Kobayashi length bounds on bordered surfaces and generalized integral points on abelian varieties
Paolo Dolce · 2026
Read Kobayashi length bounds on bordered surfaces and generalized integral points on abelian varieties by Paolo Dolce (2026) 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-2603-24193.txt.
Introduction
Dolce sharpens a bound on Kobayashi length for loops on bordered Riemann surfaces. The setup removes finitely many disks from a compact surface and fixes a loop. As more points are removed, the shortest loop in the same homotopy class can only grow. Phung had shown growth between root-s over log s and linear. Dolce improves the upper bound to order root s, which fixes the growth exponent at one half. He applies this to count integral points on abelian varieties, halving a known exponent.
Catalog record
| CREATOR | Paolo Dolce |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62343 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.24193
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Kobayashi length bounds on bordered surfaces and generalized integral points on abelian varieties Paolo Dolce Abstract Let B B be a compact Riemann surface and B 0 ⊂ B B_{0}\subset B a bordered hyperbolic subsurface obtained by removing finitely many disjoint closed disks. Fix a nontrivial loop α \alpha in B 0 B_{0} . For s ≥ 0 s\geq 0 , let L ( α , s ) L(\alpha,s) denote the supremum, over all finite subsets S ⊂ B 0 S\subset B_{0} with # S ≤ s \#S\leq s , of the minimal Kobayashi length of a loop in B 0 ∖ S B_{0}\smallsetminus S that is freely homotopic to α \alpha in B 0 B_{0} . Phung in [ 8 ] proved that L ( α , s ) L(\alpha,s) grows at most linearly and at least as s / log s \sqrt{s}/\log s . We sharpen the upper bound to O ( s log s ) O(\sqrt{s\log s}) , which determines lim s → ∞ log L ( α , s ) log s = 1 2 \lim_{s\to\infty}\frac{\log L(\alpha,s)}{\log s}=\frac{1}{2} , answering a question raised in [ 8 , Question 1.4] . As an application, we improve the counting bound for generalized integral points on abelian varieties over complex function fields: for an abelian variety of dimension n n over ℂ ( B ) \mathbb{C}(B) , Phung proved that the number of ( s , B 0 ) (s,B_{0}) -generalized integral points modulo the constant trace grows at most as s 2 n k s^{2nk} , where k = rk ( π 1 ( B 0 ) ) k=\rk(\pi_{1}(B_{0})) . We sharpen this to s n k + ε s^{nk+\varepsilon} for every ε > 0 \varepsilon>0 , halving the exponent. 1 Introduction 1.1 Generalized integral points and the Lang–Vojta conjecture Let k be an algebraically closed field and let B be a non-singular, irreducible algebraic curve over k with function field K=k(B) . Let X be a smooth, integral, projective K -variety and fix a reduced effective divisor D\subset X . A model of (X,D) is a pair (\mathcal{X},\mathcal{D}) where \mathcal{X} is a normal k -variety equipped with a proper, flat morphism f\colon\mathcal{X}\to B satisfying \mathcal{X}_{K}\cong X , and \mathcal{D}\subset\mathcal{X} is a horizontal divisor with \mathcal{D}_{K}\cong D . Such a model always exists. We also define \Sing(f):=\{b\in B\colon\mathcal{X}_{b}\text{ is singular}\}\,. The set of rational points X(K) corresponds bijectively to the sections of f : for any P\in X(K) , let \sigma_{P}\colon B\to\mathcal{X} be the section obtained as the Zariski closure of P in \mathcal{X} . By abuse of notation we will simply write \sigma_{P}\in X(K) . Given a finite set S\subset B , the set of (S,\mathcal{D}) -integral points of \mathcal{X} is \mathcal{O}_{S,\mathcal{D}}(\mathcal{X}):=\{\sigma_{P}\in X(K)\colon f(\sigma_{P}(B)\cap\mathcal{D})\subseteq S\}\,, that is, the set of sections whose image doesn’t intersect the divisor \mathcal{D} “above” B\smallsetminus S . Recall that the pair (X,D) is said to be of log general type when D is a normal crossings divisor and the log canonical bundle \omega_{X}(D) is big. A special case is when X=A is an abelian variety: the canonical bundle \omega_{A} is trivial, so the log canonical bundle \omega_{A}(D)\cong\mathcal{O}_{A}(D) is big whenever D is big. The geometric Lang–Vojta conjecture predicts strong constraints for the integral points of pairs of log general type Conjecture 1.1 (Geometric Lang–Vojta) . Let (X,D) be a pair of log general type and let (\mathcal{X},\mathcal{D}) be a model. Then there exist a proper closed subset Z\subset X\smallsetminus D and a constant m=m(\mathcal{X},\mathcal{D})>0 with the following property: for every \sigma_{P}\in\mathcal{O}_{S,\mathcal{D}}(\mathcal{X}) with \sigma_{P}(B)\not\subset\mathcal{Z} , where \mathcal{Z} denotes the Zariski closure of Z in \mathcal{X} , we have \deg_{B}\sigma_{P}^{*}\mathcal{D}\;\leq\;m\,(2g(B)-2+\#S)\,. (1) Roughly speaking it says that all integral points, apart from a subset of