3D Simulations of Ground Motions of the 1989 Loma Prieta Earthquake Using the USGS SF06 Velocity Model

David Dolenc, Doug Dreger, and Shawn Larsen (LLNL)

Introduction

The United States Geological Survey has recently released an updated 3D velocity model (USGS SF06) of the greater San Francisco Bay Area that will be used to evaluate shaking hazard for possible future events in the region. In this study we used the 3D model and the 3D finite-difference code E3D (Larsen and Schultz, 1995) to simulate ground motions of the 1989 $M_{w}$ 6.9 Loma Prieta event. Comparison of simulated ground motions with observations served to validate the velocity model for scenario studies of future large earthquakes in the region. This work was done as part of the SF06 Simulation Project Group modeling of the 1906 $M_{w}$ 7.8 San Francisco earthquake (http://www.sf06simulation.org).

Modeling

The modeling of the great 1906 San Francisco earthquake for the event's 100th anniversary was one of the reasons for the USGS SF06 velocity model release. The model extends to 45 km depth and consists of a detailed model with approximate dimensions of $290 \times 140$ km. This detailed model is nested inside an extended model which has approximate dimensions of $650 \times 330$ km. The extended model was provided mainly to accommodate the 1906 San Francisco earthquake simulations and as a buffer zone for the simulations with either sources or stations close to the edge of the detailed model. The model's main improvement over the previous model versions is that more realistic material properties were assigned to individual geologic units (Brocher, 2005). Also, the model now includes topography and material properties of water for the Pacific Ocean and the San Francisco Bay.

To use the USGS SF06 model with the finite-difference code, we queried the model to extract values on a grid with a 125-m spacing. For the Loma Prieta earthquake, we employed Wald et al. (1991)'s strong motion and strong motion/teleseismic finite-source models. In addition, we used the Beroza (1991) strong motion finite-source model as a comparison. The dimensions of the simulation were $220 \times 135 \times 50$ km and the grid spacing was 125 m. Because of computation limitations, the slowest velocities in the model were increased to a minimum S-wave velocity of 500 m/s, corresponding to a maximum modeled frequency of 0.8 Hz. In our simulations we included water (S-wave velocity 0.0 m/s), but not topography or attenuation. The computations were performed on the BSL Linux cluster. We compared results obtained with the three different finite-source models to observations. In addition, we compared the recent results to those obtained with the previous version of the USGS (ver. 2) and UCB velocity models (Stidham et al., 1999) that had 250-m grid spacing and 1 km/s minimum S-wave velocity.

Results

Maximum horizontal velocity maps for the Loma Prieta simulations with the USGS SF06 velocity model and three different finite-source models are shown in Figure 11.1 (top). Results show strong NW directivity of the Wald et al. SM/T finite-source model. Amplification of the ground motions over the basins is well observed for all three models.

Recorded and synthetic waveforms at some of the stations that were included in the simulations are compared in Figure 11.1 (bottom). Velocity waveforms were lowpass filtered at 0.8 Hz. The traces from top to bottom are observations (gray), followed by the USGS SF06 velocity model synthetics for the three different finite-source models (Beroza, Wald SM/T, Wald SM). The results showed that the overall agreement between the USGS SF06 model simulations and the observations is very good. Comparison to the results obtained with the previous USGS (ver. 2) model and the UCB model showed that the results obtained with the new USGS SF06 model better matched the observations. Important differences between the results obtained with the three different finite-source models can also be observed. The results obtained for the Loma Prieta event as well as for other smaller events in the region (Dolenc et al., 2006; Rodgers et al., 2006) indicate that the remaining misfit between the simulations and the observations at some stations requires further model refinement.

Figure 11.1: Top: Maximum horizontal velocity maps for the Loma Prieta simulations with the USGS SF06 velocity model and three different finite-source models. Left: Beroza (1991) strong motion model; Center: Wald et al. (1991) strong motion/teleseismic model; Right: Wald et al. (1991) strong motion model. Bottom: Recorded and synthetic waveforms at some of the stations that were included in the simulations. Velocity waveforms were lowpass filtered at 0.8 Hz. The traces from top to bottom are observations (gray), and USGS SF06 velocity model synthetics for the three finite-source models (Beroza, Wald SM/T, Wald SM).

References

Beroza G. C. (1991). Near-source modeling of the Loma Prieta earthquake: Evidence for heterogeneous slip and implications for earthquake hazard, Bull. Seism. Soc. Am. 81, 1603-1621.

Brocher, T. M., Compressional and shear wave velocity versus depth in the San Francisco Bay Area, California: Rules for USGS Bay Area Velocity Model 05.0.0, U.S. Geol. Surv. Open-File Report 2005-1317, 2005.

Dolenc, D., D. Dreger, and S. Larsen, 3D simulations of ground motions in Northern California using USGS SF06 velocity model, Seism. Res. Lett., 77, 300, 2006.

Larsen S. and C. A. Schultz (1995). ELAS3D: 2D/3D elastic finite-difference wave propagation code, Technical Report No. UCRL-MA-121792, 19 pp.

Rodgers, A., A. Petersson, S. Nilsson, B. Sjogreen, and K. McCandless, Ground predictions using the USGS seismic velocity model of the San Francisco Bay Area: Evaluating the model and scenario earthquake predictions, Seism. Res. Lett., 77, 281, 2006.

Stidham, C., M. Antolik, D. Dreger, S. Larsen, and B. Romanowicz, Three-dimensional structure influences on the strong-motion wavefield of the 1989 Loma Prieta earthquake, Bull. Seism. Soc. Am., 89, 1184-1202, 1999.

Wald D. J., D. V. Helmberger, and T. H. Heaton (1991). Rupture model of the 1989 Loma Prieta earthquake from the inversion of strong-motion and broadband teleseismic data, Bull. Seism. Soc. Am. 81, 1540-1572.