WASHINGTON, April 3 -- A U.S. scientist reports new evidence that at least some earthquakes along some faults can be predicted up to five days before they occur. The approach, which predicts earthquakes by measuring exceptionally slight creeping of the Earth's crust along the fault line, is highly specific to parts of the central California San Andreas fault. The study's author and other geologists caution that the effect is never definite and only occurs at times of high seismic activity. 'It's a step forward. This is an unusual section of the San Andreas fault,' says author Clifford Thurber, professor of geology and geophysics at the University of Wisconsin-Madison. He cautions, 'There are a lot of ifs.' Potentially, he said, this could be very significant to large quakes, but it might be limited to small earthquakes on certain types of faults. 'We just don't know.' Other geologists are less confident. 'It's not conclusive,' said Jim Savage, a geologist with the U.S. Geological Service in Menlo Park, Calif. 'He has good evidence that something is happening, but it seems to only occur in special consequences,' Savage said of Thurber's research. Thurber found that shallow movements, called 'creep,' occurred before more than half of the earthquakes that struck between 1970 and 1994 on a particular section of the San Andreas fault during periods of increased seismic activity. Thurber's results, published in today's issue of Nature, refer to the fault from San Jan Batutista to the Pinnacles National monument in Central California.
This section is unusual because much of its displacement occurs as creep. Creep, when two sides of a fault 'slide' across each other, occurs at a very slow rate of a few inches a day. 'What I'm optimistic about' Thurber said, is using this effect to predict the next earthquake in Parkfield, California. 'It's still expected they they'll be a magnitude 6 there.' He noted that there 'was something, probably a creep event' nine hours before the 1986 Parkfield quake. 'An irrigation pipe broke, so it (the fault) started to move before.' Savage, however, is doubtful. I'd like to see an earthquake predecessor, he said. 'Looking at the statistics, anyone really would think there's something there, but when you go to the standard proof, it's not completely compelling. It's not conclusive.' The data was collected by a network of creepmeters. These seismic trip wires instantly detect even the smallest movement of the Earth's crust. Positioned along a relatively quiet section of the San Andreas fault, the creepmeters are operated by the U.S. Geological Survey. Scientists have collected creep data for more than 25 years, but according the Thurber, 'These examples of creep, happening before small to moderate earthquakes, stand in sharp contrast,' Thurber said, to previous results. 'Seismologists are trying to understand how earthquakes start -- what kicks them out and gets them running,' Thurber said. 'It really means we have to keep an eye on this.' (Written by Aries Keck, from Durham, North Carolina)