them having bounded height, can be confined in an exceptional (proper) closed set Z ; moreover the height bound can be expressed in terms of the geometry of the base B . The geometric Lang-Vojta conjecture has been settled in very few cases: when X is a curve; when X=\mathbb{P}^{2} and D has a special shape; when X=A is an abelian variety with trivial K|k trace or when k=\mathbb{C} and A is defined over \mathbb{C} (constant case). For a quick review of the known cases we refer the reader to [ 8 , Section 1] and we remark that for a general complex abelian variety — in particular, when \mathrm{Tr}_{K/\mathbb{C}}(A)\neq 0 and A is nonconstant — the conjecture remains open. From now on we fix the base field k=\mathbb{C} , a model (\mathcal{X},\mathcal{D}) of (X,D) , and a Riemannian metric \rho on B inducing a path-length \ell_{\rho} and a distance function d_{\rho}(\cdot,\cdot) . A (closed) disk in B is a set of the form \overline{\Delta}(x_{0},R):=\{b\in B\colon d_{\rho}(b,x_{0})\leq R\} for R>0 and x_{0}\in B . The notion of (S,\mathcal{D}) -integral point can be generalized by allowing S to vary while keeping only its cardinality s=\#S bounded, and by requiring the intersection condition to hold only on the complement of finitely many disjoint closed disks. This notion of generalized integral points was introduced in [ 8 ] and we recall it below. Definition 1.2 . Let t\in\mathbb{Z}_{\geq 0} . Consider t disjoint disks \overline{\Delta}_{1},\ldots,\overline{\Delta}_{t} in B such that \Sing(f)\subset\bigsqcup_{i=1}^{t}\overline{\Delta}_{i} and distinct points of \Sing(f) are contained in distinct disks. Set B_{0}:=B\smallsetminus\bigsqcup_{i=1}^{t}\overline{\Delta}_{i} , for any s\in\mathbb{Z}_{\geq 0} I(s,B_{0}):=\{\sigma_{P}\in X(K)\colon\#\left(f(\sigma_{P}(B_{0})\cap\mathcal{D})\right)\leq s\} is the set of generalized (s,B_{0}) -integral points of (\mathcal{X},\mathcal{D}) . One immediately observes the inclusion \bigcup_{\begin{subarray}{c}S\subset B\\ \#S\leq s\end{subarray}}\mathcal{O}_{S,\mathcal{D}}(\mathcal{X})\;\subseteq\;I(s,B_{0})\,, so I(s,B_{0}) is potentially much larger than the integral points for any single fixed S . Understanding the growth of \#I(s,B_{0}) as a function of s (with B_{0} fixed) is the main object of this paper. 1.2 Main results Building on Parshin’s foundational work [ 7 ] , Phung in [ 8 ] proved an important quantitative bound on \#I(s,B_{0}) for abelian varieties, working in the following setting. (P) Let A be an abelian variety of dimension n over K=\mathbb{C}(B) and let D\subset A be a reduced effective divisor. We fix a model (\mathcal{A},\mathcal{D}) , where the proper flat morphism is f\colon\mathcal{A}\to B and \sigma_{O}\colon B\to\mathcal{A} is a fixed section. We assume: (i) There exist t\in\mathbb{Z}_{\geq 0} disjoint disks \overline{\Delta}_{1},\ldots,\overline{\Delta}_{t} in B such that \Sing(f)\subset\bigsqcup_{i=1}^{t}\overline{\Delta}_{i} and distinct points of \Sing(f) are contained in distinct disks. We define B_{0}:=B\smallsetminus\bigsqcup_{i=1}^{t}\overline{\Delta}_{i} so that f\colon\mathcal{A}_{B_{0}}\to B_{0} is a family of abelian varieties (A_{b},\sigma_{O}(b))_{b\in B_{0}} . (ii) B_{0} is hyperbolic (this can always be achieved by taking t\geq 3 ). (iii) D\subset A does not contain any translates of nonzero abelian subvarieties. Phung’s main result is the following. Theorem 1.3 (Phung [ 8 , Theorem A] ) . Assume that the setting (P) holds. Then there exists m:=m(B_{0})\in\mathbb{R}_{>0} such that \#\left({\mathchoice{\raisebox{3.75pt}{$\displaystyle{I(s,B_{0})}$}\mkern-5.0mu\diagup\mkern-4.0mu\raisebox{-3.75pt}{$\displaystyle{\Tr_{K/\mathbb{C}}(A)(\mathbb{C})}$}}{\raisebox{3.75pt}{$\textstyle{I(s,B_{0})}$}\mkern-5.0mu\diagup\mkern-4.0mu\raisebox{-3.75pt}{$\textstyle{\Tr_{K/\mathbb{C}}(A)(\mathbb{C})}$}}{\raisebox{2.625pt}{$\scriptstyle{I(s,B_{0})}$}\mkern-5.0mu\diagup\mkern-4.0mu\raisebox{-3.41666pt}{$\scriptstyle{\Tr_{K/\mathbb{C}}(A)(\mathbb{C})}$}}{\raisebox{1.875pt}{$\scriptscriptstyle{I(s,B_{0})}$}\mkern-5.0mu\diagup\mkern-4.0mu\raisebox{-3.41666pt}{$\scriptscriptstyle{\Tr_{K/\mathbb{C}}(A)(\mathbb{C})}$}}}\right)\leq m(s+1)^{2n\rk(\pi_{1}(B_{0},b_{0}))} (2) Note that the bound behaves coherently: if B_{0} shrinks then the set of generalized integral points grows, but the number of punctures of B also increases, making \rk(\pi_{1}(B_{0},b_{0})) larger. The proof of Theorem 1.3 rests on a Parshin cocycle argument. Let k:=\rk(\pi_{1}(B_{0},b_{0})) and fix generators \alpha_{1},\ldots,\alpha_{k} . By Ehresmann’s theorem, the fibration \mathcal{A}_{B_{0}}\to B_{0} gives rise to a short exact sequence 1\to\Gamma\to\pi_{1}(\mathcal{A}_{B_{0}},w_{0})\to G\to 1 , where \Gamma=H_{1}(A_{b_{0}},\mathbb{Z})\cong\mathbb{Z}^{2n} and G=\pi_{1}(B_{0},b_{0})