Steklov zeta-invariants and a compactness theorem for isospectral families of planar domains

TL;DR

This paper investigates the inverse Steklov spectral problem for planar domains, introduces zeta-invariants related to the Dirichlet-to-Neumann operator, and proves a compactness theorem for isospectral families using these invariants.

Contribution

It establishes a lower bound for zeta-invariants for real functions and proves a compactness theorem for isospectral planar domains in the smooth topology.

Findings

Zeta-invariants are determined by the Steklov spectrum for positive functions.

A lower estimate for zeta-invariants is obtained for real functions.

Steklov isospectral families of planar domains are compact in the $C^ abla$-topology.

Abstract

The inverse problem of recovering a smooth simply connected multisheet planar domain from its Steklov spectrum is equivalent to the problem of determination, up to a gauge transform, of a smooth positive function $a$ on the unit circle from the spectrum of the operator $a\Lambda$, where $\Lambda$ is the Dirichlet-to-Neumann operator of the unit disk. Zeta-invariants are defined by $Z_m(a)={\rm Tr}[(a\Lambda)^{2m}-(aD)^{2m}]$ for every smooth function $a$. In the case of a positive $a$, zeta-invariants are determined by the Steklov spectrum. We obtain some estimate from below for $Z_m(a)$ in the case of a real function $a$. On using the estimate, we prove the compactness of a Steklov isospectral family of planar domains in the $C^\infty$-topology. We also describe all real functions $a$ satisfying $Z_m(a)=0$.