September 30, 2002
The Northridge fault surprised residents of greater Los Angeles with a magnitude 6.7 earthquake on January 17, 1994, killing 60, injuring more than 7,000 and causing more than $20 billion in damage. Now, it has surprised scientists again.
"Recent measurements indicate the Northridge fault has slowed to a crawl," said Dr. Andrea Donnellan, a geophysicist at NASA's Jet Propulsion Laboratory, Pasadena, Calif. "Following a quake, Earth's crust readjusts to changing forces. We expected the Northridge region to re-adjust for at least another 20 years, but it looks as though that readjustment is largely done."
So what put the brakes on? Donnellan says evidence points to the physical state of rocks deep in the region's crust. The Northridge quake began in the lower crust, which unlike Earth's brittle upper crust is usually "gooey." "But with Northridge," says Donnellan, "measurements of the speed seismic waves travel through the region indicate the lower crust is actually cold and hard -- like frozen molasses. This cold crust apparently brought Northridge to the conclusion of its earthquake cycle much sooner than expected, or perhaps is causing the region to adjust so slowly that we can't detect it. In contrast, the lower crust in places such as Landers in the Mojave Desert -- site of a 1992 magnitude 7.6 quake-is 260 to 538 degrees Celsius (500 to 1,000 degrees Fahrenheit) warmer, so quake responses in such regions can be measured for a longer time."
Donnellan cautioned that this information doesn't mean there still couldn't be aftershocks, or that neighboring faults couldn't rupture. "It's an interacting system," she said. "A quake on one fault may turn on or turn off a quake on another. While an earthquake is not likely to happen directly on the fault that broke in 1994, we don't know the status of Northridge's neighboring faults -- where they are in the cycle of building up stress, releasing the stress through a quake, and then building up stress again. There are indications neighboring faults were affected, and it's possible one of them may break, but we don't know when. Understanding the way those faults were affected will take time."
In the study, presented recently to the Seismology Society of America, Donnellan and colleague Dr. Gregory Lyzenga, a JPL geophysicist and professor at Harvey Mudd College, Claremont, Calif., analyzed data from global positioning system (GPS) receivers arrayed throughout Southern California and radar images taken before and after the Northridge quake.
When the quake struck, Donnellan and other scientists had been monitoring the area with GPS for eight years. Each receiver continuously measures its location, detecting surface changes as small as 3 millimeters (a tenth of an inch) horizontally and 7 millimeters (a third of an inch) vertically.
Measurements show that during the 1994 quake, the fault slipped two to three meters (6.5 to 9.8 feet), starting about 18 kilometers (11 miles) deep in Earth's crust and rupturing up to about 5 kilometers (3 miles) below the surface. In the 18 months following the quake, it moved an additional 30 centimeters (about one foot), raising the nearby Granada Hills about 12 centimeters (about 5 inches).
"Ninety percent of the movement after the initial earthquake was a quiet sliding motion, or aseismic," said Donnellan. "Only 10 percent was seismic -- the shaking motion a seismometer measures. That's why new tools like GPS and advanced radar are so important. We're just now seeing how Earth behaves with geodetic measurement -- we're able to observe these quiet processes and get insight into the whole system."
In response to the Northridge disaster, NASA led a collaboration of agencies to develop and install the Southern California Integrated GPS Network, which now includes more than 250 stations in 14 California counties and Mexico. The network allows scientists to monitor movements of Earth's plates. "We can now see how faults interact and look at whole fault systems, not just individual ones," said Donnellan.
But while GPS receivers are measuring continually, they're not measuring everywhere. For the most complete map of every change a quake makes on the surface, scientists turn to satellite interferometric synthetic aperture radar. By combining detailed radar images taken before and after a quake, they can pinpoint surface changes caused by the initial quake and track its aftermath. "You would need a GPS receiver every 20 meters (about 66 feet) to get the same information," said JPL geologist Dr. Gilles Peltzer, a specialist in interpreting radar data. Radar images of Northridge from the European Space Agency's European Remote Sensing Satellite-1 taken two months before the quake, combined with others from December 1995, help fill in the Northridge picture. Those results confirmed the GPS results, indicating the Northridge fault continued to slip after the event. They also shed light on the extent of the region subject to deformation as a result of the quake.
Donnellan said increased use of space-based technologies will lead to better computer models and should, in the next 10 to 20 years, give scientists much clearer insights into how earthquakes behave and where and when they may occur.
The two are now studying other California faults -- Oak Ridge, Sierra Madre, and San Cayetano -- developing computer models to see how stress is transferred between them.
JPL is a division of the California Institute of Technology in Pasadena.
Contacts: JPL/Alan Buis (818) 354-0474