Berkeley - Pushing the limits of today's techniques for monitoring earthquake fault activity, a geophysicist at the University of California, Berkeley, assessed movement along the northern Hayward fault and found less chance of a major quake originating on that segment than previously thought.

With the help of radar interferometry and data from global positioning satellites (GPS), plus analysis of repeating microquakes 6 miles below the surface, he and his colleagues concluded that the deep portions of the fault steadily slip at about the same rate as the surface does. This means the rocks deep below the surface aren't locked and building up strain that could be released in a catastrophic quake.

"Our research shows no evidence of locking at any depth, which means the threat from one of our worst hazards, right in our backyard, is much reduced," said Roland Bürgmann, assistant professor of geology and geophysics at UC Berkeley. "However, other hazards - from the southern Hayward fault, the San Andreas fault and other nearby faults - leave the need to build reinforced homes and the need to be prepared just as high as before."

Bürgmann and his colleagues at UC Berkeley, the Lawrence Berkeley National Laboratory, the Jet Propulsion Laboratory in Pasadena, Calif., and UC Davis report their findings in the Aug. 18 issue of Science magazine.

The Hayward fault, considered one of the most dangerous faults in California, stretches more than 60 miles from San Pablo Bay in the north to below Fremont in the south, and is a branch of the more famous San Andreas fault that extends much of the length of California. Last year a state-wide team of seismologists estimated a 32 percent chance of a major quake originating somewhere on the Hayward fault in the next 30 years. A major quake is one of magnitude 6.7 or greater.

The segment of the Hayward fault from San Pablo Bay south to the border between Berkeley and Oakland is referred to as the northern Hayward fault, which may connect under the bay with the Rogers Creek fault that runs through Napa County. Until recently, the northern Hayward fault also was ranked high in terms of the chance of a major quake. The latest assessment, that of the U.S. Geological Survey Working Group on California Earthquake Probabilities issued last October, lowered this risk, in part based on preliminary findings supplied by Bürgmann's team.

Bürgmann set out several years ago to clarify the confusing history of earthquake activity along the northern Hayward fault. If, as trenching evidence suggests, the northern segment was the site of a major quake sometime between the mid-1600s and the arrival of Spanish colonists in 1776, why hasn't another quake occurred since then, Bürgmann wondered. Perhaps, he thought, the fault slips freely and large quakes do not occur on the northern segment.

"We know the Hayward fault creeps at about 5 millimeters per year at the surface, but we don't know how deep this creep goes," Bürgmann said. "We decided to use all the data that exists to try to say how deep the creep goes, and whether the fault is locked at depth."

The techniques Bürgmann used to study activity along the fault have just recently become available. Only within the past few years has interferometric synthetic aperture radar (InSAR) from satellites been used to measure ground motion along faults. Thanks to detailed mathematical analysis, it is possible to determine the surface displacement that has occurred between successive orbits of the satellite, even when the orbits are years apart. With data taken in 1992 and 1997 by a pair of European satellites, ERS-1 and ERS-2, plus analysis software developed at JPL, Bürgmann was able to determine the surface creep within a few millimeters along the northern Hayward fault.

"The global coverage of the European radar satellites allows the same interferometry technique used in this study to be applied to active faults in other parts of the world," said co-author Eric Fielding, a JPL geophysicist. "There are few places in the world that have the detailed ground information that was available for this study, but radar satellites image nearly everywhere. This allows us to study active faults in regions such as Turkey, Iran and Tibet to learn more about how faults behave. Because faults may behave differently at different times, it is important to look at a wide variety of faults to understand all of the possible types of behavior."

As a check on these measurements, Bürgmann took advantage of regional GPS stations that have been in place for nearly a decade. Data from the GPS network supply only regional slip rates, however. The GPS stations are not close enough to the northern Hayward fault to give precise slip rates for that segment.

In addition, seismologists at UC Berkeley and LBNL have just recently discovered that repeating microquakes - quakes too small to be felt but indicative of small patches of the fault suddenly slipping deep underground - can reveal the amount of movement below the surface. This technique was calibrated at a study site on the San Andreas fault near Parkfield, 165 miles south of San Francisco, by Robert Nadeau, a researcher in the Berkeley Seismological Laboratory, and Thomas McEvilly, a professor emeritus of geology at UC Berkeley. Both are members of the Earth Sciences Division at LBNL.

"They found that some of these microquakes were occurring at exactly the same spot, and that the microquake clusters could be used to infer how fast the fault is creeping near these stuck fault patches deep underground," Bürgmann said. "We found clusters of repeating microquakes as deep as 6 miles under Berkeley, which is evidence of structural creep far below the surface."

Putting all this information together, he estimated that the northern Hayward fault slips underground at a rate of about 5 to 7 millimeters per year, essentially the same rate as at the surface. The similar rates indicate that the fault is slipping freely without locking, he said.

Over long periods, and counting the slippage that occurs during and after earthquakes, the entire Hayward fault moves on average about 10 millimeters per year. The northern segment moves less than this because it is pinned by the southern segment, which is locked. In fact, though the entire fault moves at about 10 millimeters per year, surface creep along the southern segment is only 5 millimeters per year, which means strain builds up that can only be released in an earthquake.

Most faults outside California do not slip freely, but lock at depth. Bürgmann said what may allow the northern Hayward fault - and some other state faults - to move freely is a greenish rock that underlies much of central and northern California and could serve as a lubricant: serpentinite, often called serpentine. Serpentinite, the official state rock, is soft and fractures easily.

Bürgmann hopes to continue his studies of the Hayward fault and other faults that underlie the area.

"We can use these same techniques to measure the strain built up in the faults in surrounding regions, making the Bay Area a natural laboratory for the study of earthquake faults," Bürgmann said.

Co-authors of the paper with Bürgmann, Nadeau and McEvilly are Eric Fielding of JPL, graduate student David Schmidt, M. D'Alessio and Mark Murray, all of the Berkeley Seismological Laboratory, and D. Manaker of UC Davis.

The work was supported by the National Science Foundation, the Solid Earth and Natural Hazards program of the National Aeronautics and Space Administration, and the U.S. Geological Survey's NEHRP program.