On the Beilinson-Bloch-Kato conjecture for Rankin-Selberg motives (Q2118064)

This paper proves an interesting and important result in the framework of Bloch-Kato conjecture for motives associated to Rankin-Selberg products of conjugate self-dual automorphic representations. We state the main results of the paper in a somehow simplified version. Let \(F/F^+\) be be a totally imaginary quadratic extension of a totally real number field \(F^+\neq \mathbb{Q}\). Fix an integer \(n\geq 2\), and denote by \(n_0\) and \(n_1\) the unique even and odd numbers in \(\{n,n+1\}\), respectively. We also fix two cuspidal automorphic representation \(\Pi_0\) and \(\Pi_1\) of \(\mathrm{GL}_{n_0}(\mathbb{A}_F)\) and \(\mathrm{GL}_{n_1}(\mathbb{A}_F)\), respectively. One assumes that these representations are conjugate self dual, and for any Archimedean place \(w\) of \(F\) one also assumes that \(\Pi_{i,w}\) is isomorphic to the principal series representation induced by the characters \((\arg^{1-n_i},\arg^{3-n_i},\dots,\arg^{n_i-3},\arg^{n_i-1})\) where \(\arg\) is the argument character defined by \(\arg(z)=z/\sqrt{z\bar{z}}\) (these representations are called \emph{relevant} in the paper; in particular, they are regular algebraic). One also assumes some further tecnical conditions on \(\Pi_i\): the existence of a suitable inert prime (called \emph{very special}) of \(F\) where \(\Pi_0\) is Steinberg and \(\Pi_1\) is unramified with distinct Satake parameters, and a finite place where \(\Pi_i\) is supercuspidal. The main results of the paper are the following implications: \begin{itemize} \item If \(L(1/2,\Pi_0\times\Pi_1)\neq0\), then the Bloch-Kato Selmer group \(H^1_f(F,\rho_{\Pi_0,\lambda}\otimes_{E_\lambda}\rho_{\Pi_1,\lambda}(n))\) vanishes, for all \(\lambda\) except possibly a finite number; \item If \(L(1/2,\Pi_0\times\Pi_1)=0\) and \(L'(1/2,\Pi_0\times\Pi_1)\neq0\), then the Bloch-Kato Selmer group \(H^1_f(F,\rho_{\Pi_0,\lambda}\otimes_{E_\lambda}\rho_{\Pi_1,\lambda}(n))\) has dimension \(1\). \end{itemize} Here \(\lambda\) is a prime ideal in a suitable coefficient field \(E\) (called \emph{strong coefficient field for both \(\Pi_0\) and \(\Pi_1\)}), \(E_\lambda\) is the completion of \(E\) at \(\lambda\), and \(\rho_{\Pi_i,\lambda}\) is the Galois representation of \(F\) with coefficients in \(E_\lambda\) associate with \(\Pi_i\). These results apply to all \(\lambda\), except possibly a finite number. More general results, included a beautiful applications to symmetric powers of elliptic curves, are given. The main argument of the paper uses a level-raising approach, which follows from the work of Bertolini-Darmon on anticyclotomic Iwasawa Main Conjecture for elliptic curves, which has been also used by one of the authors (W.Z.) to obtain the \(p\)-part of the BSD conjecture for elliptic curves. Although the paper is very technical, the main ideas and arguments are clear and clean, and the results proven are very important in this research area.