Earthquake Science Center Seminars

Validation of Two Recent CyberShake Ground Motion Simulation Studies

Date & Time

Wednesday, July 2, 2025 at 10:30 AM Pacific

Location

Online-only seminar via Microsoft Teams

Host

Jeanne Hardebeck

Summary

The CyberShake platform, developed by the Statewide California Earthquake Center (SCEC), performs physics-based ground motion simulations. CyberShake uses 3D wave propagation simulations with a reciprocal approach to synthesize seismograms, intensity measures, site-specific hazard curves, and regional hazard maps. Two recent simulation studies have just been completed: Study 22.12 in southern California and Study 24.8 in northern California. Compared to previous studies, several critical changes are made in the two studies, which include 1) the use of an updated USGS San Francisco Bay 3D velocity model; 2) the use of a regional velocity model constructed through tiling the USGS San Francisco Bay region 3D model, the CCA-06 Central California tomography-derived model, and a 1D background model, with a near-surface low velocity layer applied using the Ely & Jordan approach; 3) a reduction in minimum Vs to 400 m/s; 4) the use of an updated rupture generator (i.e., Graves-Pitarka v5.5.2).

To validate the two recent studies, we develop CyberShake-based GMMs and compare them to NGA-West2 GMMs. The median predictions from Study 22.12 are very close to NGA-West2 GMMs, while the median predictions from Study 24.8 are significantly higher especially at 2s. The overprediction in Study 24.8 is caused by the strong directivity effects, which are likely results of how rupture velocity along the faults is defined in the rupture generator. Moreover, we find that the variability of source and site effects are underrepresented in CyberShake studies at shorter periods, due to small stress drop variations and imperfection of 3D velocity models. Finally, based on our analysis, we propose various ways to improve future CyberShake simulations.