2
$\begingroup$

Let $ Z $ be an $ n $-dimensional, projective, variety, over a field of arbitrary characteristic and let $ \iota: \mathbb G_{m}^{n} \to Z $ be a morphism such that for any $ z \in Z $, the fibre $ \iota^{-1}(z) $ is either empty or a single point. Note that I am not requiring that $ \iota(\mathbb G_{m}^{n}) $ is an open, sub-variety of $ Z $. Does anyone know of some object which measures whether $ Z $ is toric? Namely, does someone know of something that measures when the action of $ \mathbb G_{m}^{n} $ extends to an action on $ Z $?

$\endgroup$
3
  • 1
    $\begingroup$ $\mathbb G_m$ worked for me, I edited your post $\endgroup$ Mar 10, 2023 at 7:54
  • $\begingroup$ Thank you @NicolasHemelsoet. It may be a computer issue. $\endgroup$
    – Schemer1
    Mar 10, 2023 at 7:56
  • 3
    $\begingroup$ There are many conditions. For example, if $Z$ is smooth and the complement $D$ of the torus is a divisor with normal crossings, then $Z$ is toric (in the sense that the action of the torus extends) if and only if the log cotangent bundle $\Omega_Z(\log D)$ is trivial, by a theorem of Winkelmann. $\endgroup$ Mar 10, 2023 at 9:53

2 Answers 2

5
$\begingroup$

Even in the case when the image of the morphism is open, there are plenty of non-toric examples. Consider the standard toric action on $\mathbb{CP}^3$ and consider one of the torus invariant divisors $H \cong \mathbb{CP}^2$, then blow up any smooth curve in $H$ which is not a line. The result has automorphism group of rank $<3$ hence non-toric, and contains a copy of $(\mathbb{C}^*)^3$.

In these examples, the non-toricness is measured by the rank of the automorphism group. So perhaps a reasonable condition to add is that it has rank $3$, although maybe in that case there are also counter-examples.

$\endgroup$
5
$\begingroup$

If you consider the pair $(Z,\Delta)$, where $\Delta=Z\setminus i(\mathbb{G}_m)$ is the divisor given as the complement of the image of the torus, then there is an numerical invariant of $(Z,\Delta)$ called the complexity which can be used to characterise whether or not $Z$ is toric. It was conjectured by Shokurov and proved in this paper by Brown, M$^{\text{c}}$Kernan, Svaldi & Zong.

$\endgroup$

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.