70 Percent Chance That Large Earthquake Will Strike San Francisco By 2030

The report was compiled by a group of scientists known as a "working group" or "WG99," comprised of more than 70 geologists, seismologists, geophysicists and statisticians. The group is chaired by Dr. David Schwartz, a USGS geologist who heads that agency's San Francisco Bay Area Earthquake Hazards Project.

Schwartz pointed out that there is also an 80 percent chance of one or more magnitude 6 to 6.7 earthquakes before 2030. "While magnitude 6.7 or greater earthquakes can cause damage throughout the Bay region, these smaller quakes, centered in urbanized areas, could also cause serious damage."

Schwartz said an important difference between the new probabilities figures and previous earthquake estimates is that the earthquake hazard is spread broadly across the entire Bay region, from the Pacific coast to the Sacramento Delta, not just restricted to the those areas closest to the Bay.

"Since much of this broad region is becoming heavily urbanized and many of these faults run right through urban areas, future earthquakes have the potential to cause much more damage than the 1989 Loma Prieta earthquake," Schwartz said. "For example, there is a 30 percent chance of a damaging earthquake on one of the four newly characterized faults along and east of the of the rapidly growing I-680 corridor."

Coastal areas from Half Moon Bay to Monterrey are also included in the new probabilities study, with a 25 percent chance of one or more earthquakes centered on the nearby San Andreas or San Gregorio faults. "These areas are also experiencing rapid growth," said Schwartz, "and thus increasing the odds for damage from earthquakes."

While the new study concludes that there is a 70 percent chance (plus or minus 10 percent) for at least one large earthquake across the region in the next 30 years, it also lists the odds for earthquakes on individual faults in the San Francisco Bay area, with those odds having an uncertainty factor of 5-10 percent. Those faults are:

the Hayward-Rodgers Creek fault that runs from Fremont to Santa Rosa, with a 32 percent probability of a 6.7 or larger earthquake, has the highest odds for any Bay area fault. a 21 percent probability of a 6.7 or larger earthquake occurring on the San Andreas fault, from San Jose to north of San Francisco. an 18 percent probability of a 6.7 or larger earthquake occurring on the Calaveras fault, which runs from Gilroy on the south to Mount Diablo. a 10 percent probability of a 6.7 or larger earthquake occurring on the San Gregorio fault, which runs offshore, from Monterey on the south to the Golden Gate. a six percent probability of a 6.7 or larger earthquake occurring on the Greenville fault, which runs through the East Bay hills from south of Livermore to east of Mt. Diablo, and also a six percent probability for the Concord-Green Valley fault which runs northward from Walnut Creek, across Suisun Bay to an area east of Napa. a four percent probability of a 6.7 or larger earthquake occurring on the Mt. Diablo thrust fault. This "blind thrust" fault under the landmark East Bay mountain, is similar to the fault that caused the 1994 Northridge earthquake. Schwartz explained that the new probability figures were derived by balancing two processes--the motions of the plates that make up the Earth's outer shell and the slip on faults, which occurs primarily during earthquakes. "The continual northwestward motion of the Pacific Plate past the North American Plate loads strain onto the network of active faults that slice through the region, Schwartz said, "and earthquakes sporadically release and redistribute this strain."

He said that development of the Global Positioning System, which uses satellites to locate points precisely on the earth's surface, has allowed geophysicists to make accurate measurements of how the current plate motion distributes strain onto individual faults. That plate motion is currently about 1.5 inches per year in the Bay area.

New geologic studies, such as digging trenches parallel to and across faults, also contributed to the new probability report. On the San Gregorio fault near Moss Beach, for example, trenching revealed that a buried stream channel has been offset about 1,000 feet during the past 80,000 years, indicating a strain rate there of about one-sixth of an inch per year.

The other side of the probability balance is the slip, or movement on faults that can occur either during an earthquake or during slow, aseismic (without large earthquakes) creep. On the Hayward fault, for example, creep is slowly offsetting street curbs, even in the absence of large earthquakes. Where aseismic creep occurs, it affects the balance between plate motions and earthquakes by relieving the strain, which can either change the likelihood or lower the magnitude of future earthquakes.

Trenching that revealed strain on the San Gregorio fault also revealed the past history of tremors on several other Bay Area faults. Soil layers in a trench in El Cerrito, for example, showed that four to seven large earthquakes had occurred on the Hayward fault during the past 2,200 years. A "large" earthquake is one of magnitude 6.7 or greater.

In addition to identifying fault segments, WG99 looked for previously unknown faults. Most faults in the Bay Area have "strike-slip" motion, in which the two sides of the fault slip horizontally past each other. In contrast, ramp-like "thrust" faults have vertical motion and often do not reach the Earth's surface, making them difficult to find. The importance of locating these hidden faults was underscored by the devastating 1994 Northridge earthquake in southern California, which occurred on a previously unknown thrust fault. Of the known faults in the Bay Area, only the Mount Diablo thrust fault has a high enough slip rate to be included in the WG99's calculations, according to Schwartz.

In calculating the probabilities of earthquakes occurring during the next 30 years, members of the WG99 developed a new set of computer models that use both physics and statistics. In these models, earthquakes are caused by a combination of constant plate motions and a random process that accounts for variations in earthquake sizes and occurrence, a scenario that closely mimics the occurrence of earthquakes around the world.

Because every earthquake changes strain on nearby faults, another important element in the WG99 method is the inclusion of interactions of faults in the San Francisco Bay region. For example, the rate of large quakes in the region was high in the late 1800s, but abruptly dropped after the 1906 earthquake, probably because the San Andreas fault slipped so much over such a great length in 1906 that the strain was reduced on most faults throughout the Bay region. Strain on those faults, however, has been slowly building up again, the scientists said, and strong quakes began to occur again in the 1980s, although not yet reaching the level of the late 1800s.

In order to assign probability figures to the various faults and fault segments, WG99 assigned weights to the various choices so that all of them were included in their overall calculations. This method helps ensure that the WG99 probabilities are reliable estimates of the earthquake threat faced by the San Francisco Bay region between 2000 and 2030. The WG99 method also allows the scientists to update these probabilities as new insights are gained.

In releasing the new probability figures, Schwartz and other members of WG99 stressed that the purpose of their research and report is to inform Bay Area residents of the earthquake risk so that they can make informed decisions about their personal risk and make preparations to survive any of the earthquakes that might strike the region. "The whole purpose of our research is to use science to contribute to safer communities," said Schwartz, " and although we can't predict or prevent earthquakes, we all can prepare for them." The work of USGS and other scientists is part of an ongoing program of the National Earthquake Hazard Reduction Program.

In addition to explaining the science of forecasting earthquakes, the WG99 report and accompanying fact sheet advise citizens of agencies such as the California Governor's Office of Emergency Services (OES), the Red Cross of California, the Federal Emergency Management Agency (FEMA) and others that have pamphlets and Web sites that contain emergency preparedness information. The information is also available on the USGS Web site http://quake.wr.usgs.gov, or by phone at 650-329-4085.