Prediction of Quake Gets Experts Talking

An international team of earthquake scientists said the 6.5-magnitude San Simeon earthquake fell within its broad prediction that a significant temblor would hit the region.

The prediction has caught the interest of other scientists prominent in the study of earthquakes, although they caution that more work is needed to verify that the technique works. They said it would take time to determine whether the group’s system for forecasting quakes made solid predictions or simply matched random coincidence.

In documents and verbal advisories submitted to other scientists, the team predicted last June that a quake of magnitude 6.4 or larger would occur within nine months somewhere along a 310-mile stretch of California extending from Fort Bragg in Mendocino County to Cambria in San Luis Obispo County.

The actual quake occurred a little more than six months later in December toward the southern end of that zone. It was centered six miles northeast of San Simeon.

The same group -- led by UCLA scientist Vladimir Keilis-Borok -- is now predicting a quake of at least magnitude 6.4 somewhere in a broad swath of Southern California’s desert by Sept. 5.

With areas as wide as these and such long times between the predictions and the possible quakes, many may feel that such information is not precise enough to be very useful. How many government agencies or private individuals will pay much attention if the predicted quake might be 300 miles away?

But the predictions are getting more precise, and scientists say that with more research they may be able to further narrow the parameters.

The team’s predictions are based on perceived “chains” of small earthquakes, magnitude 3.0 to 5.0, that are comparable to past patterns preceding larger quakes in the same area.

Keilis-Borok, 82, who came to the U.S. from Russia, has been researching quake prediction for more than 20 years.

His team forecast an earthquake in Japan of magnitude 7.0 or higher by this past Dec. 28, in a region that included Hokkaido island. A magnitude 8.1 quake did strike Sept. 25 off the coast of Hokkaido.

Previously, the UCLA scientist has related, his team made “intermediate-term” predictions years in advance. For instance, the 6.7 Northridge earthquake of 1994 occurred 21 days after the expiration of an 18-month prediction window that a quake of magnitude 6.6 would occur within 120 miles of the epicenter of the 7.3 Landers earthquake of 1992.

Keilis-Borok has declined interviews about his techniques in hopes of publishing an article in the journal Nature, which asks its authors to embargo statements on what they are writing about until the piece has appeared.

But last month he was quoted extensively in a UCLA news release, and he has published other articles on the subject.

“We look backward to make our earthquake predictions,” he said in the release. “First, we search for quickly formed long chains of small earthquakes. Each chain is our candidate to [be] a newly discovered short-term [quake] precursor.

“In the vicinity of each such chain,” he said, “we look backward and see its history over the preceding years -- whether our candidate was preceded by certain seismicity patterns. If yes, we accept the candidate as a short-term precursor and start a nine-month alarm. If not, we disregard this candidate.”

He has acknowledged that the team’s predictions have not always been right. He has told colleagues that he expects, for now, to be right 50% of the time.

Still, said John Vidale, acting director of the Institute of Geophysics and Planetary Physics at UCLA, “If their predictions are effective 50% of the time, that represents a vast improvement over what we’ve seen in the past.”

Lucy Jones, scientist in charge of the Pasadena office of the U.S. Geological Survey and a past skeptic about earthquake prediction, said she felt that the team’s successes must be recognized as significant. But she still viewed the predictions as “a test,” with more predictions necessary before firm conclusions could be drawn on the validity of the predictive technique.

“We want to know not only whether some quakes that are predicted do not occur, but also whether some unpredicted quakes do,” Jones said.

Caltech seismologist Egill Hauksson said it should be recognized that “some progress has been made,” but it still made him somewhat uncomfortable that scientists seemed to be moving toward some correct quake predictions without yet understanding the earthquake process, such as what exactly initiates a temblor.

Tom Jordan, director of the Southern California Earthquake Center, said the California Earthquake Potential Evaluation Council, a body headed by an official of the California Geological Survey, will meet later this month to formally evaluate the Keilis-Borok predictions thus far.