On reduction maps and support problem in \(K\)-theory and abelian varieties (Q852539): Difference between revisions

The support problem for number fields was formulated by P. Erdös and solved by \textit{C. Corrales-Rodrigáñez} and \textit{R. Schoof} [J. Number Theory 64, No. 2, 276--290 (1997; Zbl 0922.11086)], where they also stated and solved the analog of the problem for elliptic curves. The case of abelian varieties for which one can get some important facts about the image of the Tate \(\ell\)-adic representation, as well as the \(K\)-theoretical version of the support problem was treated by \textit{G. Banaszak, W. Gajda} and the reviewer [J. Number Theory 100, No. 1, 133--168 (2003; Zbl 1088.11040) and also \textit{G. Banaszak, W. Gajda} and \textit{P. Krasoń}, J. Number Theory 100, No. 1, 117--132 (2003; Zbl 1056.11034)]. The support problem for abelian varieties was independently studied by \textit{C. Khare} and \textit{D. Prasad} [J. Number Theory 105, No. 2, 322--332 (2004; Zbl 1080.11046)]. The above mentioned papers give affirmative answer to the support problem for abelian varieties satisfying certain specific conditions. However in general the answer is negative [\textit{M. Larsen}, J. Number Theory 101, No. 2, 398--403 (2003; Zbl 1039.11040) and also \textit{M. Larsen} and \textit{R. Schoof}, J. Number Theory 116, No. 2, 396--398 (2006; Zbl 1097.14036)]. The author considers the generalization of the support problem for abelian varieties and \(K\)-theory to the case of two sets of points. In the case of abelian varieties his theorem is as follows: Theorem 8.2. Let \(A\) be an abelian variety defined over number field \(F\) such that End\((A)\) is an integral domain. Let \(P_{1} \dots P_{n}, Q_{1}\dots Q_{n} \in A(F)\) be the points of infinite order. Assume that for almost every prime \(\ell\) the following condition holds: For every set of positive integers \(m_{1}\dots m_{n}\) and for almost every prime \(v\) \[ m_{1}r_{v}(P_{1})+\cdots + m_{n}r_{v}(P_{n})=0 \qquad {\text{ implies }}\qquad m_{1}r_{v}(Q_{1})+\cdots + m_{n}r_{v}(Q_{n})=0 \] Then there exist \(k_{i}\in {\mathbb Z}\backslash\{0\}, {\beta}_{i}\in \text{End}(A)\backslash\{0\}\) such that \(k_{i}P_{i}+ {\beta}_{i}Q_{i}=0\) for every \(i\in \{ 1,\dots , n\}\). The methods rely on those developed in [\textit{G. Banaszak, W. Gajda} and \textit{P. Krasoń}, J. Number Theory 115, No. 2, 322--342 (2005; Zbl 1089.11030)].