begingroup$ Ah. Actually, I think that definition probably does not work at all. You need to think of $A$ as an object in the derived category of chain complexes of sheaves. $\endgroup$
    – Zhen Lin
    Commented Jan 16, 2023 at 12:19
  • 1
    $\begingroup$ The point is that you have to give an ($\infty$-categorical) definition of $R\Gamma$. There are some tricky point such as hypercompleteness. $\endgroup$
    – Z. M
    Commented Jan 16, 2023 at 13:40
  • 1
    $\begingroup$ @MarkusZetto So you are asking for a proof that the derived category of chain complexes of sheaves is equivalent to the category of sheaves of objects in the derived category of chain complexes (inserting $\infty$ appropriately)? That is actually a non-trivial fact, but the direction you are interested in is the easy one. $\endgroup$
    – Zhen Lin
    Commented Jan 16, 2023 at 15:12
  • 1
    $\begingroup$ No, I do not have a reference. But I'm sure someone does and will give you one if you ask a separate question specifically about it. At the most basic level this is a question about strictification: if we forget about coverages for a moment this reduces to showing that the homotopy $(\infty, 1)$-category of the category of presheaves valued in a model category $\mathcal{M}$ is equivalent to the $(\infty, 1)$-category of presheaves valued in the homotopy $(\infty, 1)$-category of $\mathcal{M}$. $\endgroup$
    – Zhen Lin
    Commented Jan 20, 2023 at 22:55

0

You must log in to answer this question.