Remarks on log pluricanonical representations

Osamu Fujino and Jinsong Xu · 2025

Read Remarks on log pluricanonical representations by Osamu Fujino and Jinsong Xu (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-2501-16568.txt.

Introduction

Osamu Fujino and Jinsong Xu prove a finiteness result in algebraic geometry in a 2025 maths paper. The objects are log pluricanonical representations, which act on spaces of pluricanonical forms. Such actions arise in the study of log pairs and their models. The team assume a good minimal model exists for the setting. Under that assumption, the representations are shown to be finite. The short note settles the question on those terms.

Catalog record

CREATOROsamu Fujino and Jinsong Xu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE15 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-16568

Browse the full library · About the Tower of Babel

Full text

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

Remarks on log pluricanonical representations Osamu Fujino and Jinsong Xu Abstract . We show the finiteness of log pluricanonical representations under the assumption of the existence of a good minimal model. 2020 Mathematics Subject Classification Primary 14E30; Secondary 14E07 1. Introduction This short paper is a supplement to [ FG ] (see also [ HX ] ). One of the main purposes of this paper is to establish: Theorem 1.1 . Let (X,\Delta) be a projective log canonical pair such that K_{X}+\Delta is \mathbb{Q} -Cartier. Assume that (X,\Delta) has a good minimal model. Then there exists a positive integer k such that the image of \rho_{km}\colon\Bir(X,\Delta)\to\Aut_{\mathbb{C}}H^{0}(X,\mathcal{O}_{X}(km(K_{X}+\Delta))) is a finite group for every positive integer m , where \Bir(X,\Delta):=\{\sigma\mid\text{$\sigma\colon(X,\Delta)\dashrightarrow(X,\Delta)$ is $B$-birational}\}. We make five important remarks on Theorem 1.1 . Remark 1.2 . In Theorem 1.1 , k is the smallest positive integer such that k(K_{X}+\Delta) and k(K_{X^{\prime}}+\Delta^{\prime}) are both Cartier, where (X^{\prime},\Delta^{\prime}) is a good minimal model of (X,\Delta) . Remark 1.3 . If K_{X}+\Delta is semiample, that is, (X,\Delta) itself is a good minimal model of (X,\Delta) , then Theorem 1.1 is nothing but [ FG , Theorem 1.1] (see also Remark 1.2 ). In this paper, we will show that we can reduce Theorem 1.1 to [ FG , Theorem 1.1] . We note that [ FG , Theorem 1.1] plays a crucial role in the study of the abundance conjecture for semi-log canonical pairs. Remark 1.4 (Log canonical pairs of log general type) . If (X,\Delta) is a projective log canonical pair such that K_{X}+\Delta is big, then \Bir(X,\Delta) is already a finite group (see [ FG , Corollary 3.13] ). Therefore, Theorem 1.1 is obvious when K_{X}+\Delta is big. Remark 1.5 (Kawamata log terminal pairs) . We have already established some better results for kawamata log terminal pairs. For the details, see [ FG , §3.1] , where we do not need the minimal model program. Hence, in this paper, we are mainly interested in a log canonical pair (X,\Delta) which is not kawamata log terminal. For smooth varieties, see also [ U , §14] . Remark 1.6 . In a recent preprint [ F5 ] , the first author established the finiteness of relative log pluricanonical representations in a suitable complex analytic setting. For the details, see [ F5 ] . Similarly to Theorem 1.1 , we can prove the following theorem. Theorem 1.7 . Let V be a smooth quasi-projective variety and let X be a smooth projective completion of V such that \Delta:=X\setminus V is a simple normal crossing divisor on X . Assume that (X,\Delta) has a good minimal model. Then there exists a positive integer k such that the image of \rho_{km}\colon\PBir(V)\to\Aut_{\mathbb{C}}H^{0}(X,\mathcal{O}_{X}(km(K_{X}+\Delta))) is a finite group for every positive integer m , where \PBir(V):=\{\sigma\mid\text{$\sigma\colon V\dashrightarrow V$ is proper birational}\}. Note that Theorem 1.7 implies Theorem 1.8 . Theorem 1.8 (Affine varieties) . Let V be a smooth affine variety and let X be a smooth projective completion of V such that \Delta:=X\setminus V is a simple normal crossing divisor on X . Then there exists a positive integer k such that the image of \rho_{km}\colon\Aut(V)\to\Aut_{\mathbb{C}}H^{0}(X,\mathcal{O}_{X}(km(K_{X}+\Delta))) is a finite group for every positive integer m , where \Aut(V):=\{\sigma\mid\text{$\sigma\colon V\to V$ is an isomorphism}\}. To the best knowledge of the authors, Theorem 1.8 is nontrivial and new. Acknowledgments . The first author was partially supported by JSPS KAKENHI Grant Numbers JP20H00111, JP21H04994, JP21H04994, JP23K20787. The authors would like to thank Professors Hideo Kojima and Takashi Kishimoto for comments. Throughout this paper, we will work over \mathbb{C} , the field of complex numbers. We will freely use the standard notation as in [ F2 ] and [ F3 ] . 