Evaluating and Improving Ground Motion Predictions for Scenario Earthquakes in The San Francisco East Bay by Integrating Earthquake Ground-Motion Simulations and Noise-Derived Empirical Green's Functions

Taka'aki Taira, & Arthur J. Rodgers

Submitted August 14, 2018, SCEC Contribution #8483, 2018 SCEC Annual Meeting Poster #003

Simulations of scenario earthquake ground motions play an increasingly important role in improving seismic hazard assessment. In the San Francisco Bay Area (SFBA) the USGS has developed a 3D geologic/seismic model. Simulations of large magnitude (M) 7 Hayward Fault scenario earthquakes show dramatic differences in peak ground motions across the fault, arising from wavespeed differences, which introduce the uncertainty in ground motion estimation. Preliminary comparison of observed and simulated waveforms from M 4 earthquakes reveals bias in the travel times of direct S-waves for paths crossing the East Bay Hills, east of Hayward Fault. To reduce the uncertainty in ground motion prediction for scenario earthquakes, we systematically evaluate bias in the current USGS 3D velocity model of the SFBA with emphasis on the East Bay Hills where strong ground motions are predicted. Our work involves performing earthquake ground motion simulations incorporating the USGS 3D velocity model for moderate Bay Area local earthquakes (M 3.5-4.5) including the recent 2018 Mw 4.4 Berkeley earthquake to quantify the accuracy of 3D model predictions by comparing observed and simulated waveforms. We also retrieve noise-based empirical Green’s functions (NEGFs) from ambient noise cross-correlation and earthquake coda waves to illuminate path-specific bias in the USGS 3D velocity model. We will report resultant path-specific bias to update the current 3D velocity model. Results from our work will provide the groundwork for the USGS 3D velocity model improvement and will result in improved accuracy of scenario ground-motion maps at the SFBA.

Citation
Taira, T., & Rodgers, A. J. (2018, 08). Evaluating and Improving Ground Motion Predictions for Scenario Earthquakes in The San Francisco East Bay by Integrating Earthquake Ground-Motion Simulations and Noise-Derived Empirical Green's Functions . Poster Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Ground Motions