A converse theorem for quasimodular forms (Q2121542)

The converse theorem of Hecke, generalized to higher levels by Weil, proves that the meromorphic continuation of a Dirichlet series, with a boundedness condition and a functional equation of an appropriate completion, determine this Dirichlet series to be the \(L\)-function a modular form. This paper extends this converse theorem to more general Dirichlet series, relating them quasimodular forms. In more detail, Hecke showed that if a Dirichlet series can be extended meromorphically to the entire complex plane, is bounded on vertical strips, and multiplying it by the Gamma function and a power of \(2\pi\) yields a function that is invariant under the transformation \(s \mapsto k-s\), then this series is the \(L\)-function associated with a modular form. Weil established a more general result, in which two Dirichlet series are related by the functional equation, and then they are associated (under similar conditions) with a modular form and its image under the Fricke involution. The twisting by a character is also allowed, extending the level further. In this paper the authors consider pairs of finite systems of Dirichlet series, with functional equations in which one Dirichlet series from one side of the pair is associated with an appropriate linear combination of the system in the other side of the pair. They show that each such system of Dirichlet series arises from the Fourier expansion of a quasimodular form (and the components of its quasimodular transformation), and the other system in the pair corresponds to the image of this quasimodular form by an appropriately defined action of the Fricke involution. As consequences they establish the quasimodularity of an explicit series defined by Ramanujan, as well as the result that when a quasimodular form and its Fricke image have no constant terms, their Fourier coefficient change sign infinitely often in any direction. The paper is divided into 6 sections. Section 1 is the Introduction with the statement of results, and Section 2 defines the basic notions of quasimodular forms, the associated nearly holomorphic modular forms, and the action of the Fricke involution. Section 3 proves the converse theorem involving untwisted quasimodular forms. Section 4 then establishes some intermediate results about the twisting operation, and Section 5 uses these results to prove the converse theorem also with the twists. Finally, Section 6 deduces the two stated consequences.