HR: 0830h
AN: S41C-0089    [PDF]
TI: Rationalizing Hybrid Earthquake Probabilities
AU: * Gomberg, J
EM: gomberg@usgs.gov
AF: U.S. Geological Survey, 3876 Central Ave., Suite 2, Memphis, TN 38152
AU: Reasenberg, P
EM: reasen@usgs.gov
AF: U.S. Geological Survey, MS 977, 345 Middlefield Rd, Menlo Park, CA 94025
AU: Beeler, N
EM: nbeeler@usgs.gov
AF: U.S. Geological Survey, MS 977, 345 Middlefield Rd, Menlo Park, CA 94025
AU: Cocco, M
EM: cocco@ingv.it
AF: Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Rome, 00143 Italy
AU: Belardinelli, M
EM: elina@ibogfs.df.unibo.it
AF: Universit… di Bologna, Viale Berti-Pichat 8, Bologna, 40127 Italy
AB: An approach to including stress transfer and frictional effects in estimates of the probability of failure of a single fault affected by a nearby earthquake has been suggested in Stein et al. (1997). This `hybrid' approach combines conditional probabilities, which depend on the time elapsed since the last earthquake on the affected fault, with Poissonian probabilities that account for friction and depend only on the time since the perturbing earthquake. The latter are based on the seismicity rate change model developed by Dieterich (1994) to explain the temporal behavior of aftershock sequences in terms of rate-state frictional processes. The model assumes an infinite population of nucleation sites that are near failure at the time of the perturbing earthquake. In the hybrid approach, assuming the Dieterich model can lead to significant transient increases in failure probability. We explore some of the implications of applying the Dieterich model to a single fault and its impact on the hybrid probabilities. We present two interpretations that we believe can rationalize the use of the hybrid approach. In the first, a statistical distribution representing uncertainties in elapsed and/or mean recurrence time on the fault serves as a proxy for Dieterich's population of nucleation sites. In the second, we imagine a population of nucleation patches distributed over the fault with a distribution of maturities. In both cases we find that the probability depends on the time since the last earthquake. In particular, the size of the transient probability increase may only be significant for faults already close to failure. Neglecting the maturity of a fault may lead to overestimated rate and probability increases.
DE: 7209 Earthquake dynamics and mechanics
DE: 7223 Seismic hazard assessment and prediction
DE: 7260 Theory and modeling
DE: 7299 General or miscellaneous
SC: Seismology [S]
MN: 2003 Fall Meeting