In the past 10 days, California has felt earthquakes of magnitude 3.0 and greater. While the quakes haven't caused any significant damage, they are an unnerving reminder that the state is due for the Big One.
Could a warning system help -- even if it only offered mere seconds to prepare? The state is banking on it.
California is setting up a network of seismic monitors that would detect the first vibrations from a quake, and transmit a warning to affected areas before the shaking gets violent. This could offer citizen precious seconds to brace for the shaking.
"It's matured to the point where it's sort of duct tape and bailing wire," said Doug Given, a USGS Geophysicist and coordinator for the earthquake early warning project. "It's not ready for public use yet ... But we want to move that into a more fully functional and robust system."
This isn't earthquake prediction -- that's not possible yet. While scientists can say where they think stronger quakes will happen, there is no way to forecast them the way meteorologists do the rain. But earthquakes aren't instantaneous, and that's where an early warning system can help.
The key is the different types of waves that make up a typical earthquake. Initially there is the P-wave (primary wave), which is a pressure wave like the sound waves in air. It is often felt as a sudden vertical motion (though not always). That arrives ahead of the S-waves and surface waves. Those are more like the waves in the ocean.
The surface waves produce the swaying and up-and-down motion that makes a building collapse. The gap between the two depends on how far one is from the epicenter of the earthquake.
The speed of the waves is limited by the speed of sound in rock. That's still fast -- P-waves can hit 5 to 8 miles per second and S-waves travel at about 70 percent of that. But an electronic signal is faster. It's most useful during the more powerful temblors, and there is a "golden donut" around the epicenter far enough away that a warning of a few seconds might do some good, and close enough that there is danger. At some distances warnings might be 90 seconds -- enough time to take action and get to a safer place.
USGS studies put the likelihood of a 6.7 or greater quake in California's San Francisco area at about 62 percent in the next 20 years. A 2008 study placed the odds of a 7.5 or greater quake at 46 percent out to 2038. The latter would be more powerful than the Northridge quake of 1994 (magnitude 6.7) or the Loma Prieta quake in 1989 (magnitude 6.9). So setting up some kind of early warning system is an urgent need.
In the proposed set-up, each sensor would be connected a broadcast-type system. When a P-wave is detected at more than one location they would fire off a signal, broadcasting a warning ahead of the surface wave. The most important function would probably be for facilities such as power plants and public transportation. A nuclear power plant, for instance, could shut down, or a subway train could slow, stop or pull into the next station.
Another function is to tell how strong the earthquake will be. It's possible to tell the strength an earthquake is likely to have, so seismologists can say a quake will be magnitude 7 or 8 from the get-go, and warnings can reflect that.
The optimal spacing for sensors is about one every 20 kilometers (12.4 miles). There are already about 300 such sensors active in southern California and they get data from another 100 around the state (mostly in the Bay Area). Ideally, the system would need another 50-100 sensors in place.
Many cities have implemented early warning systems. In Japan, the technology has been in place since the first Shinkansen (Bullet Trains) were built. They were originally designed to give enough warning to slow and stop the trains during earthquakes. Now, when an earthquake is felt in Japan, warnings pop up on smart phones and televisions. Other cities with similar systems are Istanbul, Mexico City, and Bucharest. In that sense California is a bit behind the rest of the world, Given said.
Once a warning system is in place, however, people need to know what to do with the information.
"This carries with it the problem of being a relatively rare event -- for people to respond appropriately to something that occurs only every couple of decades," Given said.
Christie Rowe, assistant professor of Earth and planetary sciences, at McGill University, said government structure matters. "There was plenty of time for warning of the Boxing Day 2004 before it arrived in Sri Lanka and east Africa -- but no government structures in place to inform people, and no societal memory of similar events to tell them how to react," she wrote in an email.
Another wrinkle is telling people how much time they have. While it is possible to send messages to smart phones, in order to be accurate the transmitter has to know where the receiver is. There are a number of ways to do that but neither implementing the technology nor the calculations are trivial. In Japan the ability to send millions of text messages simultaneously has been in place for some time; the U.S. wireless providers are a bit behind in that respect (though that is changing fast).
The last issue is funding. Given said to build and run the network necessary to cover California, it will cost about $80 million over five years. Originally the project was to be federally funded, but the amounts hoped for may not come through, and the state doesn't look like it will be paying for such a project either. The California project is currently a cooperative effort between Cal Tech, the University of California at Berkeley and the United States Geological Survey.
Donald Turcotte, a geology professor at the University of California, Davis, who is unconnected with the early warning project, has been studying how to predict where earthquakes are likely to happen.
"The early warning situation is really pretty straightforward," he said. "The basic question is how willing are you to spend money to implement it."