Geophysics Seminar: Bridging the Gap between Strong-Motion Seismology and Geodesy: Examples from Japan and California

Abstract:

Due to the large number of unknowns, slip inversions are generally underdetermined and non-unique. This issue is usually addressed by imposing smoothness and/or dumping constraints. Indeed, regularizing the set of equations can turn the problem overdetermined and more stable, but it does not render the solution more trustable. Additionally, slip distributions on segments located farther away from the geodetic network are (much) less well-resolved than those on segments closer to the network. Here we present a new approach aimed at enhancing the resolving power of the inversion procedure in places where the geodetic-only inversion cannot resolve the final slip.

We study the static slip distribution due to the M6.0 2004 Parkfield earthquake, which ruptured a ~35 km long densely monitored segment of the San Andreas Fault. Since near-fault 1 Hz GPS data are only available from 11 sites located to the northwest of the epicenter, the static slip distribution to the southeast of the epicenter is poorly resolved. To address this issue, we have developed an approach for extracting permanent offsets from body-wave spectra of strong motion seismograms recorded at the near-field, where the velocity spectra at frequencies well below the corner frequency approaches the permanent offset. We validated this approach at sites of collocated (or nearly collocated) strong motion sensors and 1 Hz GPS receivers in Japan. By incorporating permanent offsets inferred from strong-motion records satisfying the near-field criteria, we increased the number of permanent offsets in Parkfield by a factor of 4, and dramatically enhanced the resolution of the slip inversion. Preliminary results reveal notably more co-seismic slip than previously inferred southeast of the hypocenter.

Seminar Organizers: Prof. Moshe Reshef and Dr. Alon Ziv