Explicit surjectivity of Galois representations of products of elliptic curves over function fields
Alina Cojocaru, Frederick Saia · 2025
Read Explicit surjectivity of Galois representations of products of elliptic curves over function fields by Alina Cojocaru, Frederick Saia (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-2510-20910.txt.
Introduction
Alina Cojocaru and one co-author prove an effective surjectivity result for Galois representations attached to products of elliptic curves over function fields of characteristic zero. Such results bound how large the image of the representation can be, with explicit control rather than mere existence.
The proof routes through an isogeny degree bound in this setting, assembled from curve bounds by Griffon and Pazuki and from a function field analogue of the Frey Mazur conjecture. Techniques trace back to Serre and to Masser and Wüstholz in the number field case. The curves treated are non isotrivial and pairwise non isogenous, the generic and most useful case.
Catalog record
| CREATOR | Alina Cojocaru, Frederick Saia |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-20910
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Explicit surjectivity of Galois representations of products of elliptic curves over function fields Alina Cojocaru, Frederick Saia Abstract . We prove an explicit surjectivity result for products of non-isotrivial, non-isogenous elliptic curves over a function field of arbitrary characteristic. This is by way of an isogeny degree bound in this setting, generated from bounds for elliptic curves by Griffon–Pazuki, and techniques originated by Serre and Masser–Wüstholz in the number field setting. We apply our result to prove that most members of a family of products of elliptic curves over ℚ \mathbb{Q} with no extra endomorphisms have no exceptional primes above a specified constant which depends neither on the elliptic curve factors nor on the dimension of the product. 2020 Mathematics Subject Classification Primary 11G10, Secondary 11G05 1. Introduction Let E be an elliptic curve over a field K , and let N be a positive integer. The absolute Galois group G_{K}:=\Gal(K^{\textnormal{sep}}/K) acts naturally on the torsion points of E over a separable closure K^{\textnormal{sep}} of K , and in particular acts on the subgroup E[N] of points of order dividing N . This action is captured in the mod- N Galois representation \rho_{E,N}:G_{K}\rightarrow\GL(E[N])\simeq\GL_{2}(\mathbb{Z}/N\mathbb{Z}). By a celebrated result of Serre [ Ser72 , Théorème 2] , if K is a number field and E does not have complex multiplication over K^{\textnormal{sep}} (is non-CM), then there exists a constant \ell_{0} depending on E such that \rho_{E,\ell} is surjective for all primes \ell\geq\ell_{0} . Serre also proved an analogous result [ Ser72 , Théorème 6] for a product of two elliptic curves, stating that the mod- \ell Galois representation associated to a product of two non-CM, non-(geometrically)-isogenous elliptic curves over K is as large as possible for all \ell sufficiently large with respect to the two elliptic curve factors. For K the function field of a smooth, projective, and geometrically integral curve over a perfect field of constants k , we have a theorem of Igusa [ Igu59 , Theorem 3] whose proof preceded Serre’s theorem. (We also refer the reader to [ BLV09 ] for an alternate proof of this result.) Igusa’s theorem implies that if E is a non-isotrivial elliptic curve over K , then there exists a constant \ell_{0} depending on E such that for all \ell\geq\ell_{0} the representation \rho_{E,\ell} is as large as possible in the sense that \text{Image}(\rho_{E,\ell})\cap\SL_{2}(\mathbb{Z}/N\mathbb{Z})=\SL_{2}(\mathbb{Z}/N\mathbb{Z}). Naturally, one asks for effective versions of the above results: what explicit lower bounds on \ell guarantee that the image is as large as possible in either setting? Following Serre’s work, an effective theorem for the product of an arbitrary number of elliptic curves over a number field was proven by Masser–Wüstholz [ MW93a , Theorem, Proposition 1] as an application of their work on isogeny degree bounds in [ MW93b ] . They prove that surjectivity holds for all primes \ell larger than a constant depending on the Weil heights of the elliptic curve factors and the degree of the number field. Since their work, the subject of effective surjectivity results over number fields has seen ongoing interest, with many working to improve the bounds of Masser–Wüstholz or to extend these types of results to other classes of abelian varieties (see, e.g., [ Coj05 , LV14 , Lom15 , LF16 , Lom16b , Lom16a , Zyw22 , MW24b , MW24a ] ). An explicit, effective surjectivity result for elliptic curves over a function field, which we recall as Theorem 