Treating Active Data As Noise - Low-frequency Surface Wave Inversion Using SPAC

We apply a passive array technique, the spatial autocorrelation (SPAC) method to active seismic data, by treating the frequencies below the vibroseis sweep as noise. Rayleigh wave phase velocities are measured locally within the spread. A phase velocity map is then generated by interpolation of many such measurements. We compute SPAC curves in the inline direction only, rather than azimuthal averaging, which could compensate for possible bias in a directional noise field. Using one direction only, the assumptions regarding source distribution are identical to seismic interferometry. We verified that differences in velocity by measuring in different directions (using adjacent receiver lines) are consistent with the local heterogeneity and are not caused by a potential bias due to the noise directionality. Integrating the low-frequency dispersion curves with active surface wave analysis enables inversion for a deeper shear wave model.