The chairman of the USC geology department says he expects a magnitude 6.0 to 6.5 earthquake to occur in Central California by July 9, perhaps close to the Parkfield area, where such a quake has long been predicted.
If Charles G. Sammis is correct, it will mark a milestone in earthquake prediction. If he isn't, it will be the latest in a long line of failed prognostications, particularly at Parkfield, where seismologists have been expecting a large quake on the San Andreas fault since 1985 and have installed a major array of instruments to analyze it.
Sammis, aided by graduate geophysicist David D. Bowman, has plotted an upsurge of bursts of quakes above magnitude 3.0 in the Parkfield area since 1987 and compared it with a similar upsurge that preceded the 7.1 Loma Prieta quake in the Bay Area in 1989.
Based on Sammis' analysis of the intermediate quake patterns--which represent an accumulation of strain--he says a sizable Central California quake should occur during a 160-day period that began Feb. 1 and continues to July 9.
"When you see deterministic patterns of this kind, you would tend to believe there would be a quake by July 9," Sammis said in an interview.
The forecast, kept within the scientific community until now, is getting considerable attention from those most involved in the continuing but frustrating effort to predict earthquakes with reliability.
Tom Henyey, director of the Southern California Earthquake Center at USC, said this week that much of the current research in earthquake prediction involves close study of such precursory phenomena. But he said, "Dr. Sammis has gone further than most scientists in putting his views on the line about when an earthquake will actually occur."
Sammis said that because he cannot pinpoint where in Central California the quake will occur, he is technically making more of a general "forecast" than a precise prediction.
He said he has also plotted an increase in intermediate earthquakes in the Kommandorsky Islands at the end of the Aleutian chain and expects to see a large quake there within about a year.
Another form of this research is the "backpredicting" of earthquakes. This entails studying the seismicity leading up to large earthquakes in the past, such as Loma Prieta, calculating from the upsurge when the quake would be expected, and then examining when it actually hit to see if the theory is verified.
Sammis said backpredicting confirms that Loma Prieta occurred within the window in which it could have been expected.
The chairman of the USC geology department, Charles G. Sammis, says that a magnitude 6.0 to 6.5 earthquake may occur on or near the San Andreas Fault in Central California, perhaps near Parkfield, by July 9. But some previous forecasts of Parkfield-area quakes have not been realized.