Adelic Mordell-Lang and the Brauer-Manin Obstruction
Brendan Creutz · 2025
Read Adelic Mordell-Lang and the Brauer-Manin Obstruction by Brendan Creutz (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-2510-25931.txt.
Introduction
Brendan Creutz, working alone, proves a result about rational points on varieties inside abelian varieties over function fields. An abelian variety is a higher dimensional analogue of an elliptic curve, and a rational point is a solution with coordinates in the base field. He assumes the variety has no positive dimensional isotrivial quotient, a technical condition that rules out constant parts after extension.
Creutz shows that any adelic point on the subvariety that is a limit of rational points on the ambient variety is already a limit of rational points on the subvariety itself. If the Tate Shafarevich group is finite, rational points are then dense in the Brauer set, a set cut out by a known obstruction. Similar results follow over totally imaginary number fields under a stated conjecture.
Catalog record
| CREATOR | Brendan Creutz |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-25931
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Adelic Mordell-Lang and the Brauer-Manin Obstruction Brendan Creutz Abstract . Let X X be a closed subvariety of an abelian variety A A over a global function field k k such that the base change of A A to an algebraic closure does not have any positive dimensional isotrivial quotient. We prove that every adelic point on X X which is the limit of a sequence of k k -rational points on A A is a limit of k k -rational points on X X . Assuming finiteness of the Tate-Shafarevich group of A A , this implies that the rational points on X X are dense in the Brauer set of X X . Similar results are obtained over totally imaginary number fields, conditionally on an adelic Mordell-Lang conjecture. 1. Introduction Let A be an abelian variety over a global field k . By the celebrated Mordell-Weil theorem the set A(k) of rational points on A is a finitely generated abelian group. Moreover, the standard arithmetic duality theorems describe, at least conjecturally, how A(k) sits inside the set A({\mathbb{A}}_{k}) of adelic points. The topological closure \overline{A(k)} of the image of A(k) in the space A({\mathbb{A}}_{k})_{\bullet} of connected components of A({\mathbb{A}}_{k}) is isomorphic to the profinite completion of A(k) [ SerreII , MilneCongruence ] . It is conjectured that the Tate-Shafarevich group {\mbox{{Sh}}}(A) of A is finite and, in particular, that its maximal divisible subgroup is trivial. If {\mbox{{Sh}}}(A)_{\textup{div}} is trivial, then \overline{A(k)} is equal to the set A({\mathbb{A}}_{k})^{\Br}_{\bullet} of connected components of A({\mathbb{A}}_{k}) which are orthogonal to the Brauer group of A [ Manin , Wang , PoonenVoloch ] . We prove the converse of this last statement in Theorem 4.2 below. For X a closed subvariety of A , the Mordell-Lang conjecture describes X(k) as the set of rational points on a finite union of special subvarieties. This was proved by Faltings [ Faltings ] in the number field case and by Hrushovski [ Hrushovski ] in the function field case (See also [ Voloch , AV , RosslerML ] ). However, it is less well understood how X(k) sits inside X({\mathbb{A}}_{k}) . The set X({\mathbb{A}}_{k})^{\Br} of adelic points orthogonal to the Brauer group is closely related to the intersection of X({\mathbb{A}}_{k}) with the closure of the image of A(k) in A({\mathbb{A}}_{k}) . In the function field case it is conjectured that the sets \overline{X(k)} , X({\mathbb{A}}_{k})\cap\overline{A(k)} and X({\mathbb{A}}_{k})^{\Br} are all equal [ PoonenVoloch , Conjecture C] . A similar statement with modifications at the archimedean primes is conjectured in the number field case. These conjectures originate from questions posed independently by Scharaschkin [ Scharaschkin ] and Skorobogatov [ Skorobogatov , p. 133] for curves over number fields. There is extensive numerical and theoretical evidence for these conjectures, though predominantly in cases where X(k) is finite [ Flynn , Poonen , Stoll , BruinStoll , PoonenVoloch , BGW , CreutzWA ] . The main results of this paper establish these conjectures over global function fields, assuming a nonisotriviality hypothesis on the base change A_{\overline{k}} of A to an algebraic closure of k and triviality of {\mbox{{Sh}}}(A)_{\textup{div}} . To the best of our