Tian’s theorem for Moishezon spaces
Abstract.
We prove that the Fubini-Study currents associated to a sequence of singular Hermitian holomorphic line bundles on a compact normal Moishezon space distribute asymptotically as the curvature currents of their metrics.
Key words and phrases:
Bergman kernel function, Fubini-Study current, singular Hermitian metric, normal complex space, Moishezon space2010 Mathematics Subject Classification:
Primary 32L10; Secondary 32A60, 32C20, 32U05, 32U40Contents
1. Introduction
Let be a positive Hermitian holomorphic line bundle on a projective manifold and set . Kodaira’s embedding theorem states that for all sufficiently large, the Kodaira map associated to is an embedding. Hence one can consider the Fubini-Study forms on , , where denotes the Fubini-Study form on a projective space. A celebrated theorem of Tian [T] shows that as , in the topology on (see also [R]). Tian’s theorem follows from the first term asymptotics of the Bergman kernel function associated to the space endowed with the inner product determined by and a volume form on . We refer to the book [MM] for an exposition of these topics as well as for the full asymptotic expansion of the Bergman kernel in different contexts.
In [CM1] we extended Tian’s theorem to the case when is a singular Hermitian holomorphic line bundle with strictly positive curvature current on a compact Kähler manifold , the above convergence now being in the weak sense of currents. Later, we extended Tian’s theorem further to general classes of compact Kähler spaces [CM2, CMM]. In all these situations one has to replace the space with the Bergman space of square integrable holomorphic sections.
In [CMM, Theorem 1.1] we generalized Tian’s theorem by considering sequences , , of singular Hermitian holomorphic line bundles over a compact normal Kähler space , in place of the sequence of powers of a line bundle . Assuming that the curvature currents satisfy a natural growth condition, we proved that the Fubini-Study currents associated to the Bergman spaces (see (1.4)) distribute asymptotically like .
The purpose of this note is to show that the preceding result holds more generally for compact normal spaces which are not assumed to be Kähler. The precise setting is the following:
(A) is a compact, reduced, irreducible, normal complex space of dimension , denotes the set of regular points of , denotes the set of singular points of , and is a Hermitian form on .
(B) , , is a sequence of holomorphic line bundles on with singular Hermitian metrics whose curvature currents verify
(1.1) |
We let and assume that
(1.2) |
Condition (B) implies that are big line bundles, hence is a Moishezon space.
Let , so . We consider currents on in the sense of [D2], and denote by the set of positive closed currents of bidegree on which have local plurisubharmonic (psh) potentials, i.e., holds in a neighborhood of each point of for some psh function . We refer to [CMM, Section 2.1]) for a review of the notions of differential forms, psh functions and currents on complex spaces. We denote by the set of psh functions on an open set . The notions of singular Hermitian metric on a line bundle over a complex space , and its curvature current, are defined as in the case when is smooth (see [D3], [CMM, Section 2.2]).
Let be the Bergman space of -holomorphic sections of relative to the metric and the volume form induced by on ,
(1.3) |
endowed with the obvious inner product. Let be the Bergman kernel function and the Fubini-Study current of the space . They are defined as follows. Let be an orthonormal basis of . If let be a holomorphic frame of on a neighborhood of and write with . Then
(1.4) |
We have that are independent of the choice of basis. Moreover, , where is the (meromorphic) Kodaira map associated to the Bergman space .
Our main result is the following theorem:
Theorem 1.1.
Assume that , satisfy conditions (A)-(B). Then the following hold:
(i) as , in .
(ii) as , in the weak sense of currents on .
Note that a complex space that verifies (A)-(B) is a Moishezon space. Thus Theorem 1.1 applies to any compact normal Moishezon space , which is not necessarily assumed to be Kähler. Indeed, a singular Hermitian holomorphic line bundle over with strictly positive curvature current as in (1.1) is big, hence is Moishezon (see, e.g., [CMM, Proposition 2.3], [BCMN, Propositions 3.2 and 3.3]). We recall that a (reduced) compact irreducible complex space of dimension is called a Moishezon space if there exist algebraically independent meromorphic functions on (see [U, Definition 3.5], [BCMN, Section 3]). We refer to [BCMN, Section 3] and the references therein for the definition and some basic properties of big line bundles over complex spaces.
