Seismic Tomography of Southern California and Alaska via Bayesian Joint Inversion

Abstract

This dissertation investigates seismic tomography of the crust of Southern California and the crust and upper mantle of Alaska. The main goal is to use earthquake and ambient-noise data to resolve subsurface structure via Bayesian joint inversion, which also quantifies uncertainty and sensitivity. This is achieved by computing ambient-noise cross-correlations, retrieving Rayleigh-wave phase velocity and ellipticity measurements, performing Bayesian Markov Chain Monte Carlo (MCMC) joint inversions of resolved datasets, and interpreting our velocity model results to gain new understandings of the subsurface.

To study the upper crust of Southern California, we calculated multicomponent ambient-noise cross-correlations from 1-year of continuous recordings across over 300 stations. From these, we retrieved Rayleigh-wave phase velocity and ellipticity measurements and combined these in a MCMC joint inversion to resolve the upper 10 km shear-velocity (Vs) structure across Southern California. We find high-velocities in the Sierra Nevada and Peninsular Ranges, velocity contrasts across major faults, and local-scale detailed structures within the Los Angeles and Ventura basins and Salton Trough.

Next we incorporate early-times of receiver functions with our previous datasets in a MCMC joint inversion at over 200 stations to gain stronger constraint in the near-surface of Southern California. We resolve Vs and the ratio of compressional to shear velocity (Vp/Vs) structures and find the near-surface Vs-structure more consistent with local borehole studies. We observe high Vp/Vs in fluid-saturated sediment basins and low Vp/Vs that varies according to silica-content and origin of local geological features.

From multiple years of continuous recordings on over 200 stations in Alaska, we determine Rayleigh-wave phase velocity and ellipticity measurements from both ambient-noise cross-correlations and teleseismic earthquakes, and high-quality receiver functions from teleseismic body waves to resolve subsurface interfaces. We perform a MCMC joint inversion at each station and a follow-up deterministic inversion across the region. Our imaged results include near-surface basins, thinner crust within interior Alaska, and insight on how the subsurface impacts volcanism in the Aleutians, Denali volcanic gap, and Wrangell Volcanic Field.

The success of these projects demonstrates the possibilities of joint Bayesian inversions and information from the subsurface that can be gained from these types of tomography studies.