Miyaoka–Yau equality and uniformization of log Fano pairs
Louis Dailly · 2025
Read Miyaoka–Yau equality and uniformization of log Fano pairs by Louis Dailly (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-2501-05887.txt.
Introduction
Louis Dailly proves a result in geometry in a twenty twenty-five paper. He works with log Fano pairs that have standard coefficients, a set class of shapes in algebraic geometry. The paper deals with the Miyaoka-Yau inequality, a known bound for such objects. Dailly shows what must hold when a pair meets that bound exactly. In that case, he proves, the orbifold universal cover of the pair is a projective space. The result ties an extreme numeric case to a very rigid shape.
Catalog record
| CREATOR | Louis Dailly |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 96 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-05887
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Miyaoka–Yau equality and uniformization of log Fano pairs Louis Dailly Abstract . Let ( X , Δ ) (X,\Delta) be a log Fano pair with standard coefficients. We show that if it satisfies the equality case in Miyaoka–Yau inequality, then its orbifold universal cover is a projective space. 1. Introduction At the beginning of the 20th century, Poincaré and Koebe stated that the projective line \mathbb{P}^{1}(\mathbb{C}) is the unique Riemann surface which is compact connected, simply connected. In higher dimensions, several characterizations of complex projective spaces exist. We can think about Mori’s characterization [ Mor79 ] : projective spaces are the only compact complex manifolds which have an ample tangent bundle, or Siu–Yau’s one [ SY18 ] : projective spaces are the only manifolds which can be endowed with a Kähler metric whose bisectional holomorphic curvature is positive. In another direction, we can give a characterization of projective spaces coming from uniformization theory. Let us describe the universal cover of a compact complex manifold X of dimension n , if it satisfies the following two geometric assumptions: • (KE_{\lambda}) : the manifold X can be endowed with a Kähler–Einstein metric \omega : \mathrm{Ric}\,\omega=\lambda\omega , with \lambda\in\mathbb{R} . • (MY) : the manifold X satisfies the so-called Miyaoka–Yau equality: \left(2(n+1)c_{2}(X)-nc_{1}(X)^{2}\right)\cdot\left[\omega\right]^{n-2}=0. Under these assumptions, the universal cover \widetilde{X} of X is biholomorphic to: • the complex projective space \mathbb{P}^{n} if \lambda>0 , • the complex affine space \mathbb{C}^{n} if \lambda=0 , • the unit ball \mathbb{B}^{n}=\left\{z\in\mathbb{C}^{n}\,\,|\,\,|z|<1\right\} if \lambda<0 . In the case \lambda>0 , we get that the manifold X is biholomorphic to \mathbb{P}^{n} , as X is a Fano manifold which is always simply connected. Hence, it yields the desired characterization of \mathbb{P}^{n} . Now, if we consider a non-trivial finite group G , it can be realized as a subgroup of \mathrm{PGL}\,(n+1,\mathbb{C}) for some integer n , and we can consider the orbit space \mathbb{P}^{n}/G . The group G does not act freely, and the quotient map \mathbb{P}^{n}\longrightarrow\mathbb{P}^{n}/G is ramified. The space \mathbb{P}^{n}/G is a normal projective variety with quotient singularities endowed with a branching divisor \Delta_{G} . We get an orbifold pair (\mathbb{P}^{n}/G,\Delta_{G}) that is uniformized by \mathbb{P}^{n} . A natural question is: can we adapt the previous characterization of quotients to the pair framework ? More precisely, if a pair (X,\Delta) satisfies singular versions of (KE_{\lambda}) and (MY) , can we describe its universal cover \left(\widetilde{X}_{\Delta},\widetilde{\Delta}\right) ? In the case \lambda\leq 0 , it has been shown [ CGG24 , Thms. A and B] that such characterizations hold. In the case \lambda>0 , the uniformization result holds when \Delta=0 as a consequence of [ GKP22 , Thm. 1.5] and [ DGP24 , Thm. B] . The aim of the present paper is to investigate the case \lambda>0 with a boundary divisor \Delta . On the one hand, the authors of [ GKP22 ] show the uniformization result by replacing the assumption (KE_{>0}) with the condition of semistability of the canonical extension, which is an extension of the tangent sheaf \mathcal{T}_{X} by \mathcal{O}_{X} . On the other hand, the authors of [ DGP24 ] show that the existence of a singular Kähler–Einstein metric implies the above mentioned semistability. The aim of our