An approach to the Selberg trace formula via the Selberg zeta-function (Q1089034): Difference between revisions

Selberg's trace formula for a quotient \(\Gamma\setminus H\) of the upper half plane \(H\) by a cofinite discrete subgroup \(\Gamma\) of \(PSL(2,{\mathbb{R}})\) relates data on the spectral decomposition of the Laplacian in \(L^ 2(\Gamma \setminus H)\) to geometric data. At present the standard reference for full proofs of the trace formula is [\textit{D. A. Hejhal}, The Selberg trace formula for PSL(2,\({\mathbb{R}})\), Vol. 2 (Lect. Notes Math. 1001) (1983; Zbl 0543.10020) (called ''Trace 2'' in this review)]. The idea of the proof is to consider certain integral transformations in \(L^ 2(\Gamma \setminus H)\) which are of trace class, and to compute the trace in two ways. The resulting equality is the trace formula. The zeta function of Selberg, \(Z(s)\) is a meromorphic function containing information on the lengths of the periodic geodesics of \(\Gamma\setminus H\); it is analogous to the zeta function of Riemann represented as product over the primes. A special choice of the test function in the trace formula leads to the meromorphic continuation and functional equation of Z(s). The idea of the book under review is to put \(Z(s)\) central. The trace formula is first proved for test functions corresponding to very special integral transformations, representing the resolvent in \(L^ 2(\Gamma \setminus H)\) of the Laplacian. This gives an identity between two expressions for the logarithmic derivative of \(\Xi(s)\), where \(\Xi(s)\) denotes a holomorphic function with \(Z(s)\) as a factor. The zeros of \(\Xi\) correspond to the spectrum of the Laplacian. The growth and Weierstrass factorization of \(\Xi\) are studied. The Weyl-Selberg formula for the spectral density of the Laplacian is derived. Finally the general trace formula is shown to follow from this restricted one. All this is carried out in the generality of ''version C'' in Trace 2; i.e. arbitrary real weight, \(\Gamma\) a Fuchsian group of the first kind, a unitary multiplier system with values in a finite dimensional Hilbert space, the Laplacian is generalized as the Casimir operator acting in the particular weight considered. As foundation the author uses the work of Roelcke and Elstrodt [\textit{W. Roelcke}, Math. Ann. 167, 292-337 (1966); 168, 261-324 (1967; Zbl 0152.07705); \textit{J. Elstrodt}, Math. Ann. 203, 295-330 (1973; Zbl 0239.10014); Math. Z. 132, 99-134 (1973; Zbl 0246.10016); Math. Ann. 208, 99-132 (1974; Zbl 0265.10017)]. A few results on the determinant of the scattering matrix are taken from Trace 2. This book gives an interesting approach to the trace formula. In addition, the results on the growth and factorization of \(\Xi(s)\) are new, as far as I know. Proving the trace formula asks for a lot of work, whatever method of proof one uses. The streamlined presentation of this book gives, in my opinion, an easier access to the trace formula in this generality than Trace 2. (But the latter book gives another proof and goes into many subjects not treated here.)