Time-Dependent Seismic Hazard in the San Francisco Bay Region, California

Abstract eng:
A key assumption in standard probabilistic seismic hazard analysis (PSHA) is that earthquake occurrence can be modeled as a Poisson process (time-independent). For but a few regions in the world, however, the timing information on past earthquakes are not sufficient to calculate time-dependent hazard. We have calculated timedependent probabilistic hazard for the San Francisco Bay region (SFBR) resulting from the region's seven major faults (e.g., San Andreas, Hayward-Rodgers Creek) using the range of models that were considered by the Working Group on California Earthquake Probabilities (WGCEP, 2003). Based on their results, there is an increasing probability of a large (moment magnitude [M] > 6.7) earthquake occurring in the SFBR in the period 2002 to 2031. The estimated probability in 2002 was 62%. WGCEP (2003) considered five probability models that take into account various degrees of physics, date of last rupture, recent seismicity rates, and slip in the 1906 M 7.9 San Francisco earthquake. The probability models, which are not all time-dependent, and their weights were used to compute the rates of characteristic events for each rupture source in the WGCEP (2003) model. These rates were then input into the PSHA code, along with the source geometries, characteristic magnitudes, and shape of the magnitude distributions. All other faults in the seismic source model are treated in a time-independent manner. The probabilistic hazard was calculated for the cities of San Francisco, Oakland, and San Jose to evaluate the sensitivities of the hazard in the SFBR to parameters of the WGCEP (2003) model. The difference between the time-dependent and time-independent hazard depends on the proximity to the major faults. For example, a site in San Francisco along the San Andreas fault shows little difference in the two types of hazard because of the recent occurrence of the 1906 earthquake. In contrast, a site in Oakland located adjacent to the Hayward fault shows a 5 to 10% higher time-dependent hazard because the average recurrence interval has nearly elapsed.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket