Long-Period Ground-Motion Simulations of the Mw 7.2 El Mayor-Cucapah Earthquake

The 4 April 2010 Mw 7.2 El Mayor-Cucapah earthquake generated strong shaking that was recorded by over 200 strong motion instruments throughout northern Baja California and southern California. At longer periods (T > 2 sec), rupture directivity and basin response effects lead to amplified ground motions and extended shaking durations in the Imperial/Coachella, San Bernardino and Los Angeles regions. Using a suite of hypothetical finite-fault rupture models, we test the ability of long-period ground-motion simulations of scenario earthquakes to produce waveforms consistent with these recordings. The hypothetical ruptures are constructed using the methodology proposed by Graves and Pitarka (2010) and require, as inputs, only a general description of the fault location and geometry, event magnitude and hypocenter. For each rupture model, simulations are run for two Southern California Earthquake Center (SCEC) 3D community seismic velocity models (CVM-4m and CVM-H62). While the details of the motions vary across the simulations, the median levels match the observed peak ground velocities reasonably well with the standard deviation of the residuals generally within 50% of the median. For the non-basin regions, simulations with the CVM-H62 model perform significantly better than those of CVM-4m, which we attribute to the inclusion of the Tape et al. (2009) tomographic updates within the background crustal velocity structure of CVM-H62. Within the greater Los Angeles region, the CVM-4m model generally matches the level of observed motions whereas the CVM-H62 model over-predicts the motions in the southernmost portion of the basin. Animations of the simulated wave fields demonstrate this over-prediction is created by the sharp impedance contrast along the southern margin of the Los Angeles basin in the CVM-H62 model. For both seismic velocity models, the variance in the peak velocity residuals is lowest for a rupture that has significant shallow slip (less than 5 km depth), whereas the variance in the residuals is greatest for ruptures with large asperities below 10 km depth. These results are consistent with recent source inversion studies indicating this earthquake was dominated by shallow (less than 10 km depth) moment release (e.g., Wei et al, 2010).