Multimodal inversion of guided waves in georadar data

Abstract

We analyze the dispersive characteristics of the electromagnetic guided waves (at georadar frequency) to infer the electrical properties of materials that constitute a layered wave guide. WARR (Wide Angle Reflection and Refraction) georadar acquisitions could be carried out in TE (Transverse Electric) configuration to collect the full wavefield at different offset from the source. The dispersive curves of TE modes are obtained by transforming the space-time acquisition into frequency-wavenumber domain (f-k spectrum); the relative maxima in the f-k spectrum for each frequency represent the different propagation modes. We adopt both global and local inversion algorithms for minimizing the misfit function between computed and theoretical curves in order to obtain a 1D model of the layered subsoil (thicknesses and electrical permittivity). We perform a multimodal and multilayer inversion of the dispersive events. The results of two field cases will be discussed; the first one refers to the propagation in a confined waveguide (layered subsoil) and the other in a leaky waveguide (snow cover on a glacier).