PASADENA, CALIF. — On March 18, 1997, the Mojave Desert pitched and swayed in a magnitude 5.4 earthquake centered underneath the ghost town of Calico, east of Barstow, Calif.
Within five minutes, a map appeared on the World Wide Web showing where the temblor hit, how strong it was, and how the shaking varied with distance.
It was the first map of a significant quake's ground motions to come from a new network of seismometers sprinkled throughout southern California.
Run by the California Institute of Technology in Pasadena, the US Geological Survey, and California's Division of Mines and Geology, TriNet combines the latest in digital quake detectors, computers, and communications to provide nearly immediate feedback on ground motions from quakes.
Data from the network, which received its initial funding in January, is expected to help emergency planners respond more quickly to severe quakes, forecast where ground motion is likely to be heaviest, and eventually provide early warning of looming shocks from distant large quakes.
As far back as the 1970s, seismologists were intrigued by the prospect of developing a network like this. But "the seismic instrumentation was not that good," says Hiroo Kanamori, director of the Seismology Laboratory at Cal Tech. "Now, it's feasible to build such a system."
The TriNet consortium is replacing 300 old seismometers with 200 smaller, more capable digital devices. So far, 50 of the new detectors have been installed, Dr. Kanamori says.
For seismologists, the differences between the old and new are striking. For one thing, the new seismometers are able to accurately measure much stronger and weaker motions.
As a result, researchers can fill critical gaps in ground-motion information that structural engineers need as they design earthquake-resistant high-rises and other key pieces of southern California's infrastructure.
There are two ways to get that kind of information, Kanamori says, "Wait for the next big earthquake, or gather ground-motion information from small and mid-sized earthquakes." By taking the ground's pulse daily, he says, the TriNet array will allow seismologists to estimate the ground motions of a big quake "by adding up the small ones."
The network serves up this motion information in two ways, each of interest for different kinds of buildings, according to Thomas Heaton, an engineering seismologist at Cal Tech. For homeowners, the issue is how quickly the ground accelerates during a quake, he says. As the 1994 Northridge, Calif., quake demonstrated, sudden acceleration can yank at the ground underneath a poorly anchored home, leaving several inches of an otherwise intact house hanging over the edge of its foundation.
For flexible tall buildings, he says, the ground motion's velocity, which affects the rate of swaying, is critical data for designers.
One problem the TriNet array is expected to overcome is the lag between an earthquake and the time it takes to report information about its effects. This can influence everything from how quickly emergency services respond to how quickly the state can seek federal disaster relief.
"It used to be a battle just to get epicenter maps out," says David Wald, a seismologist at the US Geological Survey here, referring to information on the location and magnitude of an earthquake. Following the Northridge quake, which caused $27 billion in damage,it took up to a half hour to estimate the magnitude and epicenter for emergency agencies and the media. Now, that information, plus preliminary ground motion estimates, hits the Web (www-socal.wr.usgs.gov/ pga.html) nearly as soon as the shaking stops.
Timely reporting also affects reconstruction efforts. "After the Northridge quake, freeway overpasses were redesigned, but based on information from the San Fernando earthquake in 1971. We didn't have the [ground-motion] data from Northridge available yet," says Cal Tech seismologist Egill Hauksson.
Moreover, the network's 200 seismometers will be so widely dispersed that they may be able to warn of approaching seismic waves from a break along distant portions of faults such as San Andreas. A heads-up of even 15 to 30 seconds, researchers say, could give officials and the public time to take safety precautions.
As a scientific tool, the network also is expected to help researchers study the Los Angeles basin to see how it changes over time and locate faults hidden deep beneath the surface.
The system is still undergoing a shakedown phase, Dr. Wald says. For example, the network captured a magnitude 4.9 earthquake in the mountains west of the Salton Sea in southern California on July 26. By one measure, ground motions in the softer sediments of the Imperial Valley south of Salton Sea were nearly twice as high as they were at the quake's epicenter. The valley might be expected to shake more, Wald says, because it's filled with softer sediments.
But one of the recording stations in the valley was offline at the time, so the number given for the strength of the shaking was a computer-driven estimate, rather than an actual measurement. Indeed, the map's Web site carries disclaimers pointing out several potential sources for inaccuracy.
Still, researchers have high hopes for the system. "This is going to give us some exciting new data for research," Dr. Hauksson says, "And it's going to fulfill needs for public safety agencies, emergency managers, and future building codes."