HR: 11:05h
AN: S42B-03    [Abstracts]
TI: Model Uncertainty, Earthquake Hazard, and the WGCEP-2002 Forecast
AU: * Page, M T
EM: pagem@physics.ucsb.edu
AF: Dept. of Physcs, University of California, Santa Barbara, CA 93117 United States
AU: Carlson, J M
EM: carlson@physics.ucsb.edu
AF: Dept. of Physcs, University of California, Santa Barbara, CA 93117 United States
AB: Model uncertainty is prevalent in Probabilistic Seismic Hazard Analysis (PSHA) because the true mechanism generating risk is unknown. While it is well-understood how to incorporate parameter uncertainty in PSHA, model uncertainty is more difficult to incorporate due to the high degree of dependence between different earthquake-recurrence models. We find that the method used by the 2002 Working Group on California Earthquake Probabilities (WG02) to combine the probability distributions given by multiple models has several adverse effects on their result. In particular, taking a linear combination of the various models ignores issues of model dependence and leads to large uncertainties in the final hazard estimate. Furthermore, choosing model weights based on data can systematically bias the final probability distribution. The weighting scheme of the WG02 report also depends upon an arbitrary ordering of models. In addition to analyzing current statistical problems, we present alternative methods for rigorously incorporating model uncertainty into PSHA.
DE: 3245 Probabilistic forecasting (3238)
DE: 3275 Uncertainty quantification (1873)
DE: 6309 Decision making under uncertainty
DE: 7200 SEISMOLOGY
DE: 7223 Earthquake interaction, forecasting, and prediction (1217, 1242)
SC: Seismology [S]
MN: Fall Meeting 2005