Researchers look to devise earthquake-warning system

They're working on algorithms that can analyze seismic data

Researchers in California are working on technology that could someday help create early warning systems to notify residents, emergency officials, utility companies and others of impending earthquakes.

Their efforts focus on the development of algorithms that can accurately analyze incoming seismic data so that earthquakes in cities, including San Francisco and Los Angeles, can be predicted in time to send out automatic warnings. Much of the work is being done by teams of researchers at the University of California-Berkeley, the University of Southern California, the California Institute of Technology and other affiliated schools.

The work is complicated because both cities are located on top of major fault zones, meaning it's unlikely that any early warning systems could provide more than 10 or 20 seconds notice before a quake actually hits.

"Some people feel that 10 to 20 seconds is not worth it because these systems won't give enough time" for people to get out of harm's way, said Bill Leith, coordinator of the advanced national seismic system at the U.S. Geological Survey (USGS) headquarters in Reston, Va. "Other people feel that just a little bit of warning can provide some benefit."

Though it seems like a short time, 10 to 20 seconds could give children time to get underneath desks at school or allow fire, police and other emergency workers to respond more efficiently, he said. It could also help personnel at utility companies, airports, highway departments and other agencies better prepare, he said.

Before researchers determine whether such a system is worth building -- a working system could cost hundreds of millions of dollars -- they must first determine whether it is feasible and how it could be engineered.

The three algorithms being developed would take raw data from existing seismic collection systems in California and process it to determine the strength of an earthquake, how much the ground is shaking and how far away it is.

Research has been under way for about a year as part of a three-year $800,00 grant from the USGS, Leith said, stressing that it's unclear when a prototype system might be ready for testing.

Other countries, including Japan and Mexico, already have earthquake early warning systems, he said.

In Japan, the systems are more highly developed because the government spent a significant amount of money to build them, Leith said. Like Mexico, Japan has large offshore earthquakes that allow for longer warning times, making such systems valuable when a quake is detected, he said.

David Oppenheimer, project manager for the USGS's Northern California Seismic Network, said the grant will pay for the development of three different algorithms that can each be analyzed to determine which one does a better job of providing early warnings.

One of the algorithm projects, called ElarmS, is being worked on at the University of California-Berkeley, where potential users of an early warning system are being asked for input. The other two algorithms are being researched at Cal Tech, while USC is collecting the data from the three algorithms in real time and analyzing it for the project, Oppenheimer said.

Once the initial work is done, other researchers will look at the technical barriers to -- and the costs of -- building a viable early warning system, Leith said. Future researchers would also determine how any warnings would be transmitted.

Even if the research ultimately shows that the system isn't cost effective, the work can improve existing methods of earthquake detection and reporting by shortening the time between detection and notification, Leith said.

"There's value to the tests even if they don't result in a warning system," he said.

Join the discussion
Be the first to comment on this article. Our Commenting Policies