Theorem 1.1 is proved in Section 2. An important special case is provided by the sequence of powers of a singular Hermitian holomorphic line bundle with strictly positive curvature current. See Theorem 3.1 in Section 3, which gives a full generalization of Tian’s theorem to the singular setting. We recall in Section 3 a few other important applications of Theorem 1.1, in particular to the asymptotic distribution of the zeros of random sequences of holomorphic sections (see Theorem 3.3).
2. Proof of Theorem 1.1
By a theorem of Moishezon ([Mo], [U, Theorem 3.6]), is bimeromorphically equivalent to a projective manifold. More precisely, since is assumed to be normal we have that and the following holds (see [BCMN, Theorem 3.1]):
Theorem 2.1.
If is a compact, irreducible, normal Moishezon space then there exists a connected projective manifold and a surjective holomorphic map , given as a composition of finitely many blow-ups with smooth center, such that is a biholomorphism, where is an analytic subset of , , , and is a normal crossings divisor.
Let verify assumptions (A)-(B) and be as in Theorem 2.1. In [CMM] we assumed that is a Kähler normal space, and we showed that the desingularization obtained by finitely many blow-ups with smooth centers as in [BM, GM] is Kähler. This is crucial for construction peak sections by using methods involving . In our present situation we obtain a projective desingularization since is Moishezon.
We will follow the arguments from the proof of [CMM, Theorem 1.1], working with instead of the desingularization of given in [CMM, Section 2.3], and using a Kähler form on the projective manifold . We recall the following lemmas that will be needed in the proof.
Lemma 2.2 ([CMM, Lemma 2.1]).
If
the map is an isometry and the Bergman kernel function of is .
Lemma 2.3 ([CMM, Lemma 2.2, Lemma 3.2]).
There exist , , a Hermitian form on , and singular Hermitian metrics on such that , and as , and on . Moreover, for every relatively compact open subset of there exists a constant such that on .
The Hermitian form is obtained as , where is a suitable metric on and is an appropriate constant. If is a sequence such that , , and if is a weight of on , one defines the metric on and shows that it has the desired properties. In particular the positivity of is needed to solve a -equation on , by using the following version of Demailly’s estimates for the -operator [D1, Théorème 5.1] (see also [CMM, Theorem 2.5]):
Theorem 2.4.
Let , , be a complete Kähler manifold and be a Kähler form on (not necessarily complete) such that its Ricci form for some constant . Let be singular Hermitian holomorphic line bundles on such that , where . If verifies and then there exists such that and .
Proof of Theorem 1.1.
Step 1. We prove that as , in . Fix , a contractible Stein coordinate neighborhood of , such that the (closed) ball , and set . Note that the currents have uniformly bounded mass. By [DS, Proposition A.16] (see also [DNS]) and [Ho, Theorem 3.2.12], we infer that there exist psh functions on such that and the sequence is relatively compact in . Since is holomorphically trivial, we can find holomorphic frames for such that are the corresponding psh weights of , so .
Let be a Kähler form on and be the Hermitian form from Lemma 2.3. Then there exists constants such that
(2.1) |
With as in Lemma 2.3, we prove that there exist and such that
(2.2) |
holds for all , , and with . The upper bound in (2.2) follows from the subaverage inequality, exactly as the upper bound from [CM1, (7)].
We show next that there exist and with the following property: if and is such that , then there exists with and
(2.3) |
This yields the lower bound in (2.2), since . To this end we work first on using the metric from Lemma 2.3. By (2.1),
We have that has a complete Kähler metric [D1, O], and on for some . Using ideas from [D4, Proposition 3.1], [D5, Section 9], we apply the Ohsawa-Takegoshi extension theorem [OT] and Theorem 2.4 as in the proof of [CM1, Theorem 5.1] to show that there exist , , such that if and , , then there is verifying and
By Lemma 2.3 and (2.1) we obtain
(2.4) |
where is so that on . As is a biholomorphism, we let and be the section of induced by . Since is normal and , extends to a holomorphic section on and (2.3) follows from (2.4).
