Bay area quake fails to fit textbook model.

Bay area quake fails to fit textbook model

In the wake of last week's earthquake, experts on the San Andreas fault find themselves alternately patting each other's backs and scratching their heads. The quare confirmed several of their predictions, but left scientists with many surprising details.

Records from distant seismic stations led the U.S. Geological Survey (USGS) this week to upgrade the quake magnitude from 6.9 to 7.1.

Last week's temblor ruptured a section of the San Andreas that runs through mountains east of Santa Cruz. Scientists named it the Loma Prieta quake for a nearby mountain peak. Historical records suggest this segment fractured during a jolt in 1865 as well as during the great San Francisco quake of 1906, and several signs indicated it was again near breaking. A 1988 report issued by the USGS calculated a 30 percent chance that a magnitude 6.5 shock would hit this fault section within 30 years.

"It worked out so well, almost unbelievably well," says Allan G. Lindh, a USGS seismologist in Menlo Park, Calif., who has long discussed the quake potential of the fault near Santa Cruz. But Lindh worries, "This will leave a lot of people feeling that we know more than we do."

As scientists study the Loma Prieta shock, the cause of more than 60 confirmed deaths, an unusual portrait is emerging. Searches above the quake center have failed to find any sign the jolt ruptured the surface along the San Andreas fault. During a temblor this size, opposite sides of the San Andreas normally slip apart by a meter or two, and geologists usually find evidence of the displacement at the surface.

Geologists have found cracks in the ground, but none that match the way the

fault moved underground, says USGS geologist David P. Schwartz. The San Andreas forms the border between the northwest-moving Pacific plate and the southeast-heading North American plate. Because the plate motion drives earthquakes along the San Andreas, the ocean side of the fault always moves to the northwest, relative to the continental side. For a person facing the fault, the other side should always slip to the right. Schwartz says investigators have not located such offsets at the surface.

What geologists seek may not exist. After the 1906 quake, researchers could not find the expected surface ruptures along this segment, and perhaps it does not normally exhibit the ground offsets that appear on most other sections of the San Andreas, Schwartz says.

Seismologists, who analyze earthquake waves traveling through the ground, may have an explanation for the lack of ground rupture. The main shock struck about 18 kilometers down, unusually deep for San Andreas earthquakes, which normally occur about 10 km below the surface. Lindh says. The rupture spread upward, but apparently did not break the surface. By measuring changes in the distance between reference stations, geophysicists have detected another unusual characteristic of this fault section. William H. Prescott of the USGS reports the Loma Prieta quake produced about 1.5 meters of horizontal slip and about a meter in the vertical direction. This is the first time that scientists have seen significant vertical movement on any section of the San Andreas. They say a local bend in the fault may explain the vertical slip.

Kerry E. Sieh, a geologist who studies rock displacements and other physical evidence of past earthquakes along faults, finds the absence of surface rupture troubling. Geologists often rely on the rupture history of a fault segment to assess the chances of a future shock. Sieh, of the California Institute of Technology in Pasadena, says if some damaging quakes do not manifest themselves at the surface, that could cause underestimates of the seismic danger.

Other scientists sound less concerned. When studying past quakes, geologists usually pick areas where the geology is less complex than the Santa Cruz mountains. These sports probably provide a true picture of the earthquake history, Linch says.