Remarks on Relative Canonical Bundles and Algebraicity Criteria for Foliations in Kähler context

Junyan CAO, Mihai PĂUN · 2025

Read Remarks on Relative Canonical Bundles and Algebraicity Criteria for Foliations in Kähler context by Junyan CAO, Mihai PĂUN (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-02183.txt.

Introduction

Junyan Cao and Mihai Paun pursue, in a 2025 note, three aims on foliations in the Kahler setting. Foliations split a space into leaf-like parts, a frame used in complex geometry. The first aim is progress on the positivity of the relative canonical bundle, a key object that steers much of this theory. That point has been open in the Kahler case, where the space holds a rich metric form. The note takes steps on it in plain terms.

The second aim is a proof of a test for algebraicity due to W Ou, a peer's recent claim. A space or leaf that passes the test counts as algebraic, built from polynomial rules. The third aim extends Ou's test for when a space is uniruled, swept by rational curves. Each part builds on the parts before it. The note is short and dense, aimed at experts.

Catalog record

CREATORJunyan CAO, Mihai PĂUN
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE71 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-02183

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Remarks on Relative Canonical Bundles and Algebraicity Criteria for Foliations in Kähler context Junyan CAO, Mihai PĂUN Abstract 1. Introduction In this note, motivated by the recent preprint [ 37 ] , we pursue three main objectives. The first is to make progress towards the fundamental conjecture proposed in [ 16 ] by Cao-Höring. Conjecture 1.1 . [ 16 ] Let p:X\to Y be a holomorphic surjective map between two compact Kähler manifolds. Let (L,h_{L}) be a line bundle over X , endowed with a possibly singular metric h_{L} such that i\Theta_{h_{L}}(L)\geq 0 on X in the sense of currents. Assume that the following conditions hold: • The multiplier ideal sheaf of h_{L} is trivial. • The restriction of the adjoint bundle K_{X}+L to the fiber X_{y} of p over y , denoted as \displaystyle(K_{X}+L)|_{X_{y}} , is pseudo-effective for generic y\in Y . Then the twisted relative canonical bundle K_{X/Y}+L is pseudo-effective. In this generality, the main difficulty is that there is no obvious candidate for a natural, positively curved metric on K_{X/Y}+L (as opposed to the projective case). Nevertheless, we establish the following particular case of Conjecture 1.1 (the definition of D(p) is recalled in the next section). Theorem 1.2 . In the setting of Conjecture 1.1 , further assume that there exists a Kähler metric \omega and a holomorphic 2-form \sigma on X such that \omega+\sigma+\overline{\sigma} is a rational class and that the restriction of \sigma to the generic fiber of p vanishes. Then K_{X/Y}+L-D(p) is pseudo-effective. Our second main result provides a simple proof of the algebraicity criterion established in [ 37 ] . Ou’s algebraicity criterion provides an important extension to the Kähler case of the results in [ 3 , 2 ] , and more recently [ 11 ] (see also [ 24 ] ). The version we prove here is as follows. Theorem 1.3 . Let X be a compact Kähler manifold, and let C\subset X be a compact submanifold. Let S_{0} be an irreducible, locally closed submanifold containing C_{0} , where C_{0} is a dense open subset of C satisfying {\rm codim}_{C}(C\setminus C_{0})\geq 2 . Let M be the Zariski closure of S_{0} in X . If \dim M>\dim S_{0} , then the conormal bundle \mathcal{N}^{*}_{C_{0}/S_{0}} is pseudo-effective (abbreviated as ”psef”) in the sense of Definition 2.4 . In particular, let {\mathcal{F}}\subset T_{X} be a holomorphic foliation on X . If {\mathcal{F}}^{\star} is not psef, then {\mathcal{F}} is induced by a meromorphic map. Finally, our third objective is to slightly extend W. Ou’s uniruledness criterion [ 37 , Thm 1.1] . This generalization is made possible by Theorems 1.2 and 1.3 . Our proof largely follows the approach in [ 11 ] . In particular, we establish the following results. Theorem 1.4 . Let X be a compact Kähler manifold whose canonical bundle is pseudo-effective. Consider a coherent, torsion-free sheaf {\mathcal{Q}} such that