10
$\begingroup$

I am looking for some references on equivariant resolution of singularities. In most references quoted on mathoverflow (for instance : Reference on an equivariant resolution of singularities), they only talk about finite group action (if I am not mistaken).

I was wondering if it is known that equivariant resolutions do not exist in general for larger group (are there counter-examples?). I am especially interested when the group is $\mathbb{C}^*$.

Thanks in advance.

$\endgroup$
4
  • 5
    $\begingroup$ Equivariant resolutions exist for the action of any algebraic group. See, for example, 3.9.1 in Kollar's book "Lectures on resolution of singularities". $\endgroup$
    – naf
    Sep 26, 2013 at 10:58
  • $\begingroup$ @ulrich Thanks! I still wonder why many papers only deal with finite group actions. But that does not really matter, now I have this reference. $\endgroup$
    – Libli
    Sep 26, 2013 at 12:10
  • $\begingroup$ I've need to reference this in a paper and used: Cor 7.6.3, O. E. Villamayor U. Patching local uniformizations. Ann. Sci. E ́cole Norm. Sup. (4), 25(6):629–677, 1992. $\endgroup$
    – Jim Bryan
    Sep 26, 2013 at 18:00
  • $\begingroup$ @ulrich In the proof of the proposition 3.9.1, it seems that for Kollar, a group action is, in particular a smooth morphism $G\times X\rightarrow X$, or am I mistaken ? (or maybe for him smooth morphism only means that any fiber is smooth??) $\endgroup$
    – pi_1
    Aug 18, 2020 at 12:28

1 Answer 1

6
$\begingroup$

To any variety $X$ over a field of characteristic zero, say $k$, one can attach a resolution of singularities, say $X'\to X$, with the following properties:

  1. $X'\to X$ is an isomorphism over $X_{\mathrm{reg}}$.
  2. if $\Theta:X\rightarrow Y$ is an isomorphism of schemes (not necessarily defined over $k$), then $\Theta$ can be extended to an isomorphism $\Theta':X'\rightarrow Y'$ compatible with $\Theta$ and the resolutions.

It follows from 1. and 2. that if $\Psi:Y\rightarrow Z$ is a second isomorphism of schemes, then $(\Psi\circ\Phi)'$ and $\Psi'\circ\Phi'$ coincide on the inverse image of $X_{\mathrm{reg}}$ in $X'$, and hence they coincide on $X'$. With a similar argument, we have $(\mathrm{id}_X)'=\mathrm{id}_{X'}$. In particular, if $G$ is any group of automorphisms of $X$ (not necessarily defined over $k$), the action of $G$ on $X$ can be lifted to an action on $X'$.

For a reference, see:

O. Villamayor: Equimultiplicity, algebraic elimination, and blowing-up. Adv. in Math. 262 (2014) 313-369. (ArXiv link, DOI link)

$\endgroup$

Your Answer

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

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