Periods, subconvexity of \(L\)-functions and representation theory (Q2494205)

Let \(\Gamma\) be a cocompact Fuchsian group acting on the upper half plane \(\mathbb{H}\). Assume that \((\varphi_j)_{j\geq 0}\) is a sequence of Maaß forms on the compact Riemann surface \(Y=\Gamma \setminus\mathbb{H}\) forming a complete orthonormal system in \(L^2(Y,dv)\) where \(v\) is the natural hyperbolic measure on \(Y\). The work under review is a contribution to the problem of estimating the triple periods \[ \int_Y\phi_j\phi_k\phi_\ell \,dv \] in terms of the corresponding eigenvalues \(\lambda_j\) etc. This problem is closely related to the problem of determining bounds on automorphic \(L\)-functions as becomes evident from a beautiful formula of T. Watson's Princeton Thesis (2001) which can be viewed as a far-reaching generalization of the Rankin-Selberg formula. In the present work the authors fix two of the Maaß forms, say, \(\phi\) and \(\phi'\), in the triple product and estimate the coefficient \[ c_j:=\int_Y \phi\phi'\phi_j\,dv \] as \(j\to\infty\). In view of previous work (see below) of the authors it is natural to consider the renormalized coefficients \[ d_j:=\gamma (\lambda_j)|c_j|^2 \] where \(\gamma(\cdot)\) is an explicit expression in terms of the gamma function. The coefficients \(d_j\) were shown to satisfy the bound \[ \sum_{|\lambda_j|\leq T}d_j\leq AT^2\quad (T>1)\tag{*} \] for some explicitly computable constant \(A\) whence \(c_j\) decays exponentially in \(|\lambda_j|\). The best individual bound which can be deduced from (*) corresponds to the convexity bound for the corresponding \(L\)-function. The main contribution of the work under review is an outline of the proof of Theorem 1.1: There exist effectively computable constants \(B\), \(b>0\) such that \[ \sum_{|\lambda_j|\in I(T)}d_j\leq BT^{5/3} \] where \(I(T)\) is the interval of length \(bT^{1/3}\) centered at \(T\). This result leads to a subconvexity bound for the value \(L(\frac 12,\varphi\otimes\varphi'\otimes\varphi_j)\) of the triple \(L\)-function which is the first subconvexity bound for an \(L\)-function of degree 8. The sketch of proof given here is based on ideas from representation theory which are more fully worked out in the authors' work in Moscow Math. J. 4, No. 1, 19--37 (2004; Zbl 1081.11037), arXiv:math. RT/0305351.