5
$\begingroup$

Let $k$ be an algebraically closed field of characteristic $p>0$ and write $W:=W(k)$ for its ring of Witt vectors. Consider a smooth cubic fourfold $X_{0}\subset\mathbb{P}^{5}_{k}$ and let $F(X_{0})$ be its Fano scheme of lines. Then the incidence correspondence \begin{equation*} F(X_{0})\leftarrow I(X_{0})\subset X_{0}\times_{k} F(X_{0})\rightarrow X_{0} \end{equation*} induces a morphism on cohomology \begin{equation*} H_{\text{cris}}^{4}(X_{0}/W)\rightarrow H_{\text{cris}}^{2}(F(X_{0})/W) \end{equation*} Does $a$ induce an isomorphism on the primitive crystalline cohomology?


Motivation:-

For primes $\ell\neq p$, consider the morphism \begin{equation*} H_{\text{ét}}^{4}(X_{0},\mathbb{Z}_{\ell})_{\text{prim}}\rightarrow H_{\text{ét}}^{2}(F(X_{0}),\mathbb{Z}_{\ell})_{\text{prim}} \end{equation*} induced by the incidence correspondence. This is an isomorphism because $X_{0}$ lifts to a cubic hypersurface $X/W$, and then we use the smooth proper base change theorem to reduce to considering the corresponding morphism over $\mathbb{C}$, which is an isomorphism by Beauville and Donagi's paper ``La variete des droites d'une hypersurface cubique de dimension 4''.

$\endgroup$
2
  • $\begingroup$ Is something wrong with the proof of the isomorphism in the Beauville--Donagi paper? Doesn't it work in the crystalline situation? $\endgroup$
    – Sasha
    Commented May 11, 2016 at 16:52
  • $\begingroup$ @Sasha , that's a good question. I guess the reason for my question is that I'm just not confident in saying that the argument does or does not carry over to the crystalline case. If it does work then I'd be very happy. If it doesn't, and somebody points out why, then I'll be equally happy since then I would've learned something about what can go wrong when trying to transplant this type of argument into a crystalline context. $\endgroup$ Commented May 12, 2016 at 8:56

0

You must log in to answer this question.

Browse other questions tagged .