Scattering theory for twisted automorphic functions (Q4211074)

The author systematically develops a scattering theory. Let us consider \(\Gamma\) a discrete cofinite but not cocompact group of motions acting on a hyperbolic plane \({\mathcal H}\) and let \(\rho\) be an irreducible unitary representation of the group \(\Gamma\) acting on a Euclidean vector space \({\mathcal V}\) of dimension \(n.\) Let us suppose that \(u\) is a \((\Gamma, \rho)\)-automorphic \({\mathcal V}\)-valued function on \({\mathcal H}\) such that \(u(\gamma z)=\rho(\gamma)u(z).\) The aim of the author is to study a scattering theory for the \((\Gamma,\rho)\)-automorphic wave equation \[ \partial_t^2 u = \Delta_{\rho}u,\qquad u(z,0)= f_1(z),\qquad \partial_t u(z,0)= f_2(z) \] where all the functions are \((\Gamma,\rho)\)-automorphic. The technique used in this note is the Lax-Phillips approach used in previous papers of these two authors. It is based on two translation representations, incoming and outcoming, for the solution operator of the wave equation. The scattering operator is the mapping from the incoming translation representation to the outgoing representation. It is possible to write the corresponding spectral representation of this kind of operator in terms of the constant term of the Eisenstein series; then the author proves that, if the dimension of \(\rho\) equals 1, no element of the scattering operator vanishes, and when the dimension \(\rho\) is more then 1, this is no longer the case. A useful example shows that some elements of the operator described can vanish for \(\Gamma(2)\)-representation of dimension 2.