MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

If a group $G$ acts on a topological space $M$, and a representation of $G$ on a vector space $V$, why $M \times_G V$ is a local constant sheaf over $M/G$?

share|cite|improve this question
What structure maps are you pulling back? – Harry Gindi Dec 6 '10 at 15:22
@Harry: the notation is not for a fiber product. It means the quotient of $M×V$ by the diagonal $G$-action (perhaps with an inverse thrown in somewhere). – Sheikraisinrollbank Dec 6 '10 at 15:41
Maybe I'm mistaken, but if $G={1}$ is the trivial group then what you wrote is not a constant sheaf. But maybe you should explain better what sheaf you are actually talking about. G invariant sections of the bundle? – Michael Bächtold Dec 6 '10 at 15:48
Presumably what the OP means is why (or under what circumstances, I'm slightly uncomfortable with the paucity of hypotheses) is $M \times_G V$ locally trivial, rather than locally constant. – Sheikraisinrollbank Dec 6 '10 at 15:55

Dear HYYY,

You should assume that the group $G$ acts freely and properly discontinuously on $M$. Also, you should equip $V$ with the discrete topology. (This way you will get a local system rather than a vector bundle, which was a point of confusion in the comments above.)

If $G$ acts freely and properly discontinuously, then if $m$ is any point of $M$, it has a neighbourhood $U$ such that $Ug$ is disjoint from $U$ for all non-trivial $g \in G$. (Here I am writing the $G$-action on $M$ on the right, as is implicit in the question.)

Thus $U$ maps injectively into $M/G$, and $U \times V$ maps injectively into $M\times_G V$. Hence $M\times_G V$ is locally constant (because, as we have just shown, its pull-back over the open subset $U$ of $M/G$ is constant).

share|cite|improve this answer
@Emerton, thanks!but how can we see the action of $G$ on $V$ from this local system? If we choose a point $x$ in $M$,do we get a section $c\in V$ of this local system over the point $[x]$ in $M/G$?And then if we choose another point $y$ which in the same orbit of $x$,assume $yg=x$,then we get another section $c'\in V$ over $[y]=[x]$,is that $gc=c'$? – HYYY Dec 7 '10 at 7:30
Are you asking how to go backwards, i.e. how to recover $V$ from the associated local system? – Emerton Dec 8 '10 at 4:25

For a bit of intuition, consider $V = \mathbb{C}^2$ and $M = \mathbb{C}$ with $G = \mathbb{Z}/2$ acting diagonally as $\pm 1$ in both cases.

Then for any open set $U$ not containing the origin in the base, the sheaf is just the constant sheaf $\mathbb{C}^2$. If, however, $U$ does contain the origin, then it is the constant sheaf $\mathbb{C}^2/\pm 1$, and so this sheaf is locally constant.

In general, the fibre over a point $[x]$ in the base should be a copy of $V/stab_G(x)$. You should be able to study these to obtain your result.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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