• A ShakeMap for the 1994 M6.7 Northridge earthquake with a typical seismogram recorded in Pasadena superimposed.
    Credit: Caltech
  • The earthquake early warning (EEW) UserDisplay in action for a scenario M7.8 earthquake. The most intense colors correspond to very strong ground shaking. The banner on top shows expected shaking at the user site. The number '14' on the left indicates warning time, and the expected intensity at the user site is shown in roman numerals, VII. Other information indicates the epicenter and date/time of the earthquake.
    Credit: Caltech
  • Shows maps of the seismic station distribution in 1994 at the time of the Northridge earthquake, and the station distribution in 2014. The stations represented by magenta large triangles have modern digital sensors appropriate for EEW while the small triangles represent old analog sensors, which need to be upgraded for EEW.
    Credit: Caltech

Preparing for Earthquakes with ShakeAlert

A few seconds may not seem like long, but it is enough time to turn off a stove, open an elevator door, or take cover under a desk. And before an earthquake strikes, a few seconds of warning can save lives. The U.S. Geological Survey aims to provide those seconds of warning with ShakeAlert, an earthquake early-warning system now being tested on the west coast of the United States. On July 30, the USGS announced approximately $4 million in awards to Caltech, UC Berkeley, the University of Washington and the University of Oregon, for the expansion and improvement of the ShakeAlert system.

"Caltech's role in ShakeAlert will focus on research and development of the system so that future versions will be faster and more reliable," says Thomas Heaton (PhD '78), professor of engineering seismology and director of Caltech's Earthquake Engineering Research Laboratory. "We currently collect data from approximately 400 seismic stations throughout California. The USGS grant will allow Caltech to upgrade or install new stations in strategic locations that will significantly improve the performance of ShakeAlert."

Earthquakes radiate two kinds of seismic waves: fast-moving and often harmless P-waves, followed by S-waves, which can cause strong ground shaking. A system of seismometers called the California Integrated Seismic Network (CISN) acquires data streams literally at the speed of light and uses several algorithms to quickly pinpoint the earthquake's epicenter and determine its strength. ShakeAlert analyzes the first P-waves in the CISN data streams to send out digital alerts, providing the "early warning" to a region before the slower, destructive S-waves arrive.

While predicting when and where an earthquake will occur is impossible, this early-warning system can give necessary seconds of preparation. Current beta-test users receive these alerts as a pop-up on their computers, displaying a map of the affected region, the amount of time until shaking begins, the estimated magnitude of the quake, and other data. In the future, alerts may be available through text messages and phone apps.

Though still technically in testing stages, ShakeAlert has already provided successful warnings. In August 2014, the system provided a nine-second warning to the city of San Francisco during a magnitude 6.0 earthquake in South Napa. In May, during a magnitude 3.8 quake in Los Angeles, an alert was issued before S-waves had even reached the earth's surface.

"With this new USGS funding, we will be able to add 20 new sensors to CISN, making coverage more robust and thus lengthening warning times," says Egill Hauksson, a research professor of geophysics and a principal investigator along with Heaton on the ShakeAlert project. "Caltech and its partners will be able to continue the high-quality seismological research that is such a necessary foundation for a reliable earthquake early-warning system."

In 2011, Caltech, along with UC Berkeley and the University of Washington, Seattle, received $6 million from the Gordon and Betty Moore Foundation for the research and development of ShakeAlert.

Written by Lori Dajose