Moments of central values of quartic Dirichlet \(L\)-functions (Q2043492)

There has been a long history of research on moments of quadratic Dirichlet \(L\)-functions, but results on moments of higher order Dirichlet \(L\)-functions are sparse in the literature. Cubic \(L\)-functions have been investigated by \textit{S. Baier} and \textit{M. P. Young} [J. Number Theory 130, No. 4, 879--903 (2010; Zbl 1204.11135)]. The authors study the quartic case and prove the following result. \\ Theorem 1.1: Let \(w: (0,\infty)\rightarrow \mathbb{R}\) be smooth and compactly supported with Fourier transform \(\hat{w}\). Assuming the truth of the Lindelöf hypothesis, we have \[ \sum\limits_{(q,2)=1} \ \sideset{}{^\ast}\sum\limits_{\substack{\chi \bmod{q}\\ \chi^4=\chi_0}} \ L(1/2,\chi)w(q/Q)=CQ\hat{w}(0)+O\left(Q^{9/10+\varepsilon}\right), \] where \(C\) is an explicit positive constant and the asterisk on the sum over \(\chi\) restricts the sum to primitive characters \(\chi\) such that \(\chi^2\) remains primitive. \\ Using the eight moment of Dirichlet \(L\)-functions, they deduce under the Lindelöf hypothesis that there exist infinitely many primitive Dirichlet characters \(\chi\) of order 4 such that \(L(1/2,\chi)\not=0\). More precisely, the number of such characters with conductor less or equal \(Q\) is \(\gg Q^{6/7-\varepsilon}\). Second moment bounds are also provided in the paper under review. The proof of Theorem 1.1 starts with the approximate functional equation and then uses a correspondence between quartic Dirichlet characters and quartic residue symbols \(\left(\frac{m}{n}\right)_4\) with \(m\in \mathbb{Z}\) and \(n\in \mathbb{Z}[i]\) (and a relation between the corresponding Gauss sums). These quartic residue symbols are Hecke characters with trivial infinity type for \(\mathbb{Q}(i)\). After some transformations and an application of Mellin inversion, the evaluation boils down to expressions involving the corresponding Hecke \(L\)-functions. A main term is isolated which comes from those of these Hecke \(L\)-functions which are associated to a trivial character and therefore have a pole at \(s=1\) (i.e., they agree with the Dedekind zeta function for \(\mathbb{Q}(i)\) up to a finite Euler product). This main term is asymptotically evaluated. The rest is bounded using the Lindelöf Hypothesis and analytic properties of a Dirichlet series involving Gauss sums.