Using Spies in Sky to Lower the Toll of Temblors

In the quest to understand the subtle twitching, squeezing, and stretching of Earth's crust, California researchers are turning to help from above - and below.

A combination of Pentagon satellites in the sky and dozens of radio receivers anchored to southern California's terra not-so firma are part of a new system for providing earthquake-related information to scientists and emergency planners with unprecedented speed, detail, and geographic reach.

The aim, seismologists say, is to provide better estimates of the size and frequency of quakes, to give rapid guidance for police, fire, and emergency-medical services, and perhaps even early warning of impending violent quakes outside the Los Angeles metropolitan area.

"This is a new direction," says Hiroo Kanamori, director of the Seismological Laboratory at the California Institute of Technology, of the efforts to take basic research devices and turn them into tools for a wider group of users.

"Science cannot make very precise predictions" of where or when an earthquake will occur, he says, "so we have to use science for more-effective hazard mitigation."

Perhaps with the exception of Japan, nowhere else in the world is as wired for wiggle as is this region. For Los Angelenos, the Earth's creeping crust presents such a significant threat to life and property that this area has emerged, over the past half-century, as a premier lab for quake study.

Lately, the emphasis on hazard mitigation has grown as researchers have come to realize how extremely difficult short-term predictions are, Dr. Kanamori says.

Even for longer-term estimates of the likelihood of quakes, the devil is in the details. By digging trenches across a fault and reading its history in layers of sediment, researchers such as Cal Tech's Kerry Sieh have been able to spot patterns in how often quakes occur. The evidence suggests an average of one major temblor along this section of the San Andreas fault every 150 years. The last major break there came in 1857. The US Geological Survey says another significant break is likely during the next few decades.

But Dr. Sieh's research shows that the quakes over the past 2,000 years occurred in clusters of two or three over a 100-year span. Each cluster was separated by 200 to 300 years. Since the southern San Andreas had two major quakes in the 19th century, it could be due for another big one to complete the "cluster," or it could be entering one of its dormant periods.

But no one has yet found reliable tipoffs to an imminent quake, says Kanamori. Researchers understand that faults seem always to be on the verge of snapping, needing only a gentle nudge. "It does not take much to set off an earthquake," he says. "That's why you don't see precursors."

To look for the subtle nudging in the crust, geophysicists have turned to the Pentagon's Global Positioning System (GPS). Using highly accurate time signals from a constellation of 24 satellites orbiting Earth and low-cost GPS receivers not much bigger than a laptop computer, scientists can track minute changes in receiver positions as the crust adjusts to strain around faults.

Typically, geologists have taken GPS receivers out once or twice a year. Today, 45 receivers are permanently in place as part of the Southern California Integrated Global Positioning System Network. Another 205 receivers are to be added during the next two years to monitor Southern California's shifting landscape.

The network can help locate hidden faults, which lie beneath the surface and leave no traces above ground. Such a fault beneath the San Fernando Valley caused the Northridge quake in 1994, a 6.7 magnitude temblor that left $27 billion in damage.

The network also gives a continuous view of how strain in the crust throughout the region shifts - not only in response to movement along the San Andreas fault, which forms the boundary between two vast plates in the Earth's crust, but also in response to large earthquakes along other faults in the region. These shifts occur more frequently than large quakes and so are a major factor in gauging Southern California's earthquake potential.

Underscoring the role of the GPS, the August issue of the Journal of Geophysical Research reports the first observation of the crust gradually adjusting after an earthquake out to distances of 155 miles over several years.

The team, led by Yehuda Bock of the Scripps Institution of Oceanography's Institute of Geophysics and Planetary Physics, found that strain released by the Landers quake (the strongest in Southern California in 45 years) put a slow squeeze on the Los Angeles basin, adding stress to the fault responsible for the Northridge quake. The '92 Landers earthquake "in a sense helped trigger" the '94 Northridge event, Dr. Bock says.

Using continuous data from a network of seven GPS receivers operating since 1991, the team reconstructed changes in the crust following the Landers earthquake. Long after the quake rent a series of faults in the Mojave Desert, the crust continued to subtly deform, shifting 3 to 5 millimeters per year.

"Post-seismic deformation has been seen before," but not over these distances, Bock says. Gathering large amounts of GPS data daily to map the direction of heaviest stress buildup "gives you a strong tool for seeing how strain migrates, and could really improve hazard assessments."

* Next: Taking the measure of a quake.