Earthquake scientists are proposing that “crowdsourcing” hundreds or even thousands of volunteers with their highly sensitive mobile phones could create a seismic early warning system to alert users of oncoming seismic shocks.
Seismologists in Menlo Park and UC Berkeley are testing the phones and foresee them as particularly useful in developing regions, like Southeast Asia and parts of Africa, that are prone to large and often devastating earthquakes but where more sophisticated warning systems don’t exist.
A study led by scientists at the U.S. Geological Survey found that networks of today’s smartphones could alert the onset of ground-shaking from earthquakes of magnitude 7 or larger. Such a network, however, would be ineffective — at least for now — for smaller quakes that can often be just as damaging as more powerful ones when they hit in crowded cities.
Benjamin A. Brooks, who led the study with USGS colleague Sarah E. Minson, said Thursday the team has received a $250,000 grant from the U.S. Agency for International Development to run an early study involving 250 smartphone detectors in Chile, where quakes of magnitude 7 or larger strike every year or two.
No early warning system exists for quakes in coastal Chile, but the country is now developing a tsunami warning system based on smartphone networks because many large quakes have triggered the often deadly coastal events.
“We hope that in a year or so we’ll know how effective the phones can be for earthquake early warning,” Brooks said.
Crowdsourcing as a way to recruit volunteers for projects is a standard online technique today: Kickstarter, for example has successfully found thousands of supporters for volunteer art projects, as well as investors in start-up companies. The crowdsourced SETI@home numbers 6 million participants reporting their efforts on line as they search for signs of extraterrestrial life with their computers.
It is unclear just how big a “crowdsourced” network must be to work as an accurate tool warning populations that a major ground shaking event is coming, Brooks said.
“Any false alarms in a warning system would be untenable,” he said, so the accuracy of any warning network would depend how big the system is.
In a simulated test of a magnitude 7 quake on the Hayward fault, the researchers reported that at least 100 smartphones could have detected the temblor, and that 5,000 phones would have detected it within 5 seconds. That’s “a sufficient amount of time to issue a warning to major population centers like San Francisco and San Jose,” Brooks said.
California is already using its highly sophisticated but still incomplete early warning system called ShakeAlert, and the scientists propose that a smartphone network could well supplement the ShakeAlert system.
At the UC Berkeley Seismological Laboratory, the system’s regional center, seismologist Peggy Hellweg called the crowdsourced smartphone network “a really good and interesting idea.”
“But I’m still wondering, how do you discriminate between me running down the stairs with my phone in my pocket and an earthquake,” Hellweg said.
At the Berkeley seismology lab, Kong Qingkai, a fourth-year graduate student from China, is researching his own small quake network with 80 smartphones, and is testing the phones on the shaketable at UC’s Richmond Field Station, where engineers shake model structures violently to determine their ability to withstand large quakes.
“I live on the second story of a house in Berkeley,” Kong said, “and when I sleep, I keep two of my own smartphones on the desk next to my bed. The Napa quake last August was 40 kilometers away, and when I woke up I saw that both phones detected it right away. They worked.”
©2015 the San Francisco Chronicle. Distributed by Tribune Content Agency, LLC.
