Reconciling Bias in Moderate Magnitude Earthquake Ground Motions Predicted by Numerical Simulations

Sajan K C, Chukwuebuka C. Nweke, Jonathan P. Stewart, & Robert W. Graves

Accepted March 16, 2023, SCEC Contribution #12717

Earthquake simulation platforms are useful for predicting ground motion in areas where data is scarce. Recent studies have used simulations to investigate ground motions associated with historical and/or expected high-impact earthquakes. Simulations are gaining attention as a resource for engineering design, but their accuracy needs to be validated before they can be fully embraced. One approach to validate simulated ground motions is to simulate earthquakes with well-recorded data and compared to identify any bias. A study by Nweke et al. (2022) conducted simulations for 13 moderate magnitude earthquakes in southern California and found a positive bias in ground motion levels relative to the observed data. This current study aims to understand the source of this bias by evaluating ground motion residuals and exploring its dependence on parameters like magnitude, rupture distance, and source depth. Currently, additional events are being analyzed to see if the bias persists. The goal of this work is to develop a refined simulation methodology that is better able to capture the absolute levels of ground motions observed for earthquakes, thus building additional confidence in the application of the simulation approach for high impact scenarios.

Citation
K C, S., Nweke, C. C., Stewart, J. P., & Graves, R. W. (2023, 03). Reconciling Bias in Moderate Magnitude Earthquake Ground Motions Predicted by Numerical Simulations. Poster Presentation at Lamont-Doherty Seismology Student Workshop.

Related Projects & Working Groups
Reconciling Bias in Moderate Magnitude Earthquake Ground Motions Predicted by Cybershake Simulations, Ground Motion Simulation Validation TAG