Fully 3D Monte Carlo ambient noise tomography over Grane field

We extracted Scholte waves from cross-correlations of twelve hours of continuous ambient noise recorded on 3458 sensors over Grane oil field. The fundamental modes of those cross-correlations are separated using a dispersion compensation method. We then automatically picked phase velocity dispersion curves for the fundamental mode and determined phase velocity maps using Eikonal tomography. To characterize the shallow subsurface structure, we performed many 1D Monte Carlo depth inversions to estimate 3D shear-velocity structures. However, the traditional independent depth inversion method loses lateral spatial correlations and introduces errors in final velocity models. Thus, we also inverted for a 2D section using a 2D parametrization that includes lateral spatial correlations and compared those results with the traditional 1D inversion. Finally, to determine a 3D shear-velocity model, we applied direct 3D Monte Carlo inversion. Comparisons are informative about methods and structures.