FAULT-SOURCE MODELING FOR THE 2014 UPDATE OF THE USGS NATIONAL SEISMIC HAZARD MAPS

The U.S. Geological Survey released an updated probabilistic seismic hazard model for the lower 48 states in July 2014. The model integrates current geologic, paleoseismologic, geodetic, and seismologic data to reduce risk from earthquakes and improve public safety through its application in engineering design. The fault-source component of the model in the Western United States is defined by a variety of geologic data and, in this update of the model, incorporates geodetic models. The primary components of the fault-source model are one or more generalized three-dimensional representations and either mean slip rate or return time for each source. The resulting predicted magnitude-frequency distribution of future earthquakes on each fault source is determined the over a range of earthquake sizes that could result in damage.

Although new geologic and paleoseismic data are critical in keeping the seismic source model current, the significance of individual studies in the model varies considerably. The most notable changes in hazard from the fault-source model in this update come from (1) including new sources not previously characterized, (2) major changes in slip rates or return times, and (3) major changes in fault-source geometry. In some locations where modest changes were made to the fault-source model, the overall seismic hazard, for the spectral frequency and probability of exceedance of interest, didn’t change significantly. Whether a specific fault results in a noticeable contribution to hazard depends on its rate of slip, the dimensions of the fault that constrain moment magnitude, the spectral frequency and probability of exceedance of interest, and if the fault has nearby neighbors. Low slip rate faults generally contribute only minor amounts of hazard.

A critical, and sometimes neglected, step in seismic-hazard assessment is validating the magnitude-frequency distribution for each source to ensure it resembles the known paleoseismic record. In order to accomplish that task, additional data needs include estimates of surface displacement during individual earthquakes at many locations as well as displacement variability at a site and recurrence model. These data can inform moment-magnitude estimates that we implement in the model, and thus, the frequency of future events of a given magnitude.