Abstract:Let $E/F$ be an elliptic curve defined over a number field $F$ with complex multiplication by the ring of integers of an imaginary quadratic field $K$ such that the torsion points of $E$ generate over $F$ an abelian extension of $K$. In this paper we prove the $p$-part of the Birch and Swinnerton-Dyer formula for $E/F$ in analytic rank $1$ for primes $p>3$ split in $K$. This was previously known for $F=\mathbb{Q}$ by work of Rubin as a consequence of his proof of the Mazur--Swinnerton-Dyer ``main conjecture'' for rational CM elliptic curves, but the problem remained wide open for general $F$.
The approach in this paper, based on a novel application of an idea of Bertolini--Darmon--Prasanna to consider a carefully chosen decomposable Rankin--Selberg convolution of two CM modular forms having the Hecke $L$-function of interest as one of the factors, circumvents the use of $p$-adic heights and Bertrand's $p$-adic transcendence results in previous approaches. It also yields a proof of similar results for CM abelian varieties $A/K$, and for CM modular forms of higher weight.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about elliptic curves \(E\) with complex multiplication (CM) defined over a number field \(F\). When \(F(E_{\text{tors}})/K\) is an abelian extension, prove the \(p\)-part of the Birch and Swinnerton - Dyer formula under analytic rank 1. Specifically, in this paper, the author Francesc Castella proved that for elliptic curves \(E\) satisfying certain conditions, the \(p\)-part of the Birch and Swinnerton - Dyer formula holds, where \(p\) is a prime number greater than 3 and split in \(K\). Previously, this problem was only solved for the case of \(F = \mathbb{Q}\) by Rubin through proving the Mazur - Swinnerton - Dyer main conjecture, but for the general case of \(F\) it was still an open problem.
The main contribution of the paper lies in proposing a new method. This method is based on the ideas of Bertolini - Darmon - Prasanna and considers a suitably chosen decomposable Rankin - Selberg convolution of two CM modular forms, where one factor is the Hecke L - function of our interest. This method bypasses the \(p\)-adic heights used in previous methods and Bertrand's \(p\)-adic transcendence results, thus not only providing a new proof for \(F=\mathbb{Q}\), but also extending to the more general case of \(F\), and also applicable to higher - dimensional CM Abelian varieties \(A/K\) and CM forms of higher weights.
In short, this paper solves the problem of verifying the \(p\)-part of the Birch and Swinnerton - Dyer conjecture over a wider range of number fields, especially for elliptic curves with complex multiplication and related Abelian varieties.