Ambient noise characterisation with three-component beamforming to inform and improve interferometric Green’s function estimates

Abstract

In the GeoHEAT project to advance geothermal exploration in Thurgau, Switzerland, passive seismic imaging plays a major role in mapping the topography of the crystalline basement and identifying potential sedimentary troughs and deep fractured zones. To this end, ambient noise tomography (ANT) is performed, which relies on Green’s function estimates obtained from ambient noise cross-correlations (interferometry). Since the quality of the retrieved Green’s function estimate depends on the properties of the ambient noise wavefield, we use three-component beamforming to analyse the dominant contributions to the wavefield recorded on a temporary nodal network. Deciphering the particle motion allows us to discriminate retro- and prograde Rayleigh waves and shows that the latter dominate a significant part of the frequency range of interest. Frequency-wavenumber analysis further reveals that this prograde motion belongs to the first higher mode Rayleigh wave. We conclude that this higher mode is also controlling the Green’s function estimates retrieved from ambient noise interferometry at the corresponding frequencies. This explains the difficulties encountered when trying to fit dispersion curves retrieved from interferometry for certain station pairs and highlights the need for improved interferometry schemes that use additional wavefield information from beamforming.