Singular intersections on families of abelian varieties

Nicola Ottolini · 2025

Read Singular intersections on families of abelian varieties by Nicola Ottolini (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-2506-15344.txt.

Introduction

Some intersections in geometry are so unlikely that mathematicians expect them to be rare, and proving that rarity is a whole research program. Nicola Ottolini studies a family of abelian varieties built over a curve, with a second curve sitting inside the family. He asks how often that inner curve can meet the family's subgroup schemes tangentially, touching rather than crossing. His result: only finitely many such tangential meeting points exist.

The heart case, elliptic curves, already follows from earlier work by Corvaja, Demeio, Masser, and Zannier. Ottolini gives a different proof of that case using the Pila-Zannier method, which counts rational points to bound special ones. He notes that this route could one day yield an effective version, with explicit bounds. The result sits inside the theory of unlikely intersections and varies a known conjecture for abelian varieties.

Catalog record

CREATORNicola Ottolini
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE81 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-15344

Browse the full library · About the Tower of Babel

Full text

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

Singular intersections on families of abelian varieties Nicola Ottolini Abstract . Let S S be a smooth irreducible curve defined over ℚ ¯ \overline{\mathbb{Q}} , let 𝒜 \mathcal{A} be an abelian scheme over S S and 𝒞 \mathcal{C} a curve inside 𝒜 \mathcal{A} , both defined over ℚ ¯ \overline{\mathbb{Q}} . In this paper we prove that the set of points in which 𝒞 \mathcal{C} intersects proper flat subgroup schemes of 𝒜 \mathcal{A} tangentially is finite. This fits in the framework of the so-called problems of unlikely intersections, and can be seen as a variation of the relative Pink conjecture for abelian varieties. 1991 Mathematics Subject Classification 11G10, 11G05, 11G50, 14C17 1. Introduction Let \lambda:\mathcal{L}\to\mathbb{C}\setminus\{0,1\} be the Legendre elliptic scheme, where for every \lambda_{0}\in\mathbb{C}\setminus\{0,1\} the fiber \mathcal{L}_{\lambda_{0}} is an elliptic curve defined by (1.1) Y^{2}=X(X-1)(X-\lambda_{0}). Masser and Zannier [ MZ08 , MZ10 ] showed that there are at most finitely many complex numbers \lambda_{0}\neq 0,1 such that the two points \displaystyle(2,\sqrt{2(2-\lambda_{0})}),\quad \displaystyle(3,\sqrt{6(3-\lambda_{0})}) both have finite order on the elliptic curve \mathcal{L}_{\lambda_{0}} . Later, the same authors proved in [ MZ12 ] that one can replace 2 and 3 by any two distinct complex numbers (\neq 0,1) or even choose distinct X -coordinates (\neq\lambda) defined over an algebraic closure of \mathbb{C}(\lambda) , provided that the corresponding points P_{1},P_{2} are not identically linearly related, i.e. there are no integers a_{1},a_{2}\in\mathbb{Z} not both zero such that a_{1}P_{1}+a_{2}P_{2}=O identically on \mathcal{C} . This was further generalized by Barroero and Capuano [ BC16 ] , who proved that, taking an arbitrary number of points defined over \overline{\mathbb{Q}(\lambda)} such that they are not identically linearly related, the set of complex numbers \lambda_{0} such that they satisfy two independent linear condition with integer coefficient is again finite. The same authors further generalized this result to product of powers of two non-isogenous elliptic schemes [ BC17 ] and then to general abelian schemes [ BC20 ] , also generalizing a result in [ MZ20 ] , where the authors consider only the intersection with the torsion; the main result is the following. Theorem 1.1 ( [ BC20 ] , Thm 1.1) . Let S be a smooth irreducible curve defined over \overline{\mathbb{Q}} . Consider \mathcal{A}\to S an abelian scheme and \mathcal{C} an irreducible curve in \mathcal{A} not contained in a proper subgroup scheme of \mathcal{A} , even after a finite base change. Let \mathcal{C}^{\{2\}}=\bigcup_{\overset{G\text{ flat}}{\text{codim }G\geq 2}}\mathcal{C}\cap G, where G runs through all the flat subgroup schemes of codimension at least 2, that is, all the subgroup schemes of codimension at least 2 whose irreducible components dominate the base S ; then, \mathcal{C}^{\{2\}} is a finite set. All these results fit in the framework of very general conjectures, formulated independently by Zilber [ Zil02 ] and Bombieri, Masser and Zannier [ BMZ99 ] in the toric case