Effective termination of general type MMPs in dimension at most five

Jingjun Han, Jihao Liu, Ziquan Zhuang · 2025

Read Effective termination of general type MMPs in dimension at most five by Jingjun Han, Jihao Liu, Ziquan Zhuang (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-2509-01501.txt.

Introduction

Jingjun Han, Jihao Liu, and Ziquan Zhuang prove the effective termination of minimal model programs of general type in dimension at most five. The minimal model program is a stepwise method in algebraic geometry for simplifying shapes. For dimension at most four, they also give explicit bounds, in terms of topological invariants and divisor volumes, on how many steps the program can take.

Catalog record

CREATORJingjun Han, Jihao Liu, Ziquan Zhuang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE83 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-01501

Browse the full library · About the Tower of Babel

Full text

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

Effective termination of general type MMPs in dimension at most five Jingjun Han, Jihao Liu, Ziquan Zhuang Abstract . We prove the effective termination of general type MMPs in dimension at most 5 5 , and give explicit bounds (in terms of topological invariants and volume of divisors) on the number of steps in the MMP when the dimension is at most 4 4 . 2020 Mathematics Subject Classification 14E30, 14J30, 14J35. Throughout this paper, we work over \mathbb{C} , the field of complex numbers. 1. Introduction The termination of the minimal model program (MMP) is one of the major open problems in birational geometry. The termination is known in full generality in dimension three by [ 28 , 22 , 19 , 29 ] . Several special cases are also known in dimension four [ 18 , 10 , 1 , 3 , 31 , 4 , 14 , 7 , 27 ] , and in higher dimensions we have the termination of MMP with scaling (a special kind of MMP) in the general type or big boundary case [ 2 ] . Beyond these, the termination conjecture remains extremely difficult. In particular, it is an open question whether every fivefold general type MMP terminates. Our first result in this paper is the effective termination of general type MMPs, or more generally, of pseudo-effective MMP with a big boundary, in dimension five. Theorem 1.1 (see Theorem 3.9 ) . Let (X,\Delta) be a projective klt pair of dimension 5 . Assume that at least one of the following holds. (1) K_{X}+\Delta is big. (2) K_{X}+\Delta is pseudo-effective and \Delta is big. Then there exists a positive integer m depending only on (X,\Delta) such that every (K_{X}+\Delta) -MMP terminates after at most m steps. We also give some explicit bounds on the number of steps in the MMP in dimensions 3 and 4 . A sample result is the following. Here h_{4}^{\mathrm{alg}}(X) is the dimension of the subspace in the singular homology group H_{4}(X,\mathbb{R}) generated by algebraic cycles. Theorem 1.2 (see Theorems 4.15 ) . Let N be a positive integer and (X,\Delta) a projective log smooth klt pair of dimension 4 such that N\Delta has integer coefficients. Assume that (1) (X,\Delta) has at most N strata, each stratum has Picard number at most N , and h_{4}^{\mathrm{alg}}(X)\leq N , and (2) \Delta is big, \mathrm{vol}(\Delta)\geq\frac{1}{N} , and (\Delta\cdot H^{3})\leq N for some very ample divisor H on X . Then any (K_{X}+\Delta) -MMP terminates after at most M^{M} steps, where M:=(2N)^{9}! . By passing to log resolutions, Theorem 1.2 gives explicit termination results for klt MMP with a big boundary (see Lemma 2.6 ). We remark that even the termination of such fourfold MMPs was not previously known. We also prove similar explicit termination results for general type fourfold MMPs (Theorem 4.14 ), terminal fourfold MMPs (Corollary 4.9 ) and arbitrary threefold MMPs (Theorem 4.16 ). For some past works on the explicit or effective termination problem for threefolds, see for example [ 8 , 26 , 6 , 20 ] . Sketch of proof. A key ingredient in the proof of Theorems 1.1 and 1.2 