Passive multi-channel analysis of surface waves with cross-correlations and beamforming. Application to a sea dike

We introduce the use of cross-correlations in the passive multi-channel analysis of surface waves (MASW), and report an improvement in the determination of subsurface shear velocities from ambient seismic noise. Velocities are measured from phase-shifts that are also related to the source location. Consequently, the accuracy with which velocities can be inferred depends on the ability of the array to locate noise sources. The computation of cross-correlations for each receiver pair allows increasing the effective spatial sampling of the array. For this reason, we show that beamforming is more efficient with cross-correlated signals. Consequently, MASW performed with cross-correlations produces a dispersion diagram where aliasing is reduced and signal-to-noise ratio increased. The proposed method is validated with synthetic records. It is then applied on passive recordings obtained on top of a sea dike at high tide, where sea waves were acting as continuous seismic sources. Surface wave velocities that favorably compare with hammer shot measurements are inferred.