December 30, 2004
By Edie Lau -- Bee Science Writer
Sacramento Bee


University of California, Davis, scientists venturing into the controversial field of earthquake prediction say they successfully forecast Sunday's monstrous temblor in Indonesia.

Twelve days before the magnitude 9 earthquake struck, spawning deadly tsunamis across the Indian Ocean, geologist Donald Turcotte unveiled at a science meeting a world map identifying locations where he and his colleagues anticipate large quakes will occur this decade.

Among the places highlighted with a bright orange dot was the northern tip of Indonesia.

"For all intents and purposes ... we were there," Turcotte said in an interview Wednesday.

Produced in late November, the forecast map showing earthquake "hot spots" elicited no dramatic reactions when it was presented Dec. 14 at a meeting of the American Geophysical Union in San Francisco.

That's partly because disaster managers can't do much with predictions of big earthquakes happening sometime over a 10-year period.

And it's partly because mainstream science views earthquake predictions with a skeptical eye.

"For an earthquake prediction, you want to have the size (of the quake) and a range of time and a range of locations that are very narrow," said Dave Wald, a seismologist at the U.S. Geological Survey's National Earthquake Information Center in Colorado.

"If it's not narrow, if you say (a quake will hit somewhere in) the whole world, in any year, of any magnitude, you'll be right."

John Rundle, director of the UC Davis Center for Computational Science and Engineering and leader of the earthquake forecast group, is well aware that the method is a long way from producing a usable short-term prediction, but he argues that the information is nonetheless useful.

"Any kind of information is valuable and allows you to improve your planning," said Rundle, a physicist by training.

The forecast, produced through computer modeling, is based upon the locations of 25 years of earthquakes around the world of magnitude 5 or larger.

That data gave researchers a map of the quake-prone regions of the globe. They then identified hot spots by seeking signs of seismic change in those regions.

"The hot spots are the places ... where the earthquake activity is changing most rapidly," Rundle said. "So there'll be a fault that's popping along, (producing) one magnitude 3 (quake) every week, and all of a sudden, it's putting out three magnitude 3s per week, or suddenly no magnitude 3s per week."

The result is a map identifying parts of the world that the scientists say are most likely to experience a magnitude 7 earthquake or larger between 2000 and 2010. The map was produced by graduate student James Holliday.

Some fellow scientists who saw the map during the meeting in San Francisco were critical, Turcotte said, arguing that "everybody knows (quakes are) going to occur in those areas."

Turcotte and Rundle counter that they've narrowed the areas of concern. In other words, not all of the seismically active parts of the planet are on their map.

"We're telling people where are the areas where they shouldn't have to be worried for the next 10 years about large earthquakes," Rundle said.

Added Turcotte: "We don't say we have done magic, but we do say we have improved the situation."

So far, the researchers say, the track record of their model is pretty good. Out of 38 magnitude 7 earthquakes or larger that have happened since 2000, they count 30 occurring in or within 125 miles of their hot spots.

(Although the map was not produced until late November, Rundle said the data upon which it is based do not include any earthquakes since 2000.)

In addition to the Indonesian earthquake, Rundle said one other large quake has happened inside a hot spot since the map's debut: a magnitude 8.1 temblor on Dec. 23 near Australia's Macquarie Island.

USGS seismologist Wald said the forecast would be better if it more precisely estimated the power of coming quakes.

For example, Wald said, to say that the model accurately forecast the Indonesian quake glosses over the fact that a magnitude 9 quake unleashes 1,000 times the energy of a magnitude 7 quake.

"A 7 can be devastating, but we know from the (Indonesian quake) that the devastation was not from the shaking but from the tsunami," he said. "That (strength of tsunami) wouldn't have happened from a 7."

Rundle said researchers did not ask the computer to forecast locations of magnitude 9 quakes, but they can. "We didn't ask the right question because this is basically very new," he said.

The team published its forecast methodology in 2002 in the journal Proceedings of the National Academy of Sciences in a study focusing on California earthquake hot spots. Their study applying the method to the entire planet has not been published.

The UC Davis team is the second in California this year to widely discuss its efforts to forecast earthquakes. UCLA seismologist Vladimir Keilis-Borok, working with an international team of scientists, predicted that a quake of magnitude 6.4 or larger would strike within a 12,000-square-mile region of Southern California sometime during the first nine months of 2004.

The prediction did not come true.

http://www.sacbee.com/content/news/story/11895593p-12782505c.html