3
$\begingroup$

Suppose that $f:X\rightarrow S$ is a proper, separated morphism of complex spaces (with $S$ reduced) and $\mathcal{F}$ a is $f$-flat coherent sheaf on $X$.

From (well-)known results it is known that the set of points $U$ where $\mathcal{F}$ is cohomologically flat in dimension $q$ is a Zariski-open subset of $S$.

Does anybody know an example where $U$ is empty? I'm especially interested in the case where $q=0$.

Same question for a variety or scheme over $\mathbb{C}$.

$\endgroup$
6
  • 1
    $\begingroup$ If $S$ is reduced, the set $U$ is a dense Zariski open subset. (I assume that by "complex spaces" you mean "locally finitely presented algebraic spaces over $\mathbb{C}$".) $\endgroup$ Jun 12, 2017 at 13:10
  • $\begingroup$ With complex space I mean complex analytic space as in Grauert-Remmert. Gluing vanishing loci of holomorphic functions on open subsets of $\mathbb{C}^n$. $\endgroup$
    – Horstenson
    Jun 12, 2017 at 13:18
  • 2
    $\begingroup$ Then please explain what you mean by "Zariski open subset". $\endgroup$ Jun 12, 2017 at 13:23
  • $\begingroup$ The set where $\mathcal{F}$ is not cohomologically flat is an analytic set. Should probably just have written open but I'm also interested in variety/scheme case. $\endgroup$
    – Horstenson
    Jun 12, 2017 at 13:28
  • 1
    $\begingroup$ Thanks, do you have a reference for that? Or am I missing something obvious? $\endgroup$
    – Horstenson
    Jun 12, 2017 at 13:32

1 Answer 1

2
$\begingroup$

For simplicity, assume that $S$ is affine (every algebraic space has a dense open subset that is an affine scheme). In the case of a proper morphism of Noetherian algebraic spaces,$$ f:X\to S,$$ for $\mathcal{F}$ an $f$-flat coherent sheaf, there exist (locally on $S$) a bounded complex $K^\bullet$ of locally free $\mathcal{O}_S$-modules and an equivalence of contravariant functors, $$H^p(X\times_S T,\text{pr}_X^*\mathcal{F}) \cong h^p(K^\bullet\otimes_{\mathcal{O}_S(S)}\mathcal{O}_T(T)),$$ on the category of algebraic spaces $T$ with an affine morphism $T\to S$. The proof is essentially the same as the proof of the Theorem on p. 46 of the following.

MR2514037 (2010e:14040)
Mumford, David
Abelian varieties.
With appendices by C. P. Ramanujam and Yuri Manin.
Corrected reprint of the second (1974) edition.
Tata Institute of Fundamental Research Studies in Mathematics, 5.
Published for the Tata Institute of Fundamental Research, Bombay; by Hindustan Book Agency, New Delhi, 2008. xii+263 pp.

The bounded complex $K^\bullet$ has finitely many nonzero cohomology sheaves, and these are each coherent. If $S$ is reduced, these coherent sheaves are each locally free on a dense Zariski open by Grothendieck's Generic Flatness Theorem. Thus, on a dense Zariski open, the complex $K^\bullet$ has locally free cohomology sheaves. Thus, all of the cohomologies split, and the complex is quasi-isomorphic to the direct sum of its locally free cohomology sheaves. Replacing $S$ by this dense Zariski open, every $H^p(X\times_S T,\text{pr}_X^* \mathcal{F})$ is naturally equivalent to the locally free $\mathcal{O}_T$-module $h^p(K^\bullet)\otimes_{\mathcal{O}_S(S)}\mathcal{O}_T(T).$ This is compatible with arbitrary base change.

$\endgroup$
3
  • $\begingroup$ Thank you for your nice and detailed answer. I'll leave the question open for now since I'm also interested in the complex analytic case. $\endgroup$
    – Horstenson
    Jun 20, 2017 at 12:28
  • $\begingroup$ I recommend you check the articles of Bingener and Flenner (and there is probably also a book of Buchweitz and Flenner) on the complex analytic analogues of Artin's algebraization theorems. The result in my answer above is one step in the proof of Artin's algebraization theorem. I suspect that those articles prove the analogous complex analytic result or cites an article where this is proved. $\endgroup$ Jun 22, 2017 at 22:45
  • $\begingroup$ I looked through the work of Bingener but I don't understand it well enough yet. Thanks for recommendation of the book by Buchweitz and Flenner, seems great! $\endgroup$
    – Horstenson
    Jun 24, 2017 at 15:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.