and by Pink [ Pin05 ] in the more general setting of mixed Shimura varieties, including classical conjectures such as Manin-Mumford, André-Oort, known as the Zilber–Pink conjectures. Shimura varieties are a class of algebraic varieties including abelian varieties, families of abelian varieties, modular curves and moduli spaces of abelian varieties. Their structure allows us to define a certain (countable) subset of their subvarieties, called “special subvarieties”. A general description is quite complicated in general, and can be found in [ Del71 ] . In the case of tori and abelian varieties, the special subvarieties are precisely the irreducible components of algebraic subgroups. For families of abelian varieties, the situation is more complicated. Flat subgroup schemes, that is, the subgroup schemes that are dominant on the base, are special subvarieties, and actually they are the only ones that are dominant on the base. However, there are in general also special subvarieties that are not dominant. For example, in the case of non-isotrivial elliptic schemes, they are the subgroups of the CM fibers. Zilber-Pink conjectures involve the intersection of a fixed subvariety Z of a Shimura variety \mathcal{X} with the special subvarieties of \mathcal{X} of codimension bigger than the dimension of Z . The prediction is that, under some natural geometric conditions, such intersection is not Zariski dense in Z . The precise statement for curves in an abelian scheme is the following. Conjecture 1.2 . Let S be a smooth irreducible curve defined over \mathbb{C} . Consider \mathcal{A}\to S an abelian scheme and \mathcal{C} an irreducible curve in \mathcal{A} , defined over \mathbb{C} and not contained in a proper special subvariety of \mathcal{A} . Consider the set \mathcal{C}^{[2]}=\bigcup_{\overset{H\subseteq\mathcal{A}}{\text{codim }H\geq 2}}\mathcal{C}\cap H, where H runs through all the special subvarieties of \mathcal{A} of codimension at least 2. Then \mathcal{C}^{[2]} is a finite set. This conjecture is known to hold when \mathcal{A} is a fibered power of an elliptic scheme \mathcal{E} , by work of Masser and Zannier [ MZ12 ] , Barroero and Capuano [ BC16 ] , Barroero [ Bar17 ] and Barroero and Dill [ BD25 ] , and in some cases of product of (fibered) powers of two non-isogenous elliptic schemes defined over \overline{\mathbb{Q}} , due to Masser and Zannier [ MZ14 ] , Barroero and Capuano [ BC17 ] and Ferrigno [ Fer24 ] . The general case, however, is still open. More details on the Zilber-Pink conjectures on families of abelian varieties may be found in a survey by Capuano [ Cap23 ] , and a more in-depth treatment of the topic can be found in books by Zannier [ Zan12 ] and Pila [ Pil22 ] . The geometric idea behind these results is that the intersections we are considering are, for dimensional reasons, expected to be empty. When this is not the case, we say that we have an “unlikely intersection”. The statement of the conjecture can be interpreted as saying that this is indeed a rare phenomenon. On the other hand, if in the same setting as above one intersects the curve \mathcal{C} with codimension 1 subgroups, then the union of such intersections will be infinite in general. However, if we restrict ourselves to consider only the points of singular intersection, \mathcal{C}\cap_{sing}G , that is the points of \mathcal{C}\cap G where the multiplicity of intersection between \mathcal{C} and a codimension 1 subgroup G is strictly greater then 1, then we may hope to gain back finiteness. This is because the intersection of a curve with another subvariety of codimension 1 is in general expected to be transverse, that is, not singular. Geometrically, this corresponds to considering the points where for some a_{1},\dots,a_{n} in the generic endomorphism ring, the section a_{1}P_{1}+\dots+a_{n}P_{n} intersects the zero section tangentially . This kind of question has already been investigated, and answered positively, in some cases. More specifically, the case of curves which are images of sections intersecting tangentially the torsion sections in a single copy of the Legendre elliptic scheme, was treated independently by works of Corvaja, Demejo, Masser and Zannier [ CDMZ21 ] and Ulmer and Urzúa [ UU20 , UU21 ] . In a different setting, namely for curves in \mathbb{G}_{m}^{2} , the question was studied by Marché and Maurin [ MM23 ] . In this paper we extend such result to curves in more general one dimensional families \mathcal{A}\to S of abelian varieties. To obtain a result

Read the complete text