Report to the Director, Governor’s Office of Emergency Services
By the California Earthquake Prediction Evaluation Council
December 9, 2004
The Governor’s Office of Emergency Services requested that the California Earthquake Prediction Evaluation Council meet to evaluate an earthquake ”alarm “Alarm” is the term used by the Keilis-Borok research group. Although it is not defined, it appears to refer to a long-term (nine months) prediction or forecast.” issued by Dr. Vladimir Keilis-Borok and colleagues as part of their experiment in short-term earthquake prediction. The Council met on December 2, 2004.
The prediction is for a magnitude 6.4 or greater earthquake to occur on or before August 14, 2005, within a 12,660 sq. miles area of southern California that includes portions of the eastern Mojave Desert, Los Angeles County, Riverside County and Orange County (Figure 1).
The area of the southern California prediction includes a number of very active faults, including the San Bernardino and (a portion of) Mojave segments of the San Andreas fault, the northern portions of the San Jacinto and Elsinore faults, and the Newport-Inglewood fault. The area is one of the most seismically active in the state. It includes the recent Landers (M7.3) and Hector Mine (M7.1) earthquakes, which continue to have significant aftershock activity. There were 4 earthquakes with M≥6.4 in the area of the southern California prediction during the last 72 years. The probability of a M≥6.4 earthquake occurring in a random 9-month period is thus estimated to be about 4%.
As discussed in the report of our March 2, 2004 meeting to evaluate an earlier Keilis-Borok prediction for southern California, this group of researchers had previously issued two apparently successful predictions, for the 2003 Hokkaido and San Simeon California earthquakes. However, neither of these forecasts was brought before the Council and the Japanese earthquake was judged to have had a high likelihood of occurring by random chance. Since then, the group has issued three subsequent predictions. The previous prediction for southern California, for a magnitude 6.4 or greater earthquake before September 5, 2004 (the subject of our March meeting) was not fulfilled. In addition, predictions for a large earthquake in Japan and a moderate earthquake in the area of Slovenia were not fulfilled. Thus, to date, the Keilis-Borok group has predicted 5 earthquakes with the current methodology; two of the predicted events have occurred and three have not.
The Keilis-Borok method is based on identifying patterns of small earthquakes prior to large shocks. To date there is no evidence that these, or related methods, yield useful intermediate term forecasts. Furthermore, the Council has heard no valid physical basis for the Keilis Borok methodology. Given the track record so far and the lack of a physical basis, the Council does not consider the method to be a basis for public policy. Therefore, CEPEC advises OES to take no special public policy actions based on this, or similar predictions.
Nevertheless, Californians need to remember that there are significant seismic hazards throughout the state. Regardless of the validity of the prediction, CEPEC recommends that all jurisdictions review and periodically exercise existing preparedness and response plans. Likewise, citizens of California who live in areas of high seismic hazard should make sure they have undertaken all general preparedness actions recommended by emergency management organizations and the Red Cross.