How Rare Earthquakes Impact Seismic Hazard Assessments

Abstract:

Seismic hazard models in California have evolved from simple segmented prescriptive models to much more complex representations of multi-fault and multi-segment earthquakes in an interconnected fault network. During the development of the 3^rd Uniform California Earthquake Rupture Forecast (UCERF3), the prevalence of multi-fault ruptures in the modeling was controversial. Yet recent earthquakes, for example, the 2016 M7.8 Kaikoura earthquake – have validated this approach. If anything, connectivity in the UCERF3 may be underestimated, although spatio-temporal clustering in the model may mitigate some modeling simplifications.

Future iterations of hazard models should continue to improve the modeling of the connectivity of the fault system and include the largest, rarest events. The goal should not be to predict the exact rupture process likely to happen in a future Kaikoura-like rupture, but to capture enough connectivity so that the modeled magnitude distribution – including the maximum magnitude – is accurate.