article is to show the following theorem, which is analogous to [ GKP22 , Thm. 1.3] with a boundary divisor \Delta . Main Theorem . Let (X,\Delta) be a log Fano pair of dimension n . There exists a finite group G\subseteq\mathrm{PGL}\,(n+1,\mathbb{C}) , such that (X,\Delta)\simeq(\mathbb{P}^{n}/G,\Delta_{G}) if and only if the following two conditions hold: (i) The canonical extension \mathcal{E}_{(X,\Delta)} is semistable with respect to -(K_{X}+\Delta) , (ii) The pair (X,\Delta) satisfies the Miyaoka–Yau equality: (2(n+1)c_{2}(X,\Delta)-nc_{1}(X,\Delta))\cdot c_{1}(X,\Delta)^{n-2}=0. Remarks on the theorem: It is a priori unclear what object \mathcal{E}_{(X,\Delta)} is and what semistability means in assumption (i) . In the following, the assumption (i) is: There exists a strictly \Delta -adapted morphism f_{0}\colon Y_{0}\longrightarrow X such that \mathcal{E}_{X,\Delta,f_{0}} is semistable with respect to f_{0}^{*}(-(K_{X}+\Delta)) . This assumption is related to the notion of adapted canonical extension \mathcal{E}_{X,\Delta,f} for a strictly \Delta -adapted morphism f\colon Y\longrightarrow X . This notion is introduced in the third section. It is a reflexive sheaf on Y which is an extension of the adapted tangent sheaf \mathcal{T}_{X,\Delta,f} by \mathcal{O}_{Y} . This is a notion similar to the adapted differentials \Omega^{[1]}_{X,\Delta,f} introduced by Miyaoka and Campana (see [ CKT16 , Sect. I.3] ). The second assumption (ii) is a singular version of (MY) that uses the definition of orbifold Chern classes initiated by Mumford. Here, we use an equivalent definition using orbifold de Rham cohomology, see [ CGG24 , Sect. 3] . The question of uniformization in [ GKP22 ] , [ DGP24 ] , [ CGG24 ] has been considered in the more general setting of klt pairs. A part of the proof of our theorem is devoted to show that a pair satisfying assumptions (i) and (ii) has only quotient singularities. Strategy of the proof: The strategy of the proof is based on [ GKP22 , Sect. 5.2] . In this article, the authors introduce a notion of canonical extension \mathcal{E}_{X} for a klt space X . Then, they show that the Miyaoka–Yau equality implies projective flatness of \mathcal{E}_{X} on the regular locus of the variety, hence flatness of \mathrm{End}\,(\mathcal{E}_{X}) . Hence, they introduce the "maximally quasi-étale" cover \pi\colon Y\longrightarrow X [ GKP16 , Thm. 1.14] that plays the role of a global uniformization of X . The goal is to show that Y=\mathbb{P}^{n} . As previously, we have the flatness of the endomorphism bundle associated \mathrm{End}\,(\mathcal{E}_{Y}) over Y_{\mathrm{reg}}\, . Because Y is maximally quasi-étale, \mathrm{End}\,(\mathcal{E}_{Y}) extends to a locally free sheaf on Y . As a consequence of the Zariski–Lipman theorem for klt spaces, Y is smooth as \mathcal{T}_{Y} is locally a direct summand of \mathrm{End}\,(\mathcal{E}_{Y}) . It is simply connected because it is a Fano manifold. Hence, there exists a line bundle L on Y such that \mathcal{E}_{Y}=L^{\oplus n+1} , \mathcal{E}_{Y} being projectively flat on a simply connected manifold. Then, we have c_{1}(Y)=c_{1}(\mathcal{E}_{Y})=(n+1)c_{1}(L) and we conclude by Kobayashi–Ochiai’s characterization of \mathbb{P}^{n} . In the pair setting, two main difficulties arise from the presence of a boundary divisor \Delta : • The universal cover (\widetilde{X}_{\Delta},\widetilde{\Delta}) of an orbifold could still have singularities in codimension 1 given by the boundary divisor \widetilde{\Delta} . Under the assumptions of the theorem, a Zariski–Lipman type argument is needed to show that the pair is smooth i.e \widetilde{\Delta}=0 and \widetilde{X}_{\Delta} is smooth. In the proof of [ CGG24 ] for the case \lambda<0 , this is obtained by arguments coming from uniformization of Riemannian orbifolds, using the constancy of the curvature. In the case \lambda=0 in the same article, the authors use the existence of an orbi-étale map via the existence of a section of some large power of K_{X}+\Delta . Recently, the authors of [ GP24 ] study system of Hodge-bundles in the pair setting to study quotient of bounded symmetric domains. In our article, we state in the first section a criterion related to the notion of orbi-locally freeness of the tangent bundle. Then, we construct the canonical extension \mathcal{E}_{(X,\Delta)} of an orbifold (X,\Delta) . It is an extension of \mathcal{T}_{(X,\Delta)} by \mathcal{O}_{X}