Wehrl inequalities for matrix coefficients of holomorphic discrete series
Robin van Haastrecht, Genkai Zhang
TL;DR
The article extends Wehrl-type $L^2-L^{2n}$ inequalities to matrix coefficients of vector-valued holomorphic discrete series on Hermitian symmetric spaces realized as Bergman spaces. It develops a tensor-power/Cartan-component framework, identifies the sharp constants via Harish-Chandra formal degrees, and proves that reproducing kernels are precisely the maximizers. A detailed calculation of the normalization constant $c_G$ is given, including explicit formulas across classical types, with the reproducing kernel $K(z,w)$ furnishing the equality cases. An additional improved inequality is established for the unit disk, featuring a derivative-remainder term, and the work outlines open questions and potential extensions to broader Bergman spaces and quantum-channel contexts.
Abstract
We prove Wehrl-type $L^2(G)-L^{p}(G)$ inequalities for matrix coefficients of vector-valued holomorphic discrete series of $G$, for even integers $p=2n$. The optimal constant is expressed in terms of Harish-Chandra formal degrees for the discrete series. We prove the maximizers are precisely the reproducing kernels.
