By How Much Can Physics-Based Earthquake Simulations Reduce the Uncertainties in Ground Motion Predictions?

Abstract:

Probabilistic seismic hazard analysis (PSHA) is the scientific basis for many engineering and social applications: performance-based design, seismic retrofitting, resilience engineering, insurance-rate setting, disaster preparation, emergency response, and public education. The uncertainties in PSHA predictions can be expressed as an aleatory variability that describes the randomness of the earthquake system, conditional on a system representation, and an epistemic uncertainty that characterizes errors in the system representation. Standard PSHA models use empirical ground motion prediction equations (GMPEs) that have a high aleatory variability, primarily because they do not account for the effects of crustal heterogeneities, which scatter seismic wavefields and cause local amplifications in strong ground motions that can exceed an order of magnitude. We show how much this variance can be lowered by simulating seismic wave propagation through 3D crustal models derived from waveform tomography. Our basic analysis tool is the new technique of averaging-based factorization (ABF), which uses a well-specified seismological hierarchy to decompose exactly and uniquely the logarithmic excitation functional into a series of uncorrelated terms that include unbiased averages of the site, path, hypocenter, and source-complexity effects (Feng & Jordan, Bull. Seismol. Soc. Am., 2014, doi:10.1785/0120130263). We apply ABF to characterize the differences in ground motion predictions between the standard GMPEs employed by the National Seismic Hazard Maps and the simulation-based CyberShake hazard model of the Southern California Earthquake Center. The ABF analysis indicates that, at low seismic frequencies (< 1 Hz), CyberShake site and path effects unexplained by the GMPEs account 40-50% of total residual variance. Therefore, accurate earthquake simulations have the potential for reducing the aleatory variance of the strong-motion predictions by about a factor of two, which would lower exceedance probabilities at high hazard levels by more than order of magnitude. Realizing this gain in forecasting probability would have a substantial impact on the prioritization and economic costs of earthquake risk-reduction strategies.