is some (explicit) uniform bounds on the Cartier index of Weil divisors in a general type MMP [ 16 , Theorem 3.2, Corollaries 3.18 and 3.20] . By Birkar’s inductive approach [ 3 ] , if termination holds in dimension n-1 , then it also holds for pairs of nonnegative Kodaira dimension in dimension n . We refine Birkar’s argument to show that if termination holds for MMPs with discrete discrepancies in dimension n-1 , then we also have termination for MMPs with uniform index bound in dimension n . By the termination of terminal fourfold flips, it is not hard to see that MMP with discrete discrepancies always terminates in dimension 4 , and this gives the termination of general type MMPs in Theorem 1.1 . Once we have termination, the effective bound on the number of steps essentially comes from Kőnig’s lemma. The proof of Theorem 1.2 divides into two main steps: we first find explicit bounds on the number of steps in terminal MMPs, and then use the index bound from [ 16 ] to reduce to the terminal case. By a variation of difficulty function type arguments for termination (see e.g. [ 28 , 18 , 22 , 19 , 29 , 10 , 1 ] ), one may write down a sequence of invariants involving the Picard numbers of boundary divisors, the (weighted) difficulties of the pair, and certain algebraic Betti numbers, that is lexicographically decreasing in any terminal MMPs in dimensions 3 and 4 . To turn this into an explicit bound on the number of steps in the terminal MMP, we show in Section 4.2 that the above lexicographic sequence of invariants decreases in a bounded way, in the sense that each time an individual term drops by one, the terms thereafter increases by at most a uniformly fixed amount. To deal with the general klt case, we further introduce a weighted combination of the above invariants, where the weights ultimately depends on the index bound from [ 16 ] , and combine with the results in the terminal case to show that the weighted combination decreases by (at least) a fixed amount after each step of the MMPs we consider. This gives the explicit bound in Theorem 1.2 . Outline of the article. In Section 2 we summarize some preliminary results that will be used in the later part of this paper. In Section 3 we prove Theorem 1.1 . In Section 4 we prove Theorem 1.2 and its variants. Acknowledgment The authors would like to thank Jungkai A. Chen, Christopher D. Hacon, James M c Kernan, Lu Qi, Vyacheslav Shokurov, Chenyang Xu for helpful discussions and comments. JH is supported by NSFC for Excellent Young Scientists (#12322102), and the National Key R&D Program of China (#2023YFA1010600). JH is a member of LMNS, Fudan University. JL is supported by the National Key R&D Program of China (#2024YFA1014400). ZZ is partially supported by the NSF Grants DMS-2240926, DMS-2234736, a Sloan research fellowship, a Packard fellowship, and the Simons Collaboration on Moduli of Varieties. 2. Preliminary In this section, we collect some preliminary materials that will be used throughout the article. 2.1. Notation and conventions In this paper, varieties are assumed to be quasi-projective unless otherwise specified. We follow the standard terminology from [ 21 , 24 ] . Given two \mathbb{R} -divisors \Delta and \Delta^{\prime} on a variety X , we write \Delta^{\prime}\geq\Delta if \mathrm{mult}_{D}\Delta^{\prime}\geq\mathrm{mult}_{D}\Delta for any prime divisor D on X . A sub-pair (X,\Delta) consists of a normal variety X together with an \mathbb{R} -divisor \Delta on X (a priori, we do not require that K_{X}+\Delta is \mathbb{R} -Cartier). It is called a pair if \Delta\geq 0 . We denote by \mathrm{Coeff}(\Delta) the (finite) set of coefficients of \Delta (in particular, we require that 0\in\mathrm{Coeff}(\Delta) ). A log (sub-)pair is a (sub-)pair (X,\Delta) such that K_{X}+\Delta is \mathbb{R} -Cartier. A sub-pair (X,\Delta) is called log smooth if X is smooth and \mathrm{Supp}(\Delta) is a simple normal crossing divisor. A stratum (resp. open stratum )

Read the complete text