New unlikely intersections on elliptic surfaces

Douglas Ulmer, José Felipe Voloch · 2025

Read New unlikely intersections on elliptic surfaces by Douglas Ulmer, José Felipe Voloch (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-2508-06680.txt.

Introduction

Douglas Ulmer and Jose Felipe Voloch study rare meets on curved surfaces in a 2025 math paper. Their stage is a surface swept by elliptic curves, with a marked path that runs without end. Past work by Ulmer and a partner counted how often that path just grazes a twist-fixed path, a rare meet by any count.

Such counts were first done with tools that fail where the base field has prime cycles. The pair extend the counts to those fields, and add a fresh route for the classic case. Their new meets tie to a famed map of Manin, a deep bridge in the field.

Catalog record

CREATORDouglas Ulmer, José Felipe Voloch
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-06680

Browse the full library · About the Tower of Babel

Full text

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

New unlikely intersections on elliptic surfaces Douglas Ulmer, José Felipe Voloch Abstract . Consider a Jacobian elliptic surface ℰ → 𝒞 \mathcal{E}\to\mathcal{C} with a section P P of infinite order. Previous work of the first author and Urzúa over the complex numbers gives a bound on the number of tangencies between P P and a torsion section of ℰ \mathcal{E} (an “unlikely intersection”), and more precisely, an exact formula for the weighted number of tangencies between P P and elements of the “Betti foliation”. This work used analytic techniques that apparently do not generalize to positive characteristic. In this paper, we extend their work to characteristic p p , and we develop a second approach to tangency properties of algebraic curves on a complex elliptic surface, yielding a new family of unlikely intersections with a strong connection to a famous homomorphism of Manin. We also correct inaccuracies in the literature about this homomorphism. 1991 Mathematics Subject Classification Primary 14G27, 11G05; Secondary 11G25, 14D10, 14G17 1. Introduction 1.1. Motivation We study unlikely intersections in the form of tangencies of curves on elliptic surfaces. In [ 8 ] , the first author and Urzúa worked over the complex numbers and gave a bound on the number of tangencies between a section of infinite order and n -torsion (multi-)sections for varying n . Such tangencies may be viewed as “unlikely intersections.” More generally, they introduced a “Betti” foliation (considered independently by several other authors) of a dense open subset of the elliptic surface by multisections generalizing the n -torsion multisections, and they showed that there is an exact formula for the weighted number of tangencies between the given section of infinite order and the leaves of the Betti foliation. These leaves are generally of infinite degree over the base, so non-algebraic, and the bound on tangencies uses a certain non-holomorphic (but real-analytic) 1-form. It is thus not evident whether or how these bounds might be generalized to base fields of positive characteristic. In this paper, we extend their work to positive characteristic. Our results are somewhat similar but there are some differences as well. In doing so, we noticed that there is a second way of looking at tangency properties of algebraic curves on a complex elliptic surface, which we develop. Our results in characteristic p are somehow intermediate between the results of [ 8 ] and this second approach to tangencies in characteristic zero. For our purposes, an elliptic surface is a smooth projective algebraic surface \mathcal{E}/k , where k is an algebraically closed field, equipped with an elliptic fibration \pi:\mathcal{E}\to\mathcal{C} , where \mathcal{C}/k is a smooth projective algebraic curve, admitting a zero section O . The generic fiber of \pi is thus an elliptic curve E/K , where K=k(\mathcal{C}) is the function field of \mathcal{C} . We take a point P\in E(K) of infinite order and also denote by P the corresponding section of \pi . Our work depends crucially on a homomorphism of groups \mu:E(K)\to K which was introduced in characteristic zero by Manin [ 3 ] and generalized to positive characteristic by the second author [ 9 ] . (This map was also studied cohomologically in [ 7 ] and made very explicit in [ 1 ] .) The value \mu(P) is expressed as a rational function of the coordinates of P and their derivatives. Our results on tangencies with P are obtained by a careful study of the local and global properties of \mu . We sharpen the local analysis of the values of \mu done by previous authors and relate them with the tangency properties were are interested in. Interpreting the values of \mu as sections of an appropriate bundle on \mathcal{C} leads to a bound on the number of tangencies. To give the flavor of our results, we state one result in characteristic p and one over \mathbb{C} . First, let k be algebraically closed of characteristic p>3 , let N>3 be an integer prime to p , and let \mathcal{C}=\Ig_{1}(N) be the (Igusa) curve parameterizing elliptic curves with a point of order N and an Igusa structure of order p . Let K=k(\mathcal{C}) , let E/K be the universal curve, and let P\in E(K) be a point of infinite order. We define a local intersection invariant I(P,t) for t\in\mathcal{C} which measures the order of contact between P and local sections which are divisible by p . (See Definition 3.4 for details.) We consider the set T (“tangencies”) of points t\in\mathcal{C} where I(P,t)>1 , and we divide it into subsets T_{o} and T_{s} where the elliptic curve corresponding to t is ordinary or supersingular respectively. Our result is then a bound \displaystyle|T|\leq|T_{o}|+p|T_{s}| \displaystyle\leq p(2g-2-d)+(p-1)\delta, \displaystyle=p(p-1)\sigma-\delta, where g is the genus of \mathcal{C} , d is the degree of a certain bundle \omega , \delta is the number of points where E has bad reduction, and \sigma is the number of points where E has supersingular reduction. In general, the division of T into two parts is dictated by the zeroes of a certain differential \lambda which measures the failure of Kodaira-Spencer to be an isomorphism. Now assume that k=\mathbb{C} . We define “complex Betti leaves” which are parameterized by two complex numbers modulo 1 and are such that on a dense open subset of \mathcal{E} , there is a complex Betti leaf through a given point with a given tangent direction. Thus, an unlikely intersection is when P meets some complex Betti leaf to order at least 3. We define an exceptional set S\subset\mathcal{C} (essentially where \mathcal{E} has bad reduction or the period map is not an immersion) and a set of tangencies T_{\mathbb{C}}\subset\mathcal{C}\setminus S where P has order of contact at least 3 with a complex Betti leaf. We then find an upper bound: |T_{\mathbb{C}}|\leq 4g-4-d+|S|. In both characteristic zero and p , our arguments have a local aspect (relating intersection numbers to orders of vanishing of a Manin map) and a global aspect (interpreting the image of the Manin map as a section of a suitable line bundle and deducing a bound from degree considerations). In both the complex and characteristic p cases, our results are somewhat looser than those of [ 8 ] due to subtleties in the local analysis. 1.2. Acknowledgements The first author thanks L’Institut des Hautes Études Scientifiques and La Fondation Sciences Mathématiques de Paris for support and hospitality during a sabbatical year, the Simons Foundation for support via Collaboration Grant 713699, and the participants in the IHES seminar “Équations différentielles motiviques et au–delà” for questions which stimulated this work. The second author thanks the Marsden Fund administered by the Royal Society of New Zealand for financial support. 2. p -descent and tangencies Let k be a field of characteristic p (algebraically closed and with p>3 for convenience). Let \mathcal{C} be a smooth, irreducible, projective curve over k , and let K=k(\mathcal{C}) . Let E/K be an elliptic curve with j -invariant j\not\in K^{p} , and let P\in E(K) be a point of infinite order. Let \mathcal{E}\to\mathcal{C} be the Néron model of E/K (as a group scheme) with zero section O . Let \omega be the line bundle O^{*}\left(\Omega^{1}_{\mathcal{E}/\mathcal{C}}\right) . It is well known that if y^{2}=x^{3}+a_{4}x+a_{6}\qquad a_{4},a_{6}\in K (2.1)

Read the complete text