Performance evaluation of the USGS Velocity model for the San Francisco Bay Area

Camilo Ignacio Pinilla Ramos, Arben Pitarka, David B. McCallen, & Rie Nakata

Submitted September 10, 2023, SCEC Contribution #13021, 2023 SCEC Annual Meeting Poster #006

The exponential growth of the processing power in the last decades pushed the development physics-based seismological simulations, allowing to solve the wave equation at higher frequencies and softer materials. The primary sources of epistemic uncertainty in the physics-based seismological simulations are the source characterization, the deformation and attenuation properties of the velocity model at elastic deformation regimes, and alternative ways of solving non-linear deformation regimes, especially important at materials close to the surface. This work is part of the EQSIM project, which assesses the San Francisco Bay Area (SFBA) seismic hazard using 3D physics-based seismological simulations. We quantify the epistemic uncertainty induced by propagating waves through the community velocity model for the SFBA developed by the USGS. As a methodology, we simulated seven small-magnitude earthquakes that occurred in the SFBA in the last 12 years. Because of the low level of ground-motion amplitude, the soil response during these small earthquakes is believed to be linear at all recording sites, leading to assume that the epistemic uncertainty induced by a linear constitutive model is zero. Furthermore, given the small rupture area of these events, their source can be modeled by a double-couple point source mechanism without losing so much resolution in the wave number range solved in our simulations. Therefore, the simulated variability and bias between observed and synthetic waveforms can be mainly attributed to the misrepresentation of the local geology, and a lesser extent, to the source parameterization, including location, depth, and focal mechanism. The simulations solved the wave equation up to 5 Hz with a minimum shear wave velocity of 250 m/s, allowing the inclusion of geotechnical layers in the simulations. The inclusion of geotechnical layers on the velocity model has a strong impact on the waveforms, especially at high frequencies. The analysis shows areas where the velocity model systematically induces over-prediction or under-prediction. Recognizing these areas provides space to identify where and how the velocity model can be improved.

Key Words
3D seismological simulations, path effects, community velocity model evaluation

Citation
Pinilla Ramos, C., Pitarka, A., McCallen, D. B., & Nakata, R. (2023, 09). Performance evaluation of the USGS Velocity model for the San Francisco Bay Area. Poster Presentation at 2023 SCEC Annual Meeting.

Related Projects & Working Groups
Seismology