Recall that is relatively compact in , hence it is locally uniformly upper bounded in . It follows from (2.2) that there is a constant such that
(2.5) |
Moreover, if a subsequence in and a.e. on , where is psh on , we infer from (2.2) and the Hartogs lemma [Ho, Theorem 3.2.13] that
holds for a.e. and every . Thus a.e. on , and hence in by (2.5) and the generalized Lebesgue dominated convergence theorem. We conclude that as in .
Step 2. We finish the proof of by showing that there exists a compact set such that and in . Let be the Bergman spaces from Lemma 2.2. It follows by (1.2) that there exists such that
(2.6) |
Let . By (2.6), we can proceed as in Step 1 to find an open neighborhood of and holomorphic frames of with corresponding psh weights of , such that the sequence is relatively compact in . Let be an orthonormal basis of and , with . By Lemma 2.2,
We claim that in . Indeed, assume that a subsequence converges in to a psh function on . By Step 1, in , hence in . It follows that is locally uniformly upper bounded in and in . This proves our claim.
Since is compact, we infer by the above that there exists a compact set such that and in . Then
as , and the proof is finished. ∎
3. Applications
In the case of the sequence of powers of a single line bundle, Theorem 1.1 yields the following generalization of Tian’s theorem to the setting of big line bundles on Moishezon spaces:
Theorem 3.1.
Let be a compact, reduced, irreducible, normal complex space of dimension and be a singular Hermitian holomorphic line bundle on such that , where is a constant and is a Hermitian form on . If are the Bergman kernel function and Fubini-Study current of then, as ,
Proof.
If , then the assumptions (A)-(B) hold with
∎
Recall that a Kähler current is a positive closed current of bidegree on such that for some constant . Let be a singular Hermitian holomorphic line bundle on with positive curvature current , and such that has a singular Hermitian metric whose curvature is a Kähler current. As in [CMM, Corollary 5.2], Theorem 1.1 can be applied to the sequence of line bundles , where and , as . One can also apply Theorem 1.1 to the sequence of tensor products of powers of several line bundles as in [CMM, Corollary 5.11]. We refer to [CMM, Section 5] for the details.
Let us consider now the special case when is smooth, i.e., a connected compact complex manifold of dimension . If is assumed to be Kähler then the domination condition (1.2) is not needed as one can work directly on without the use of a modification . More precisely, in [CMM] we proved the following:
Theorem 3.2.
[CMM, Theorem 1.2] Let be a compact Kähler manifold of dimension and , , be a sequence of singular Hermitian holomorphic line bundles on which satisfy , where and . If are the Bergman kernel function and Fubini-Study current of , and if , then in and weakly on .
However, if is a Moishezon manifold which is not Kähler, and hence not projective, we still have to use in our proof of Theorem 1.1 the modification provided in Theorem 2.1. So we have to require the domination condition (1.2) in assumption (B).
One of the main applications of Tian’s theorem is to the study of the asymptotic distribution of the zeros of random sequences of sections in as , where is a holomorphic line bundle over a compact complex manifold . This started with the pioneering work of Shiffman and Zelditch [SZ1] in the case of a positive line bundle over a projective manifold (see also [SZ2, Sh]). It is shown in [SZ1] that for almost all sequences one has that weakly on , where denotes the current of integration over the zero divisor of a holomorphic section . In the case of singular Hermitian holomorphic line bundles we proved that similar results hold in different contexts [CM1, CM2, CMN1, CMN2].
