Hierarchical Seismic Imaging of the Crust in Southern California

Session: Seismic Imaging of Fault Zones
Type: Oral
Date: 4/29/2020
Time: 03:00 PM
Room: 215 + 220
Description: 

The standard earthquake catalog of the Southern California (SoCal) Seismic Network comprises over 350,000 analyst-reviewed events since 2000. White et al. [J. Geophys. Res., 124, 6908–6930 (2019)] processed nine years of raw waveform data (2008 through 2016) adding data from the Anza Network and dense deployments in the San Jacinto Fault Zone (SJFZ) region using automated processing procedures to derive a new catalog with more than 100,000 events in the SJFZ region alone. The rapid deployment of two dense geophone arrays following the M7.1 Ridgecrest and M6.4 Searless Valley earthquakes will yield an exceedingly rich data set. These data can benefit significantly from comprehensive and integrated analysis, but many conventional techniques overwhelm computational resources when processing data sets this large so some data are sacrificed to make processing tractable. Additional challenges arise when combining results from heterogeneous data sets to properly contextualize detailed local results within their regional setting.

Recent methodological improvements in seismic tomography combining stochastic Voronoi cells and the Fast Marching Method for solving the Eikonal equation hold promise for inverting large complementary data sets. In this work we process several data sets from SoCal using the new method for tomography developed by Fang et al. [Seismol. Res. Lett. 91 (1), 343-355 (2020)]. First, we validate the methodology by independently deriving regional and local velocity models for SoCal and the SJFZ, respectively. The results for the SJFZ show strong correlations between the distribution of microseismicity and velocity structure. Then we implement a non-linear relocation step and seek a unified model that is consistent with different independent input data sets. The accuracy of the derived velocity models is tested by comparing simulations of ground motion using candidate velocity models and sets of earthquakes with observed seismograms. Updated inversion results and further methodological developments will be presented at the meeting.

Presenting Author: Malcolm C. A. White

Authors