Joint Inversion for Seismic P- and S-wave Velocity Structures with Interfaces

Most joint inversion methods using surface wave dispersion and receiver functions invert the data for crustal S-wave velocities only and the P-wave velocities are determined by fixing the Vp/Vs ratio. In addition, they use fixed layer thicknesses and do not invert for interface (e.g. Moho) depths. This is mainly because of lack of crustal P-wave velocity information in the data and the strong trade-off between the Moho depth and crustal Vp/Vs ratio. To overcome these problems, we developed a new joint inversion method that inverts for interface depths, shear wave velocities, and Vp/Vs ratios. We combine surface dispersion data with travel-time data of interface-related P and S waves such as the Moho reflected P wave and P-to-s converted wave. The travel-time data are sensitive to the crustal P-wave velocities and Moho depth and therefore can help to constrain these parameters. We use an iterative linearized inversion by computing the partial derivatives of the data with respect to the model parameters. Smoothness constrain is imposed to velocity variation between interfaces. The velocity model obtained from the new joint inversion method has well-defined interfaces whose depths are determined by the inversion instead of picked after the inversion. We apply the new inversion method to data collected in the Wabash Valley Seismic Zone. The results show a NW dipping Moho from depth of 50 km at SE side to 60 km at NW side. The sedimentary layer shows similar dipping feature at depths of 2 to 3 km. We also noticed high velocity anomalies in the mid crustal beneath the rift axis which might be related to high-density magmatic materials.