Summary We present here a structural joint inversion (SJI) scheme to couple P-wave first arrivals from seismic refraction tomography (SRT) and Rayleigh-wave dispersion curves from surface-wave analysis (SWA). Structural similarity is established through the cross-gradient constraint to build P-wave and S-wave velocity models with improved geometrical consistency. The SJI approach is applied to two synthetic models with blocky piecewise-constant structures and one field data experiment conducted in a shallow aquifer, where initial independent analysis revealed structural similarity between the seismic data sets. For both, synthetic and field data experiments, we observe that the SJI provides better results compared to the individual inversions regarding the reconstruction of consistent model boundaries, while maintaining a similar data fit. The differences between independent and joint inversion results for the field data experiment are less pronounced. The identification of relationships between P- and S-wave velocity is nevertheless improved with the resolved models corresponding well to the known local geology and a complementary electrical data set.
Roser et al. (Sat,) studied this question.