Striking agreement of physics-based earthquake simulator and UCERF3 California seismic hazard model

Bruce E. Shaw, Kevin R. Milner, Edward H. Field, Keith B. Richards-Dinger, Jacquelyn J. Gilchrist, James H. Dieterich, & Thomas H. Jordan

Published August 9, 2017, SCEC Contribution #7458, 2017 SCEC Annual Meeting Poster #014

Probabilistic seismic hazard analysis (PSHA) provides important information to society, informing building codes and hazard mitigation efforts. The PSHA models currently applied in California combine fault-based earthquake rupture forecasts (ERFs) with ground motion prediction equations (GMPEs) to estimate the probability of exceeding a specified shaking intensity, such as peak ground acceleration (PGA). Empirical ERFs, however, rest on many uncertain assumptions and are difficult to test observationally, owing to the long recurrence times of large earthquakes. Physics-based earthquake rupture simulators are potentially helpful tools, but they also face a vast range of fundamental uncertainties. We use standard GMPEs to compare hazard estimates from the physics-based rupture simulator RSQSim against those from the state-of-the-art Uniform California Earthquake Rupture Forecast, version 3 (UCERF3). The input parameters to RSQSim are the UCERF3 fault model and slip rates, as well as the physical parameters that govern rupture nucleation, fault friction, and stress interactions in the simulator, which we determine by matching global earthquake scaling relations. We find strikingly good agreement between the hazard curves predicted by RSQSim and those derived from UCERF3. For example, at an exceedance probability of 2% in 50 years, the differences in PGA between the generic RSQSim model and the mean UCERF3 model are actually smaller on average than the differences between UCERF3 and UCERF2, its immediate predecessor. This remarkable agreement of the empirical and physics-based approaches, obtained without any local tuning, is an important cross-validation step that lends substantial support to basic aspects of the PSHA methodology. We examine a number of measures, including recurrence intervals, magnitudes dominating slip, and disaggregated hazard, along with features of the GMPEs, to assess the origins and implications of the hazard-estimate agreement.

Key Words
Seismic hazard, Earthquake simulator, UCERF3

Citation
Shaw, B. E., Milner, K. R., Field, E. H., Richards-Dinger, K. B., Gilchrist, J. J., Dieterich, J. H., & Jordan, T. H. (2017, 08). Striking agreement of physics-based earthquake simulator and UCERF3 California seismic hazard model. Poster Presentation at 2017 SCEC Annual Meeting.

Related Projects & Working Groups
Working Group on California Earthquake Probabilities (WGCEP)