knowledge, this gives the first examples of subvarieties of abelian varieties (other than abelian varieties themeselves) with X(k) infinite where these conjectures are known to hold. 1.1. Statement of the main results Theorem 1.1 . Let A be an abelian variety over a global function field k such that A_{{\overline{k}}} has no nonzero isotrivial quotient. For every closed subvariety X\subset A , we have \overline{X(k)}=X({\mathbb{A}}_{k})\cap\overline{A(k)}\,. Theorem 1.2 . Let A be an abelian variety over a global function field k such that A_{{\overline{k}}} has no nonzero isotrivial quotient. The following are equivalent. (1) {\mbox{{Sh}}}(A)_{\textup{div}}=0 , (2) \overline{A(k)}=A({\mathbb{A}}_{k})^{\Br} , (3) For every closed subvariety X\subset A we have \overline{X(k)}=X({\mathbb{A}}_{k})^{\Br} . If one fixes the subvariety X\subset A , then a more precise result is possible. A subvariety C\subset A is a coset if there is an abelian subvariety A^{\prime}\subset A and a point a\in A({\overline{k}}) such that C_{{\overline{k}}} is the translate of A^{\prime}_{{\overline{k}}} by the point a\in A({\overline{k}}) . In this case we simply write C=a+A^{\prime} . We emphasize that A^{\prime} and C are defined over k , though a need not be a k -rational point. A coset C=a+A^{\prime} is a torsor under the abelian variety A^{\prime} over k . Hence, a coset C=a+A^{\prime} with C({\mathbb{A}}_{k}) nonempty represents an element of {\mbox{{Sh}}}(A^{\prime}) . Assuming A_{\overline{k}} has no positive dimensional isotrivial quotient, the special subvarieties featuring in the Mordell-Lang conjecture mentioned above are cosets contained in X . The following gives an analogue of Mordell-Lang for the Brauer set. Theorem 1.3 . Let A be an abelian variety over a global function field such that A_{{\overline{k}}} has no nonzero isotrivial quotient. Let X\subset A be a closed subvariety. There is a finite collection of cosets C_{i}=a_{i}+A_{i}\subset X , i=1,\dots,r , such that X({\mathbb{A}}_{k})^{\Br}=\bigcup_{i=1}^{r}(C_{i}({\mathbb{A}}_{k})^{\Br})\,. Moreover, \overline{X(k)}=X({\mathbb{A}}_{k})^{\Br} if and only if {\mbox{{Sh}}}(A_{i})_{\textup{div}}=0 for all i=1,\dots,r . If X is coset free , i.e., X does not contain any coset of positive dimension, then the cosets C_{i} appearing in Theorem 1.3 are k -rational points and the theorem gives X(k)=X({\mathbb{A}}_{k})^{\Br} . Theorems 1.1 and 1.3 were proved for X_{{\overline{k}}} coset free (which implies that X is coset free, but not conversely) in [ PoonenVoloch , Theorems B and D] under an additional hypothesis on the p -primary torsion of A . Theorem 1.3 is proved for X a nonisotrivial curve of genus at least 2 (necessarily coset free) without the additional hypothesis on the p -primary torsion or the requirement that A_{\overline{k}} has no positive dimensional isotrivial quotient in [ CV3 ] . Partial results in the case of isotrivial curves may be found in [ CV , CPV ] . In all of the results mentioned in the previous paragraph, the hypotheses imply that X(L) is finite for every finite extension L/k , while there is no such requirement in the theorems stated above. An example of an X covered by Theorem 1.3 to which none of these previous results apply is given by the symmetric square of a nonhyperelliptic curve Y of genus 3 which admits a degree 2 map to a nonisotrivial elliptic curve E . Then X=\Sym^{2}(Y) contains a coset isomorphic to E . Suppose the Jacobian A of X splits up to isogeny as A_{\overline{k}}\sim E_{\overline{k}}\times B for a simple nonisotrivial abelian surface B/{\overline{k}} . Theorem 1.1 gives \overline{X(k)}=X({\mathbb{A}}_{k})\cap\overline{A(k)} . If, in addition, {\mbox{{Sh}}}(E)_{\textup{div}}=0 , then Theorem 1.3 gives \overline{X(k)}=X({\mathbb{A}}_{k})^{\Br} . Note that X(k) is infinite if E has positive rank. 1.2. Adelic Mordell-Lang A key input to the proofs of the theorems above are the ‘exceptional schemes’ appearing in Rössler’s proof of Mordell-Lang over function fields [ RosslerML ] . More recently, these were used by Wisson [ Wisson ] to prove a continuous version of Mordell-Lang in the function field case. A similar result was previously obtained by Hrushovski in [ Hrushovski ] using model theoretic methods. From Wisson’s result one can fairly easily deduce the following theorem (See Section 6 ). Theorem 1.4 . Let X\subset A be a closed subvariety of an abelian variety over a global function field k .