4
$\begingroup$

Let $E$ be a stable vector bundle over a curve $X$. $K_X$ the canonical bundle of $X$. $W$ the base of the Hitchin map.

Is the Hitchin map $H:H^0(E\otimes E^*\otimes K_X)\rightarrow W$ surjective? If not, is there any conterexample?

$\endgroup$

1 Answer 1

3
$\begingroup$

There is one important case where $H$ is surjective, namely when $E$ is very stable. This means, by definition, that any nilpotent homomorphism $E\rightarrow E\otimes K$ is zero; or, equivalently, that $H^{-1}(0)=\{0\} $. It follows on one hand that $H$ is generically finite, hence dominant for dimension reasons, and on the other hand that $H$ descends to $\bar{H}:\mathbb{P}(H^0(\mathcal{E}nd(E)\otimes K))\rightarrow \mathbb{P}_w(W)$, where $\mathbb{P}_w(W)$ is the quotient of $W\smallsetminus\{0\} $ by $\mathbb{C}^*$ acting on $W=\bigoplus H^0(K^{i})$ by $\ t\cdot (\omega _i)=(t^{i}\omega _i)$. Then $\bar{H}$ is proper, hence surjective, and therefore $H$ is surjective.

$\endgroup$
3
  • $\begingroup$ This case has been proven in more general setting here link.springer.com/content/pdf/10.1007%2Fs10711-019-00447-z.pdf $\endgroup$
    – Z.A.Z.Z
    Commented Aug 19, 2019 at 16:51
  • $\begingroup$ I am interested on non very stable ones? $\endgroup$
    – Z.A.Z.Z
    Commented Aug 19, 2019 at 16:51
  • 2
    $\begingroup$ Oh, OK. You might have mentioned that in your question. $\endgroup$
    – abx
    Commented Aug 19, 2019 at 18:56

You must log in to answer this question.

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