1
$\begingroup$

Hi,

I need a precise reference for the following fact, which is certainly well known, but I do not find any.

I consider the natural glueing map of pointed curves $\overline{M}_{g_1,n}\times \overline{M}_{g_2,n} \rightarrow \overline{M}_{g_1+g_2+n-1}$. The pullback of the Hodge bundle such be equal to the direct sum of the two hodge bundles of the factors of the product plus n-1 trivial factors. where is this written properly?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

I don't know a reference but here is an argument. It suffices to consider the case when we glue two points only.

Suppose that we are given a curve $C \to S$ (not necessarily connected) and two disjoint sections, $D$ is the result of gluing these together, and $\nu \colon C \to D$ the natural map. Then there is an exact sequence $$ 0 \to O_D \to \nu_\ast O_C \to O_S \to 0$$ of sheaves on $D$. Push forward along $p \colon D \to S$. If $C$ has two components and the sections are on distinct components, we see that the first row in the long exact sequence is exact and $R^1 p_\ast O_D \cong R^1 p_\ast O_C$. If $C$ is connected we instead get a short exact sequence $0 \to O_S \to R^1p_\ast O_D \to R^1p_\ast O_C \to 0$. Dualizing gives the result for the Hodge bundle.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Thank you. Is it clear that the extension is the trivial map? $\endgroup$
    – IMeasy
    Commented Jun 22, 2012 at 16:07
  • $\begingroup$ I'll have to think about it, but it is not immediately clear to me. $\endgroup$
    – Dan Petersen
    Commented Jun 22, 2012 at 22:44

You must log in to answer this question.

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