6
$\begingroup$

In the standard references for prismatic cohomology, most theorems are proved in a local context (i.e. with completeness assumptions), and the devissage to the global case (i.e. smooth proper varieties) is often left to the reader. My first (vague) question is: how do these devisagges work? Are there references where these arguments have been written up (possibly in other contexts, e.g. crystalline cohomology)? For example, in Bhatt's Columbia notes he says that he will work in the local $p$-adically complete setting, because standard arguments then yield the general case: I would like to know what these arguments are.

To make things concrete, let me ask a more specific question. How to compute the mod $p$ prismatic cohomology of $\mathbb{P}^n_k$ (I would expect the answer to be $\mathbb{F}_p[[T]][h]/h^{n+1}$)? Etale comparison gives us the torsion-free part, but how to show that there is no torsion?

$\endgroup$
4
  • $\begingroup$ The theory seems to be formulated for formal schemes (see e.g. the original paper, Def. 4.1). By formal GAGA, in the proper case you can think about it as varieties, but at least that explains the ubiquity of completions. In §4.3 ("generalities on computing prismatic cohomology") they say some words on the (formal) affine line. $\endgroup$ Commented Jun 21, 2020 at 22:18
  • $\begingroup$ I'm not sure I understand your concrete question at the end. What is $k$? And what is the prism structure? $\endgroup$ Commented Jun 21, 2020 at 22:20
  • $\begingroup$ Thank you for the answer to the first question. As to the second, for some reason I was thinking of crystalline cohomology. Would $k=\mathbb{Z}_p[[t]]$ work? I don't know how to answer regarding the prism structure. I mean cohomology of the structure sheaf $\mathcal{O}_{\Delta}$. Is there extra information that I have to give? Do I have comparison with crystalline and etale for all the choices of this extra information? I did not notice that while reading the notes. $\endgroup$
    – Vitay
    Commented Jun 21, 2020 at 23:23
  • $\begingroup$ I'll go back to studying. $\endgroup$
    – Vitay
    Commented Jun 22, 2020 at 0:59

0

You must log in to answer this question.

Browse other questions tagged .