Insights on the Coefficient of Variability of Ground-Rupturing Earthquakes from the UCERF3 Time-Dependent Model

Abstract:

The relaxation of segmentation assumptions in fault-based ruptures of the third Uniform California Earthquake Rupture Forecast (USGER3) also necessitated the development of a new procedure for computing time-dependent, elastic-rebound-based probabilities . This new model, which was developed and calibrated using results from physics-based earthquake simulators, is much more self-consistent and less biased than those applied previously. It also has some implications that should be considered when interpreting paleoseismic observations. One is that unless faults are well segmented, there is no good reason to assume that a point on a fault (e.g. a paleoseismic trench) will exhibit a recurrence-interval distribution that is anything like that of traditionally assumed renewal models. We avoid such assumptions by changing the “probability of what” question in the new approach. The model also supports the application of magnitude-dependent aperiodicity (coefficient of variability), the need for which is implied by the results of physics-based simulators. That is, larger magnitude events are more periodic than smaller events, presumably due to the latter being more influenced by stress heterogeneities. These and other features will be discussed in the context of the new model and information used in its development.