6
$\begingroup$

Let $n\geq2$. Let $G$ be a linear automorphisms group of prime order on $\mathbb{C}^n$. We assume that $0$ is the unique fixed point of $G$. I consider the quotient $\mathbb{C}^n/G$. It is a toric variety, so I can consider the toric blowup: $\widetilde{\mathbb{C}^{n}/G}$. I am interesting in the integral cohomology (singular cohomology) of $\widetilde{\mathbb{C}^{n}/G}$. Especially, I would like to prove that $H^{2k}(\widetilde{\mathbb{C}^{n}/G},\mathbb{Z})$ is torsion free for $k\leq n-1$. Do someone know an efficient method to deal with this kind of problems? Has this cohomology been studied somewhere?

I managed to prove that $H^2(\widetilde{\mathbb{C}^{n}/G},\mathbb{Z})$ and $H^{2n-2}(\widetilde{\mathbb{C}^{n}/G},\mathbb{Z})$ are torsion free using Danilov combinatorial result. However for the other cohomology groups, it becomes too technical to be done by hand.

$\endgroup$
1
  • $\begingroup$ Culturally, this relates to McKay correspondence, as in Miles Reid paper arxiv.org/abs/alg-geom/9702016. For example, if G acts with trivial determinant, the dimension of the total cohomology of the resolution is supposed to be equal to the number of irreducible representations. However, I suppose toric McKay correspondence is not known in higher dimension. It would be nice to formulate your question in terms of geometry of the quotient stack [C^n / G], but I don't see how to do it... $\endgroup$ Commented Nov 2, 2018 at 16:44

1 Answer 1

2
$\begingroup$

You can find an answer here (Proposition 3.14).

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .