OAKLAND, Calif. (AP) -- There is a 70 percent chance that a large

earthquake will shake the San Francisco Bay area in the next 30

years, according to a new federal study released today.


The new conclusion: There are a number of faults slicing through

heavily developed areas around San Francisco, more chances they

will rupture and more people who will be affected when they do.


The estimate from the U.S. Geological Survey is slightly greater

than that found by a 1990 study because it spreads the earthquake

hazard over a much broader region. It also lowers the bar for a

large earthquake.


The study lowered the definition of a large earthquake to

magnitude 6.7 to match the strength of the 1994 Northridge quake,

which killed 57 people and caused $20 billion in damage in Southern

California. The 1990 study had estimated a 67 percent chance of a

magnitude 7.0 earthquake by 2020.


"The whole purpose of our research is to use science to

contribute to safer communities," USGS geologist David Schwartz

said. "Although we can't predict or prevent earthquakes, we can

all prepare for them."


His team of 70 scientists developed a new set of computer models

that considered the interaction of faults.


The study looked at the earthquake hazard from the Pacific Ocean

to the Sacramento Delta about 40 miles inland. The area has seen

rapid development since 1989, when the Loma Prieta earthquake shook

the area with a 6.9 magnitude quake just as a World Series game in

San Francisco was about to start. That quake killed 67 people and

caused $6 billion in damage.


The highest odds for a specific fault line -- 32 percent -- was on

the Hayward-Rodgers Creek fault, which stretches from Fremont to

Santa Rosa, cutting through the heavily populated East Bay area.


The southern section of the fault last ruptured in 1868. The

northern section has not ruptured since between 1640 and 1776,

making it overdue for a major shakeup.


The San Andreas fault, which runs from San Jose to north of San

Francisco, had a 21 percent probability of a large quake.


The rate of large quakes was high in the late 1800s but dropped

after the catastrophic 1906 San Francisco earthquake, probably

because the San Andreas fault slipped so much that strain was

reduced over most faults in the region, the study found.


Strain on those faults has been slowly building up and strong

quakes began to occur in the 1980s, although not yet at the level

of a century ago.