Calibration of the Near-Surface Seismic Structure in the San Francisco Community Velocity Model (SFCVM)

Anupam Patel, & Kim B. Olsen

Published April 15, 2026, SCEC Contribution #15039

We have calibrated the near-surface structure of the San Francisco Community Velocity Model (SFCVM) with the aim to improve its ground motion prediction efficacy using Fourier Amplitude Spectral (FAS) bias up to 1 Hz via 3D physics-based wave propagation simulations for a series of Mw3.8–4.5 earthquakes. In areas of systematic underprediction we implement a Vs30-constrained spatially varying low-velocity taper (LVT), with the optimal station-specific thickness determined from trial-and-error simulations for different taper depths. The optimal spatially variable LVT reduces the FAS bias for areas outside the basins (defined as sites with surface Vs > 1 km/s) by up to 18% on average for the small earthquakes. In the San Francisco East Bay area, the addition of a shallow (25 - 75 m) low-velocity layer (Vs = 250 m/s) further improves the FAS bias by increasing the amplitudes and duration of the synthetics. In areas where the reference model systematically overpredicts amplitudes of the source ensemble, we calibrate the model by removing the near surface (up to 900 m) decrease in velocities, which further improves the FAS bias. The depth to the isosurface of Vs = 1.0 km/s (Z1.0) in the SFCVM is increased by up to about 100 m outside the basins due to the implementation of the optimized LVT, with implications for the depth-dependent basin amplification terms used in Ground Motion Models.

Citation
Patel, A., & Olsen, K. B. (2026, 04). Calibration of the Near-Surface Seismic Structure in the San Francisco Community Velocity Model (SFCVM). Oral Presentation at Seismological Society of America Annual Meeting.

Related Projects & Working Groups
Groun Motion, Community Earth Models