Lithospheric imaging via teleseismic scattering tomography

The coda of the teleseismic P phase consists largely of energy scattered by small inhomogeneities in the receiver-side lithosphere. Given large collections of teleseismic data from dense permanent networks, previous workers have successfully back-propagated coda energy back to scattering source points using various kinematic migration schemes, as well as by inverting using an inverse scattering/radon transform approach. Under the Born approximation, seismic scattering is a linear process; therefore it is possible to approach coda scattering as a linear waveform inversion problem, mathematically similar to transmission-based tomography. Assuming ray-theoretical propagation and Rayleigh scattering, we pose the inverse scattering problem in tomographic form, and recover perturbations in density and P and S velocities from Pp and Ps scattered data. The method is applied to data from the Southern California Seismic Network (SCSN) covering the San Jacinto-Anza region. The results show a considerable correlation between seismicity and velocity perturbation structure, particularly in the region between the Mission Creek and Banning fault branches. Features connecting the Coyote Creek and Elsinore faults at right angles are correlated with seismicity lineations and may represent conjugate faulting with no surface expression.