4
$\begingroup$

Let $T$ be a compact torus, and $X$ its blow-up in a point (or in several points). It seems that $X$ is K-stable for any Kahler form on $X$. Is there a reference to this?

Also, what can we say about the constant scalar curvature Kahler metrics on this blow-up? Do they exist in all Kahler classes? They do exist in some Kahler classes, by the famous result of Arezzo and Pacard:

Claudio Arezzo and Frank Pacard, Blowing up and desingularizing constant scalar curvature Kähler manifolds, Acta Math. Volume 196, Number 2 (2006), 179-228.

If I am right and the standard conjectures about K-stability are true, it seems that the CSC Kahler metrics should exist for all Kahler classes.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

This statement about blow ups of torus is not correct. Take any aglebraic $2$-torus with a smooth curve $C$ of genus $>1$. Blow up $C^2+1$ points on $C$ and apply Theorem 1 here: https://arxiv.org/pdf/0710.4078.pdf

Improve this answer
$\endgroup$
3
  • $\begingroup$ and what if the torus is non-algebraic? $\endgroup$
    – Misha Verbitsky
    Jun 29, 2017 at 5:00
  • $\begingroup$ For the application I have in mind, non-algebraic tori are more than sufficient $\endgroup$
    – Misha Verbitsky
    Jun 29, 2017 at 5:02
  • $\begingroup$ I would bet that no positive result in this direction that is stronger than Arezzo-Pacard is known. I would guess (but this is only a guess) that using Theorem 1.2 here arxiv.org/pdf/0807.1716.pdf one could construct example similar to what I gave you by making an iterated blow up of a non-algebraic torus. This would lead to appearance of quite "bad" divisors there, composed of rational curves. But probably you should ask experts like Yuji Odaka, or Gabor Szekelyhidi $\endgroup$
    – aglearner
    Jun 29, 2017 at 10:21

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.