5
$\begingroup$

Let $X$ be a connected projective scheme over $\mathbb{C}$ and $E$ a coherent sheaf on $X$. Consider the Quot scheme $\operatorname{Quot}_X(E,P)$ of quotients of $E$ of fixed Hilbert polynomial $P$. As far as I know, this scheme is connected at least in the following cases:

  1. for Hilbert schemes on $\mathbb{P}^n$ (by Hartshorne)
  2. for punctual Quot schemes on smooth surfaces (by Fogarty, Ellingsrud–Lehn, Baranovski, etc.).

Are there any other known cases for which $\operatorname{Quot}_X(E,P)$ is connected? Could you give some examples where the connectedness fails? I am particularly interested in higher dimensions or singular surfaces/varieties.

$\endgroup$

2 Answers 2

6
$\begingroup$

The simplest example when connectedness fails is the Hilbert scheme of lines on a quadric surface: $$ \mathrm{Hilb}_{1+t}(\mathbb{P}^1 \times \mathbb{P}^1) = \mathbb{P}^1 \sqcup \mathbb{P}^1, $$ where the polarization is by the line bundle $\mathcal{O}(1,1)$.

$\endgroup$
2
  • 2
    $\begingroup$ You cannot “fix” this by using a “refined” Hilbert polynomial: let $Y$ in $\mathbb{P}^3\times\mathbb{P}^9$ be the universal quadric surface, and let $X$ be $Y\times\mathbb{P}^1$. The cone of effective curve classes on this Fano manifold is the cone of triples of nonnegative integers: the degrees of the projections to $\mathbb{P}^1$, to $\mathbb{P}^3$, and to $\mathbb{P}^9$. Consider the curve class where the triple is $(1,4,0)$. The actual curves of arithmetic genus $0$ project to curves in $Y$ in a fiber ,I.e., a quartic in a quadric surface. The curve class might be $(2,2)$ or $(3,1)$. $\endgroup$ Feb 22, 2022 at 7:07
  • $\begingroup$ Please note that this $X$ is a Fano manifold. This does not contradict Cohen-Jones-Segal, because they require the curve class to be far from the boundary of the cone of effective curve classes. $\endgroup$ Feb 22, 2022 at 7:18
1
$\begingroup$

Let $X$ be an irreducible variety, $E$ a locally free sheaf on $X$, and $n \geq 0$ an integer. The Quot scheme $\textrm{Quot}_X(E,n)$ is connected by Thm. 1.4 here. To achieve disconnectedness, probably one has to look at nonconstant polynomials, as Sasha's example suggests.

$\endgroup$
2
  • $\begingroup$ do you think it is necessary to assume that $X$ is irreducible (rather than connected). It seems to me that the result should be true if $X$ is just connected. $\endgroup$
    – Samuel
    Feb 28, 2022 at 11:29
  • $\begingroup$ you're probably right! $\endgroup$ Mar 1, 2022 at 20:33

Your Answer

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

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