Computers Map, Track Earthquakes Using Motion Sensors
Seismic activity network could be key to early warning system during tremors.
In Chile, Andres Sepulveda never saw the magnitude 8.8 earthquake coming in February, but his laptop captured the data during the catastrophe.
As an assistant professor at the University of Concepción, Sepulveda studies oceanography. But in January, just before he left for a vacation, he installed on his five-year-old laptop a USB motion sensor device. It’s part of an expanding seismic network, which has the potential to send warnings, save lives and bolster public safety efforts when an earthquake strikes. Called the Quake-Catcher Network (QCN), the project uses inexpensive motion sensors in computers to collect earthquake data in real time.
“I had this instrument as part of scientific curiosity,” Sepulveda said. “Chile is a seismic country so I had the idea it could get something while I was away. It was just a test, so I left it on top of a box on the floor of my office. And then the earthquake happened.”
On Feb. 27, the 90-second Chilean earthquake erupted off the coast of the Maule Region, killing more than 400 people and triggering blackouts and a tsunami. It was the country’s strongest earthquake in five decades.
Photo courtesy of the Quake-Catcher Network.
It took days for Sepulveda to get back into his office because shifted furniture blocked his door. Once inside, he found that the USB device on his computer remained intact and collected not only the earthquake information, but also about one hour’s worth of data on the aftershocks.
“He was really interested in the network,” said Elizabeth Cochran, an assistant professor of seismology at the University of California, Riverside, who helped develop the QCN. “Little did he know he would end up recording this earthquake a month later.”
Four years in the making, the QCN was developed by Cochran and colleagues at Stanford University to fill gaps in current earthquake monitoring efforts, hampered by 10 to 15 second reporting delays and costly equipment.
By forming a global web of seismic sensors that captures data on the spot, Cochran said, the network can be the key to an earthquake early warning system.
In an earthquake, shock waves rip through the group, but their speed is no match for electronic signals. The QCN could send messages to nearby locations seconds in advance — precious time that could be used to tell residents to find cover or for public safety departments to stop trains, raise fire station doors and shut off water and gas lines, which can prevent fires.
“When an earthquake starts, you can quickly determine the magnitude and the location,” she said. “What fire stations would love is a few seconds warning to open doors to that fire station so they can easily get equipment out.”
A Shaky System
Monitoring earthquakes has traditionally been a dirty job. Research would include digging into the earth to install new seismometers near fault lines.
“My main frustration is we don't have a huge number of seismometers around,” Cochran said, “just because they're so expensive and it takes a lot of work to install them.”
The technology in the Information Age has given Cochran another, much cleaner, method for monitoring shaky ground: Rather than installing sensors deep in trenches, users can simply install software on their computers. The seismic network utilizes accelerometers — motion sensors that protect data on the hard drive if a laptop falls down or capture movement in video game controllers. Users can upload the sensor for $50 with a USB cable or download the program directly. Some of the newer laptop models have accelerometers already built in.
As more users install the sensors on their computers, seismologists can gather data from anywhere in the world in the case of an earthquake. The idea is to develop a dense network that feeds data to a central computer system to paint a more vivid picture of how an earthquake behaves in a given place and time.
But a sensor on a computer isn’t as sensitive as a regular seismometer. It measures ground motion in three directions and can measure an earthquake with a magnitude 4.0 or higher. But researchers must determine the difference between an actual tremor and somebody banging on a table.
“The main difference is our sensors are not as sensitive,” Cochran said, “so you get lower-resolution data.”
But when a computer senses a tremor, it shoots a signal to the researchers’ servers, Cochran said, and if the server receives multiple pings from the same area, it’s probably an earthquake.