1
$\begingroup$

A complex line bundle $L$ over a complex surface $X$ is said to be pseudo-effective if it admits a (possibly singular) Hermitian metric $h$ whose curvature is positive semi-definite in the sense of currents. That is, if we locally represent $h$ as $h = e^{\varphi}$, where $\varphi$ is an integrable real function, then $\Theta(h) = i \bar{\partial}\partial \varphi \geq 0$ (in the sense of currents).

All possible smooth compact complex minimal surfaces are listed in the famous Enriques-Kodaira classification. Denote by $K_X$ the canonical bundle of a complex surface $X$. I wonder about the following question:

  • Which minimal complex surfaces $X$ have $K_X$, $K^{-1}_X$ or both pseudo-effective?

I am sure this is well-known by experts and, in fact, one can go case by case through Enriques-Kodaira's list to get the answer. However, some of the cases in the list are riddled with subtleties, and I seem to fail to obtain a satisfactory answer. A comprehensive list answering this question would be very much appreciated, and hopefully, it would be of use to other people.

Thanks.

$\endgroup$

1 Answer 1

2
$\begingroup$

Edit: My answer deals with the projective case.

First of all, for any projective manifold, $K_X$ is pseudoeffective if and only if $X$ is not uniruled (Boucksom-Demailly-Paun-Peternell). In the case of surfaces, this is also known to be equivalent to $\kappa(X)\ge 0$.

Next, $K_X$ and $-K_X$ are both pseudoeffective if and only if $c_1(K_X)=0$. In the case of surfaces (then, automatically minimal), you recover abelian surfaces, bi-elliptic surfaces, K3 surfaces and Enriques surfaces.

Finally, in the uniruled case ($K_X$ not pseudoeffective), $X$ is birational to $\mathbb P^1\times C$ for some curve $C$. Then $-K_X$ is psef if and only if $C=\mathbb P^1$ or $C=E$ is an elliptic curve.

$\endgroup$
3
  • $\begingroup$ Thanks a lot for the answer. So, for example, if I assume $\kappa(X) = -\infty$ and neither $K_X$ nor $K^{-1}_X$ are pseudo-effective, then $X$ must be $VII^{+}_0$ or $VII^0_0$? Also, Hopf surfaces have $c_1(K_X)=0$, yet they don't have a pseudo-effective $\pm K_X$, am I correct? $\endgroup$
    – Bilateral
    Commented Oct 1, 2018 at 17:24
  • 1
    $\begingroup$ I am realizing that I forgot to insist on the fact that my answer only deals with projective surfaces. In the projective case, if you assume $\kappa=-\infty$ and $\pm K_X$ not psef, then $X$ is birational to $\mathbb P^1\times C$ where $C$ is a hyperbolic curve. Finally, $c_1(K_X)=0$ always implies (without projectivity assumption) that both $K_X$ and $-K_X$ are psef. $\endgroup$
    – Henri
    Commented Oct 1, 2018 at 17:27
  • $\begingroup$ Thanks. I was actually especially interested in the non-projective case. $\endgroup$
    – Bilateral
    Commented Oct 10, 2018 at 15:01

You must log in to answer this question.

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