Warning: You Have 30 Seconds . . .

By David Bjerklie

It's one thing to warn people that the Big One is coming -- someday soon. Engineers can design stronger bridges and sounder buildings; city planners can guide development in ways that reduce hazards. It's quite another thing to sound the alarm for an earthquake less than a minute away. Such early-warning systems are technically feasible, though there is debate about whether they would save lives or promote panic.

Convinced that every extra second is of value, Japan has invested mightily in these systems. Originally such devices were not tripped until the ground began to shudder -- scant notice indeed, though enough to prevent the derailment of a speeding bullet train. Such simple motion-detecting systems are now used throughout Japan and in parts of California to halt subway trains and nudge elevators to the nearest floor when ground motion exceeds a certain threshold. The Alaska pipeline is similarly equipped for instant shutdown.

But today more sophisticated systems can alert people as much as a minute before a city starts to shake. "This is possible," explains Massachusetts Institute of Technology geophysicist M. Nafi Toksoz, "because seismic waves propagate through the earth's crust relatively slowly, 5 to 8 km/sec. With an extensive network of sensors, we can locate the epicenter and determine the magnitude of an earthquake. This gives us the opportunity to warn people in outlying areas." How long a warning depends on the distance from the epicenter. Had such a system been in place in Mexico, for example, residents of Mexico City would have had nearly a minute to brace themselves for the two offshore quakes that killed more than 7,000 people inland in 1985.

For California, an effective system could require thousands of sensors along various high-risk faults, linked by sophisticated systems for transmitting data. It would all have to run automatically by computer, says Hiroo Kanamori of Caltech's Seismological Laboratory: "There is no room for human uncertainty or hesitation." Toksoz believes a comprehensive system could be developed over a five-year period for $100 million. A 1991 report produced for the National Academy of Sciences recommended that the Federal Government build a prototype system, but so far, nothing has been implemented.

Would it be worth it? "It's not a solution to all earthquake problems, but there are a lot of practical applications," says Allan Lindh of the U.S. Geological Survey. "To have 30 seconds' warning would sound like a helluva idea to me if I worked near a sulfuric acid vat." Japan has already built advanced systems to shut down nuclear power plants, cut the gas flow from public utilities and issue tsunami alerts. Similar systems could divert incoming aircraft, warn rescue workers of aftershocks and minimize damage to computer, telecommunication and financial data networks.

For ordinary citizens, a 30-second warning might be of limited use unless they had rehearsed what to do. Where it might be most helpful, according to sociologist Dennis Mileti, a warning expert at Colorado State University, "is in schools and offices where people can practice responding to bells." Here too Japan has led the way. Its communities regularly conduct earthquake warning drills, while in California such drills are done only sporadically in some schools and workplaces. Still, most Californians know that when the ground starts to rumble, they must get away from the windows, duck beneath a sturdy table, stand under a doorframe, or pull over if they're in a car. History shows they don't always have time for even these simple measures. But what a difference 30 seconds could make.