Some local-global applications of Kummer theory (Q1396522): Difference between revisions

\loadeufm The paper deals with three problems, that might seem unrelated at first, but that are linked by their relations with various forms of Kummer theory: (1) Let \(G\) be an algebraic group defined over a number field \(k\), and denote by \({\mathcal O}_k\) the ring of integers of \(k\). Let \(a,b\in G(k)\) be such that the reduction of \(b\) in \(G({\mathcal O}_k/{\mathfrak p})\) lies in the subgroup generated by the reduction of \(a\) for almost all primes \({\mathfrak p}\) in \({\mathcal O}_k\). Does it follow that \(b\) lies in the subgroup generated by \(a\) in \(G(k)\)? (2) Let \(A\) and \(B\) be two abelian varieties defined over \(k\), where \(k\) is either a number field or a finite field. Assume that there exists an integer \(m\geq 1\) such that \(k(A[d])=k(B[d])\) for all integers \(d\) with \((d,m)=1\). What can one say about \(A\) and \(B\)? (3) Let \(k\) be a number field and \(a,b\) be integers in \(k\) such that for all \(n\geq 1\) the largest integer dividing \(a^n-1\) is the same as the largest integer dividing \(b^n-1\). What relation holds between \(a\) and \(b\)? It is shown that the answer to Problem 1 is positive when \(G\) is either an elliptic curve or the multiplicative group (this last case was already solved by \textit{A. Schinzel}, ``On the congruence \(a^x \equiv b\) (mod\ \(p\))'', Bull. Acad. Pol. Sci., Sér. Sci. Math. Astron. Phys. 8, 307--309 (1960; Zbl 0094.25504)]. Regarding Problem 2, it is proven that if \(A\) is an elliptic curve and \(B\) is an abelian variety with no repeated simple isogeny factors, then the condition is equivalent to isogeny over \(k\). When \(k={\mathbb F}_{p^f}\) is a finite field, it is shown that, for a given \(A\) of dimension \(d\), there are only at most \(c_{f,d}\) isomorphism classes af abelian varieties \(B\) with no repeated isogeny factors that satisfy the assumption (here \(c_{f,d}\) is an effective constant depending only on \(f\) and \(d\)). Finally, the author completely solves Problem 3, by showing that the condition is equivalent to the equality between the multiplicative subgroup generated by the conjugates of \(a\) and the multiplicative subgroup generated by the conjugates of \(b\).