2. Preliminaries In this section, we recall some basic definitions and properties necessary for this paper. For the standard notation of the theory of minimal models, see, for example, [ F2 ] , [ F3 ] , and so on. Let us start with the definition of B -birational maps, which was first introduced in [ F1 ] . Definition 2.1 ( B -birational maps) . Let (X,\Delta) be a projective log canonical pair. We say that a birational map \sigma\colon X\dashrightarrow X is B -birational if there exists a commutative diagram \textstyle{Z\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \scriptstyle{\beta} \textstyle{X\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\sigma} \textstyle{X} such that Z is a normal projective variety, \alpha and \beta are projective birational morphisms, and \alpha^{*}(K_{X}+\Delta)=\beta^{*}(K_{X}+\Delta) holds. We put \Bir(X,\Delta):=\{\sigma\mid\text{$\sigma\colon(X,\Delta)\dashrightarrow(X,\Delta)$ is $B$-birational}\}. Then \Bir(X,\Delta) has a natural group structure. We take a positive integer m such that m(K_{X}+\Delta) is Cartier. Then it is easy to see that we have a group homomorphism \rho_{m}\colon\Bir(X,\Delta)\to\Aut_{\mathbb{C}}H^{0}(X,\mathcal{O}_{X}(m(K_{X}+\Delta))). We call it the B -pluricanonical representation or log pluricanonical representation for the pair (X,\Delta) . For the details of \Bir(X,\Delta) , see [ FG ] . Let us recall the notion of proper birational maps (see, for example, [ I ] ). Definition 2.2 (Proper birational maps) . Let V be a quasi-projective variety. We say that a birational map \sigma\colon V\dashrightarrow V is proper birational if p_{1} and p_{2} are projective, where \Gamma is the graph of \sigma\colon V\dashrightarrow V and p_{1} and p_{2} are projections from \Gamma to V as in the following commutative diagram (see [ I , Proposition 2.17 (i)] ). \textstyle{\Gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{p_{2}} \scriptstyle{p_{1}} \textstyle{V\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\sigma} \textstyle{V} We put \PBir(V):=\{\sigma\mid\text{$\sigma\colon V\dashrightarrow V$ is proper birational}\}. Then it is easy to see that \PBir(V) has a natural group structure (see [ I , Proposition 2.17 (iii)] ). We note that if V is affine and normal then \sigma is an isomorphism for every \sigma\in\PBir(V) by Zariski’s main theorem (see [ I , Theorem 2.19 and Corollary] ). We further assume that V is smooth and X is a smooth projective completion of V such that \Delta:=X\setminus V is a simple normal crossing divisor on X . Let \Gamma^{\prime}\to\Gamma be a projective resolution of singularities and let Y be a projective completion of \Gamma^{\prime} such that \Delta_{Y}:=Y\setminus\Gamma^{\prime} is a simple normal crossing divisor on Y . Then we have the following commutative diagram. \textstyle{\Gamma^{\prime}\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{p^{\prime}_{2}} \scriptstyle{p^{\prime}_{1}} \textstyle{\Gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{p_{2}} \scriptstyle{p_{1}} \textstyle{V\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\sigma} \textstyle{V} Let m be any positive integer. In this case, for every \sigma\in\PBir(V) , we can define \sigma^{*}\in\Aut_{\mathbb{C}}H^{0}(X,\mathcal{O}_{X}(m(K_{X}+\Delta))) as follows: \begin{split}\sigma^{*}:=(\overline{p}^{\prime*}_{1})^{-1}\circ\overline{p}^{\prime*}_{2}\colon H^{0}(X,\mathcal{O}_{X}(m(K_{X}+\Delta)))&\to H^{0}(Y,\mathcal{O}_{Y}(m(K_{Y}+\Delta_{Y})))\\ &\to H^{0}(X,\mathcal{O}_{X}(m(K_{X}+\Delta))),\end{split} where \overline{p}^{\prime}_{i}\colon Y\to X is the extension of p^{\prime}_{i}\colon\Gamma^{\prime}\to V for i=1,2 . We can easily see that it is independent of the choice of (Y,\Delta_{Y}) (see [ I , Theorem 11.1] ). Then we can consider the following group homomorphism \rho_{m}\colon\PBir(V)\to\Aut_{\mathbb{C}}H^{0}(X,\mathcal{O}_{X}(m(K_{X}+\Delta))) by setting \rho_{m}(\sigma):=\sigma^{*} for every positive integer m . For the sake of completeness, we recall the definition of good minimal models in the sense of Birkar–Shokurov. Definition 2.3 (Good minimal models) . Let (X,\Delta) be a projective log canonical pair. Let \psi\colon X\dashrightarrow X^{\prime} be a birational map and let E be the reduced \psi^{-1} -exceptional divisor on X^{\prime} , that is, E=\sum_{j}E_{j} where E_{j} are the \psi^{-1} -exceptional prime divisors on X^{\prime} . Then (X^{\prime},\Delta^{\prime}) is called a good minimal model of (X,\Delta) (in the sense of Birkar–Shokurov) if (1) (X^{\prime},\Delta^{\prime}) is a projective \mathbb{Q} -factorial divisorial log terminal pair, where \Delta^{\prime}:=\psi_{*}\Delta+E , (2) K_{X^{\prime}}+\Delta^{\prime} is semiample, and (3) a(D,X,\Delta)<a(D,X^{\prime},\Delta^{\prime}) holds for every \psi -exceptional prime divisor D on

Read the complete text