ROBERT SIEGEL, HOST:
From NPR News, this is ALL THINGS CONSIDERED. I'm Robert Siegel. When a big earthquake strikes, it takes time for the seismic waves to spread. That means there's also a time to warn people who aren't at the epicenter to get ready for the jolt. As NPR's Richard Harris reports, scientists on the West Coast are now trying to set up a system to do just that.
RICHARD HARRIS, BYLINE: Japan already has an early warning system for earthquakes, says Richard Allen who runs the seismology lab at UC, Berkeley.
RICHARD ALLEN: So they have a network of seismometers, about a thousand of them across the country, that detects earthquakes as they are happening. They locate the earthquake and they estimate the amount of shaking that you might expect. And then, they push that alert out to people. People get it on their cell phones, they get it on TV, through a whole variety of communication channels.
HARRIS: The high-speed trains put on their brakes. Delicate industrial processes shut down. Eye surgeons step away from their patients. And people take cover. Allen, along with colleagues at Caltech, the U.S. Geological Survey, and the University of Washington, figured, why not establish a system like that on the West Coast?
Allen shows me a prototype on his computer, simulating what would have happened if that system had been in place in 1989, when the deadly Loma Prieta quake struck California.
(SOUNDBITE OF ALARM BLARING)
ALLEN: So the first alert comes up and you can see there's more than 20 seconds of warning.
HARRIS: Loma Prieta's epicenter was in the Santa Cruz Mountains, 60 plus miles from Berkeley. This system would have given people good warning that they were about to be hit with a substantial quake.
ALLEN: We're counting down 10 seconds at this point until the shaking would start here in Berkeley.
HARRIS: Of course, the system wouldn't help if the Hayward fault ruptured directly beneath the campus, but there are plenty of other times when it would.
ALLEN: The best-case scenario in California is about a minute worth of warning. In the Pacific Northwest you can get up to five minutes of warning. The really big earthquakes can give you a significant amount of warning and that's when you would most want it.
HARRIS: This test system already has a few users who are trying it out, including Bay Area Rapid Transit. Kevin Copley at BART says when they test the system or simply get a false alarm, the subway's trains all slow down to 26 mph to reduce the risk of a derailment.
KEVIN COPLEY: We do that system-wide, and then we also institute a hold. Any train that happens to be at a platform at the time would cause it to hold there so it won't close its doors and depart.
HARRIS: And he says it works.
COPLEY: I ride the system as a passenger every day, and it makes me feel a lot better about it. It's not a solution for every earthquake that's going to happen, but it will give us good warning for a lot of them.
HARRIS: But Richard Allen at Berkeley says the current seismic monitoring network isn't complete.
ALLEN: The network is very dense in San Francisco and Los Angeles, but there are gaps in between. And to be able to provide the most warning, you need to be able to rapidly detect the earthquake when it starts outside of the city so you can see it coming towards the city.
HARRIS: The California Legislature overwhelmingly voted to expand the network and provide the quake-proof infrastructure that would be required to send out the warnings.
ALLEN: The catch is they didn't allocate any funding to do it.
HARRIS: Allen figures it would cost $80 million to set up the network in California and run it for five years, or $120 million for a system that would cover the Pacific Northwest as well, which someday will experience the same kind of mega-thrust quake that struck Japan in 2011.
ALLEN: If we have an earthquake this afternoon, the legislature would find the money next week and we would build this system next week. So the question becomes, why don't we just do it now?
HARRIS: He's asked state and federal officials that question and is still waiting for an answer. Richard Harris, NPR News.