The new study modifies the assumption of the Working Group on California Earthquake Probabilities, who as recently as 1990 proposed that the odds are one-to-one that, in 30 years or less, an earthquake of magnitude 7 or larger will occur along the Hayward Fault. Such an event could cost thousands of lives and many millions of dollars.

In 1868 the southern Hayward Fault ruptured from Fremont to Berkeley in an earthquake estimated at magnitude 7.0 -- so destructive that, before the 1906 quake that destroyed much of San Francisco, it was known as "the Big One." Since 1868, the southern Hayward Fault has been firmly locked at depth and creeping only slowly at the surface.

The strain that builds up along major faults can be relieved by creep, which can't be felt but can be scientifically detected -- or, if the fault locks up until strain grows too great, by large earthquakes.

Geological evidence suggests that north of Berkeley the last Big One occurred much earlier, sometime between the mid-1600s and the arrival of Spanish colonists in the Bay Area in 1776. Whether or not the northern section of the fault is locked has much to do with when and where the next big quake is likely to occur and how powerful it is likely to be.

"If you add up the slippage of all the faults in the region, from the San Andreas in the west to the Greenville Fault on the edge of the Central Valley, there is a deficit in the amount the northern Hayward Fault is slipping," says Robert Nadeau of Berkeley Lab's Earth Sciences Division, a researcher in the Seismological Laboratory at the University of California at Berkeley.

To some seismologists, the deficit suggests that the Hayward Fault has stored an enormous amount of energy and is on the verge of catastrophic release.

"This is a 'worst case' estimate," says Nadeau, "based on a model that assumes the fault is creeping on the surface but locked at depth. But the data previously available couldn't discriminate between that model and one in which the fault is creeping all the way down."

To improve the data, Roland Burgmann, an assistant professor of geology at UC Berkeley, decided to integrate traditional measurements with space-based measurements such as interferometric synthetic aperture radar (InSAR) from the Jet Propulsion Laboratory.

Nadeau, along with Thomas McEvilly of Berkeley Lab's Earth Sciences Division, a Professor Emeritus in UC Berkeley's Department of Geology and Geophysics, offered to provide Burgmann with a different perspective on the problem -- one that opens a window on fault activity kilometers beneath the surface.

For many years McEvilly and Nadeau and their colleagues have studied the San Andreas Fault near Parkfield, a tiny ranching community 165 miles south of San Francisco. There they developed a technique of comparing the timing of identical repeating "microquakes" with slippage events deep underground, whose foci and energy could be accurately pinpointed by a network of seismometers at the bottom of boreholes.

Borehole data for the Hayward Fault wasn't yet available when Burgmann's study was being prepared, says Nadeau, "so we couldn't get the very rapidly repeating small quakes that would give us the best time resolution. Nevertheless larger repeating quakes recorded with surface seismometers can be used to infer slippage at depth, and they give us a picture that complements and supports the model of the fault resulting from this study."

On the northern Hayward Fault, the researchers identified three major clusters of identical repeating small quakes along the northern fault zone, representing centers of slippage up to 7 millimeters per year at from 4 to 10 kilometers depth.

When data from all sources was collated, a remarkably detailed picture of the Hayward Fault emerged. Shallow creep falls off markedly where the fault dives under San Francisco Bay at Point Pinole; creep is at a maximum at El Cerrito, with indications of rapid slippage continuing to great depths; surface creep diminishes south of Berkeley, over the locked portion of the fault that broke in 1868.

The model suggests that no locking occurs from Berkeley north to Point Pinole and that this freely creeping stretch of the fault is not likely to be the site of origin for a major quake. Prospects for the southern Hayward Fault, however, and for the Rodgers Creek fault, north of the Bay, are less sanguine.

If after this study the potential for a huge, near-term quake centered under the northern Hayward Fault seems less than before, the researchers caution that one need only remember the 62 lives lost and billions of dollars of damage caused by the magnitude 7.1 Loma Prieta quake of 1989, centered 60 miles from the major cities of the Bay, to realize the need for constant preparedness.

"Earthquake potential along the northern Hayward Fault, California," by R. Burgmann, D. Schmidt, R. M. Nadeau, M. d'Alessio, E. Fielding, D. Manaker, T. V. McEvilly, and M. H. Murray, appears in the August 18, 2000, issue of Science magazine.

The Berkeley Lab is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California.