The study of the asymptotic distribution of zeros of random sections in the Bergman spaces for an arbitrary sequence of singular Hermitian holomorphic line bundles over a compact normal Kähler space was pursued in [CMM, BCM]. In particular we considered in [BCM, Theorem 1.1] very general probability measures on the spaces , as follows. We identify the spaces using fixed orthonormal bases and we endow them with probability measures such that the following holds:
(C) There exist a constant and for every constants such that
Note that [BCM, Theorem 1.1] holds in our present context. Indeed, we can apply the general equidistribution result [BCM, Theorem 4.1] together with Theorem 1.1. We recall one of its assertions here.
Theorem 3.3.
Assume that verify (A), (B), (C) and consider the product probability space
If then for -a.e. sequence we have, as ,
We refer to [BCM, BCHM] for general classes of measures that satisfy condition (C), including Gaussians, Fubini-Study volumes, and area measure of spheres. Note that if the measures verify (C) with constants independent of (like the Gaussians and the Fubini-Study volumes) then the hypothesis of Theorem 3.3 becomes .
We close the paper with some remarks of Moishezon manifolds. By a theorem of Moishezon, a Moishezon manifold is projective if and only if it carries a Kähler metric, see [Mo] and [MM, Theorem 2.2.26]. Moreover, any Moishezon manifold of dimension two is projective, by Theorem 2.1. Indeed, in dimension two we can blow up only points and the blow-up at a point of a compact manifold is projective if and only is projective. Hence non-projective Moishezon manifolds have dimension greater than two. The first example of this kind was obtained by Hironaka in his thesis (1961) and is described in [Ha, Appendix B, Example 3.4.1]. It’s a manifold which contains a curve which is homologous to zero, which is impossible on a Kähler manifold. Further examples can be found in [A, BV, Ko, Pe], see also [MM, Section 2.3.4].
References
- [A] M. Andreatta, Moishezon manifolds, Math. Z. 230 (1999), no. 4, 713–726.
- [BCHM] T. Bayraktar, D. Coman, H. Herrmann and G. Marinescu, A survey on zeros of random holomorphic sections, Dolomites Res. Notes Approx. 11 (2018), Special Issue Norm Levenberg, 1–19.
- [BCM] T. Bayraktar, D. Coman and G. Marinescu, Universality results for zeros of random holomorphic sections, Trans. Amer. Math. Soc. 373 (2020), no. 6, 3765–3791.
- [BCMN] T. Bayraktar, D. Coman, G. Marinescu and V.-A. Nguyên, Zeros of random holomorphic sections of big line bundles with continuous metrics, preprint, 2024.
- [BV] L. Bonavero and C. Voisin, Schémas de Fano et variétés de Moishezon, C. R. Acad. Sci. Paris Sér. I Math. 323 (1996), no. 9, 1019–1024.
- [BM] E. Bierstone and P. Milman, Canonical desingularization in characteristic zero by blowing up the maximum strata of a local invariant, Invent. Math. 128 (1997), 207–302.
- [CM1] D. Coman and G. Marinescu, Equidistribution results for singular metrics on line bundles, Ann. Sci. École Norm. Supér. (4) 48 (2015), 497–536.
- [CM2] D. Coman and G. Marinescu, Convergence of Fubini-Study currents for orbifold line bundles, Internat. J. Math. 24 (2013), 1350051, 27 pp.
- [CMM] D. Coman, X. Ma and G. Marinescu, Equidistribution for sequences of line bundles on normal Kähler spaces, Geom. Topol. 21 (2017), 923–962.
- [CMN1] D. Coman, G. Marinescu and V.-A. Nguyên, Hölder singular metrics on big line bundles and equidistribution, Int. Math. Res. Notices 2016, no. 16, 5048–5075.
- [CMN2] D. Coman, G. Marinescu and V.-A. Nguyên, Approximation and equidistribution results for pseudo-effective line bundles, J. Math. Pures Appl. (9) 115 (2018), 218–236.
- [D1] J.-P. Demailly, Estimations pour l’opérateur d’un fibré holomorphe semipositif au–dessus d’une variété kählérienne complète, Ann. Sci. École Norm. Sup. (4) 15 (1982), no. 3, 457–511.
