HR: 08:30h
AN: S41D-03    [Abstracts]
TI: A RELM earthquake forecast based on pattern informatics
AU: * Holliday, J R
EM: holliday@cse.ucdavis.edu
AF: University of California, Center for Computational Science, Davis, CA 95616 United States
AU: Chen, C
EM: chenCC@cse.ucdavis.edu
AF: National Central University, Dept of Earth Science, Taipei, 320 Taiwan
AU: Tiampo, K
EM: kftiampo@cse.ucdavis.edu
AF: University of Western Ontario, Dept of Earth Science, London, ON N6A5B7 Canada
AU: Rundle, J B
EM: jbrundle@cse.ucdavis.edu
AF: University of California, Center for Computational Science, Davis, CA 95616 United States
AU: Turcotte, D L
EM: turcotte@geology.ucdavis.edu
AF: University of California, Dept of Geology, Davis, CA 95616 United States
AB: We describe a RELM forecast for future earthquakes in California that is primarily based on the pattern informatics (PI) method. The PI method, which identifies regions that have systematic and large fluctuations in seismic activity of the smallest events, has been demonstrated through an actual forecast to be successful. A PI forecast map originally published on February 19, 2002 for southern California has, as of this writing, successfully forecast the locations of 16 of 18 M > 5 events in the region during the past three and one half years. The method has also been successfully applied to Japan and on a worldwide basis. An alternative approach to earthquake forecasting is to use a Relative Intensity (RI) map, which is a map specifying the locations of the highest seismic activity of the smallest magnitude earthquakes. Recent advances in the PI method show considerable improvement, particularly when compared with the RI method using Relative Operating Characteristic (ROC) diagrams for sequences of binary forecasts. The RELM application requires a probability for each location for a number of magnitude bins over a five-year period. We have therefore constructed a hybrid forecast in which we combine the PI method with the RI method to compute a map of probabilities for events occurring at any location, rather than just the most probable locations. These probabilities are further converted, using the Gutenberg-Richter scaling law, to anticipate rates of future earthquakes that can be evaluated using the RELM test.
DE: 0550 Model verification and validation
DE: 3238 Prediction (3245, 4263)
DE: 4460 Pattern formation
SC: Seismology [S]
MN: Fall Meeting 2005