we have \otimes^{m}T_{X}^{\star}\to{\mathcal{Q}}\to 0 for some m\geq 1 . Then \det{\mathcal{Q}} is a pseudo-effective line bundle. Corollary 1.5 . Let X be a compact Kähler manifold. Let {\mathcal{F}}\subset T_{X} be a holomorphic foliation. If \mu_{\alpha,\min}({\mathcal{F}})>0 for some \alpha\in\mathop{\rm Mov}\nolimits(X) , then {\mathcal{F}} is algebraic and its leaves are rationally connected. Next, we provide a few comments on the proofs of these results. In the context of Theorem 1.2 , the map p:X\to Y is locally projective, at least generically over the base Y . The usual techniques—using Bergman kernels and their L^{2/k} versions (cf. [ 1 ] )—allow us to construct metrics on \displaystyle(K_{X/Y}+L)|_{p^{-1}(U)} for any coordinate subset U\subset Y centered at a generic point y\in Y\setminus Z , where Z\subset Y is an analytic subset. However, a priori, it is unclear how to construct these metrics in such a way that they coincide on overlapping subsets; this represents the main challenge here. Indeed, once this issue is resolved, the resulting metric extends across Z (by the familiar estimates of their weights). The key ideas of the proof are contained in [ 30 ] and [ 39 ] . The main part of the proof of Theorem 1.3 closely follows the approach in [ 37 , Thm 1.2] . The main difference is that [ 37 ] requires certain algebraic constructions to carefully desingularize the singular locus of M and to control the loss of the pseudo-effectiveness of the conormal bundle in a highly efficient manner. In our proof, we adopt an analytic approach by comparing the Lelong numbers of a certain current and its pullback (cf. Proposition 2.2 ). This makes the proof purely analytic. Note that the analytic definition of pseudo-effectiveness for sheaves Definition 2.4 is a priori slightly weaker than the algebraic version given in [ 37 , 31 ] . However, thanks to Proposition 6.8 , Theorem 1.3 still implies the main geometric applications, such as [ 37 , Thm 1.1, Thm 1.4] , Theorem 1.4 , and Theorem 5.1 . Let us explain briefly the idea of the proof of Theorem 1.3 , i.e., [ 37 , Thm 1.2] . By using Demailly’s mass concentration method [ 20 ] and Collins-Tosatti’s extension theorem [ 18 ] , we construct an \omega_{X} -psh function \varphi on X such that the Lelong number \nu(\varphi|_{S_{0}},y_{0}) at a fixed point y_{0}\in C_{0} can be arbitrarily large cf. Proposition 4.1 . To the best of our knowledge, in algebraic context this was first introduced by Bost in [ 4 ] . In case the ambient manifold is only assumed to be Kähler, the Monge-Ampère equation is usually used as substitute for the absence of an ample line bundle, see e.g. [ 20 , 22 , 38 ] . It is also the approach followed in [ 37 ] . The main point is to show that as soon as \nu(\varphi|_{S_{0}},y_{0}) is large, so is the generic Lelong number \nu(\varphi|_{S_{0}},C_{0}) of \varphi along C_{0} (hence, no isolated singularities!), see Proposition 4.2 . At least if M is non-singular, this claim is clear: very vaguely, it is a consequence of the fact that for compact manifolds, the Lelong numbers of a closed positive current are bounded in terms of its cohomology class. Let us also mention that given a foliation {\mathcal{F}} by curves on a Kähler manifold, the most complete algebraicity result is due to M. Brunella, cf. [ 9 , 10 ] . He shows e.g. that if the generic leaf of {\mathcal{F}} is uniformized by the unit disk, then the leaf-wise Poincaré metric has a psh variation. This represents a wide generalisation of all known statements in this framework -but again, it only applies to foliations by curves. As we have already mentioned, the proof of Theorem 1.4 and Corollary 1.5 follow from the arguments in [ 11 ] , combined with additional inputs from [ 12 ] . Acknowledgements JC would like to thank Wenhao Ou for sending the preliminary versions of his very nice paper [ 37 ] , as well as for many discussions related to his paper and this note. He thanks the Institut Universitaire de France and the A.N.R JCJC project Karmapolis (ANR-21-CE40-0010) for providing excellent working condition. MP gratefully acknowledge support from the DFG. Special thanks are due to F. Campana: virtually all the

Read the complete text