4.1 , was proven by Cojocaru–Hall in [ CH05 ] . Their theorem states that surjectivity holds for all primes \ell>c(g) not equal to the characteristic of K , where c(g) is an explicit constant depending only on the genus of K . (The precise definition of c(g) appears both in Theorem 1.1 and in Theorem 4.1 .) Our main result in this work is an explicit surjectivity result for the product of an arbitrary number of elliptic curves over a function field, which provides the first extension of both Igusa’s theorem and Cojocaru–Hall’s theorem to products of dimension larger than 1 . For C a smooth, projective, and geometrically integral curve over a perfect field k , let K=k(C) denote the function field of C and let g=g(K)=g(C) denote the genus of K . For E_{1},\ldots,E_{n} elliptic curves over K , A=E_{1}\times\cdots\times E_{n} the product abelian variety of dimension n , and \ell a prime, let \displaystyle\rho_{A,\ell}:G_{K} \displaystyle\rightarrow\GL(E_{1}[\ell]\times\cdots\times E_{n}[\ell])\simeq\GL_{2}(\mathbb{Z}/\ell\mathbb{Z})^{n} be the mod- \ell Galois representation of A . Equivalently, this is the product of the mod- \ell Galois representations of the factors E_{i} . The determinants of the projections of \rho_{A,\ell} to each component must agree, each being equal to the \ell^{\textnormal{th}} cyclotomic character of K , and so the image of \rho_{A,\ell} is contained in the determinant locus \Delta_{\ell}:=\left\{(M_{1},M_{2},\ldots,M_{n})\mid\det{M_{1}}=\det{M_{2}}=\ldots=\det{M_{n}}\right\}\leq\GL_{2}(\mathbb{Z}/\ell\mathbb{Z})^{n}. Let \Psi_{\ell}:=\textnormal{Image}(\rho_{A,\ell})\cap\SL_{2}(\mathbb{Z}/\ell\mathbb{Z})^{n} . Our effective result for products is then as follows (see Theorem 4.3 ). Theorem 1.1 . Let K be the function field of a smooth, projective, and geometrically integral curve C over a perfect field k . Let E_{1},\ldots,E_{n} be pairwise non-isogenous elliptic curves over K with j(E_{i})\not\in\overline{k} for each 1\leq i\leq n , and let A=E_{1}\times\cdots\times E_{n}. Set \displaystyle e_{2}(q):=\begin{cases}1\quad&\text{ if }q\equiv 1\pmod{4}\\ -1\quad&\text{ otherwise,}\end{cases}\quad\quad e_{3}(q):=\begin{cases}1\quad&\text{ if }q\equiv 1\pmod{3}\\ -1\quad&\text{ otherwise,}\end{cases}\quad \displaystyle c(g) \displaystyle:=2+\max\left\{\ell\mid\dfrac{q-(6+3e_{2}(q)+4e_{3}(q))}{12}\leq g\right\}, \displaystyle C(g) \displaystyle=3176523\cdot\max\{1,g(K)^{3/2}\}, and finally \widetilde{C}(g):=\begin{cases}c(g)\quad\text{ if }n=1\\ C(g)\quad\text{ if }n\geq 2.\end{cases} For all primes \ell not equal to \textnormal{char}(K) satisfying \ell>\widetilde{C}(g) , we have \Psi_{\ell}=\SL_{2}(\mathbb{Z}/\ell\mathbb{Z})^{n} . A pleasing aspect of Theorem 1.1 , like the dimension 1 result Theorem 4.1 , is that it depends only on the genus g of K , and not on the individual elliptic curves factors nor on the dimension of A when n\geq 2 . The constant \widetilde{C} appearing in the above statement is defined in cases, singling out the dimension 1 case, so as to recover the theorem of Cojocaru–Hall when n=1 , given that our proof strategy only produces a bound in dimension n\geq 2 (see Theorem 4.3 ). We also prove a version ( Theorem 4.5 ) of this result which depends on the modular heights of the factors, which can provide a better bound than C(g) for individual product abelian varieties of dimension n\geq 2 . We apply Theorem 1.1 to prove a result on exceptional primes of specializations of families of products of elliptic curves over the rationals. Our setup is as follows: Let n\geq 2 , let A_{1}(t),B_{1}(t),\ldots,A_{n}(t), and B_{n}(t)\in\mathbb{Z}[t] and consider the elliptic curves over \mathbb{Q}(t) given by E_{i}:y^{2}=x^{3}+A_{i}(t)x+B_{i}(t)\qquad\textnormal{ for i }\in\{1,\ldots,n\}. Assume that the E_{i} are pairwise non-isogenous over \overline{\mathbb{Q}(t)} and that j(E_{i})\not\in\overline{\mathbb{Q}} for each i . Let A=E_{1}\times\cdots\times E_{n}, and let E_{i,t_{0}} and A_{t_{0}} denote, respectively, the specialization of E_{i} and of A to a given t_{0}\in\mathbb{Q} . Let S\subseteq\mathbb{Q} denote the set of t_{0}\in\mathbb{Q} such that at least one of the following items holds: • A has bad reduction at t_{0} , • E_{i,t_{0}} has CM for some i\in\{1,\ldots,n\} , or • E_{i,t_{0}} and E_{j,t_{0}} are isogenous over \overline{\mathbb{Q}} for some i\neq j . For T>0 a real number, let \mathcal{F}(T):=\left\{\frac{m}{n}\in\mathbb{Q}\setminus S\mid m,n\in\mathbb{Z},\gcd(m,n)=1,\textnormal{ and }\max\{|m|,|n|\}\leq T\right\}, and for \ell a prime number let \mathcal{E}_{\ell}(T):=\{t_{0}\in\mathcal{F}(T)\mid\Gal(\mathbb{Q}(A_{t_{0}}[\ell])/\mathbb{Q})\subsetneq\Delta_{\ell}\}. Our result, stating that most specializations of A have no exceptional primes at least 3176533 , is as follows