- [D2] J.-P. Demailly, Mesures de Monge-Ampère et caractérisation géométrique des variétés algébriques affines, Mém. Soc. Math. France (N.S.) No. 19 (1985), 1–125.
- [D3] J.-P. Demailly, Singular Hermitian metrics on positive line bundles, in Complex algebraic varieties (Bayreuth, 1990), Lecture Notes in Math. 1507, Springer, Berlin, 1992, 87–104.
- [D4] J.-P. Demailly, Regularization of closed positive currents and intersection theory, J. Algebraic Geom. 1 (1992), 361–409.
- [D5] J.-P. Demailly, A numerical criterion for very ample line bundles, J. Differential Geom. 37 (1993), 323–374.
- [DS] T.-C. Dinh and N. Sibony, Dynamics in several complex variables: endomorphisms of projective spaces and polynomial-like map**s, Holomorphic dynamical systems, 165–294, Lecture Notes in Math. 1998, Springer, Berlin, 2010.
- [DNS] T.-C. Dinh, V.-A. Nguyên and N. Sibony, Dynamics of horizontal-like maps in higher dimension, Adv. Math. 219 (2008), 1689–1721.
- [GM] C. Grant Melles and P. Milman, Classical Poincaré metric pulled back off singularities using a Chow-type theorem and desingularization, Ann. Fac. Sci. Toulouse Math. (6) 15 (2006), 689–771.
- [Ha] R. Hartshorne, Algebraic geometry, Grad. Texts in Math., No. 52, Springer-Verlag, New York-Heidelberg, 1977. xvi+496 pp.
- [Ho] L. Hörmander, Notions of Convexity, Reprint of the 1994 edition, Modern Birkhäuser Classics, Basel, Birkhäuser, 2007, viii, 414 pp.
- [Ko] J. Kollár, Flips, flops, minimal models, etc., Surveys in differential geometry (Cambridge, MA, 1990), 113–199, Lehigh Univ., Bethlehem, PA, 1991.
- [MM] X. Ma and G. Marinescu, Holomorphic Morse Inequalities and Bergman Kernels, Progress in Math., vol. 254, Birkhäuser, Basel, 2007, xiii, 422 pp.
- [Mo] B. G. Moishezon, On -dimensional compact complex manifolds having algebraically independent meromorphic functions. I, II, III, Izv. Akad. Nauk SSSR Ser. Mat. 30 (1966), 133–174, 345–386, 621–656. English translation: American Mathematical Society Translations Ser. 2, 63, 1967, 51–177.
- [O] T. Ohsawa, Hodge spectral sequence and symmetry on compact Kähler spaces, Publ. Res. Inst. Math. Sci. 23 (1987), 613–625.
- [OT] T. Ohsawa and K. Takegoshi, On the extension of holomorphic functions, Math. Z. 195 (1987), 197–204.
- [Pe] T. Peternell, Moishezon manifolds and rigidity theorems, Bayreuth. Math. Schr. (1998), no. 54, 1–108.
- [R] W. D. Ruan, Canonical coordinates and Bergman metrics, Comm. Anal. Geom. 6 (1998), 589–631.
- [Sh] B. Shiffman, Convergence of random zeros on complex manifolds, Sci. China Ser. A 51 (2008), 707–720.
- [SZ1] B. Shiffman and S. Zelditch, Distribution of zeros of random and quantum chaotic sections of positive line bundles, Comm. Math. Phys. 200 (1999), 661–683.
- [SZ2] B. Shiffman and S. Zelditch, Number variance of random zeros on complex manifolds, Geom. Funct. Anal. 18 (2008), 1422–1475.
- [T] G. Tian, On a set of polarized Kähler metrics on algebraic manifolds, J. Differential Geom. 32 (1990), 99–130.
- [U] K. Ueno, Classification theory of algebraic varieties and compact complex spaces, Lecture Notes in Math., Vol. 439, Springer-Verlag, Berlin-New York, 1975, xix+278 pp.