Web Extra Wednesday, December 10, 2003

Cascading earthquakes in L.A.

About a year ago, on Nov. 3, 2002, a magnitude-7.9 earthquake started on a small thrust fault known as the Susitna Glacier. The movement on that 40-kilometer fault set off a chain reaction: The tremor moved east to the Denali fault, 220 kilometers long, and then stepped onto the Totschunda fault, 70 kilometers long. The next two faults in the chain slipped several meters past each other. Thus, the first thrust fault set off a cascade of faulting along the rest.

Seismologists recognized almost immediately that the same kind of fault geometry — a short thrust fault adjoining two consecutive laterally slipping faults — exists in the Los Angeles basin. Greg Anderson of UNAVCO in Pasadena and his colleagues Brad Aagaard and Ken Hudnut of the U.S. Geological Survey in Pasadena took a closer look, to see if such a cascading event could trigger a large event in the densely populated Los Angeles metropolitan area. Now, reporting this week in Science, the team says while that event could happen, chances are slim that it would set off the San Andreas Fault.

The team examined three faults that lie very close to the San Andreas. The smaller Cucamonga thrust fault describes an arc with the Sierra Madre and San Jacinto strike-slip faults. Researchers do not know how the three faults meet at depth, and whether they adjoin the San Andreas Fault. Using paleoseismology and trenching of the Los Angeles faults, the seismologists modeled how the faults might interact if one of them slipped, using some assumptions about stress fields and other fault conditions.

"Our model suggests a Denali-style event in Los Angeles is unlikely," says Anderson, who presented the results on Monday at the American Geophysical Union annual meeting in San Francisco. The thrust fault in Los Angeles is slightly different in its orientation from the Susitna Glacier fault, with a slight slip component to the thrust; the system, though geometrically similar, is different with regard to the way it handles stress.

The model did show, however, that an earthquake starting on the San Jacinto fault would begin a cascade of seismicity that would travel down the Sierra Madre-Cucamonga fault system, potentially giving rise to an event with 7.0 or greater magnitude, just not along the San Andreas. "An earthquake that's 7.5 to 7.8 magnitude is very rare" on that trio of faults, Anderson says, from past geological research in the region. But for every 50 to 100 earthquakes on the San Jacinto, one might cascade. "It can happen, not often, and no one has seen it before in models," he says. "It must be taken into account in seismic hazards for Los Angeles."

James Dolan of the University of Southern California in Los Angeles says that the work "reinforces our sense that earthquakes on one fault can and have triggered earthquakes on another." He has dug trenches on the Cucamonga fault, showing that the fault breaks every 500 to 1,000 years, evidence that the USGS team used. The Sierra Madre ruptures every 5,000 to 10,000 and the San Jacinto every 100 to 300 years. An event cascading across all three would be the region's "doomsday quake," Dolan says. "If this ever happens, it's incredibly infrequent."

The doomsday event Ralph Archuletta, a seismologist at the University of California at Santa Barbara, imagines is the San Andreas Fault failing along with a big thrust fault, which might be hidden in the valley. But he agrees that the three faults acting alone, which are closer to metropolitan areas than the San Andreas, would be worrisome. "A 7.5-magnitude earthquake near a populated area is not insignificant — whether or not it implies a Denali-style event," he says.

However, Archuletta is less accepting of the model's conclusions on whether or not a cascade on the adjacent faults would trigger the San Andreas fault. Researchers, he says, do not know the states of stress before and after an earthquake for these faults. For the model, Anderson's team set "critical" stresses for the faults of 40 bars of pressure. Faults rupture at critical stress, Archuletta says, and releasing 40 bars of pressure — which is a typical stress drop for a large earthquake event — seemingly should have an effect on the stress regime of the nearby San Andreas Fault. "That's equivalent to an extra 40 years of loading on the San Andreas," he says, which builds up almost a bar of pressure a year.

Nevertheless, Archuletta calls the model "a good start" on what could happen in the Los Angeles basin.

Naomi Lubick


Abstract of Anderson and colleagues' talk
USGS Denali Earthquake site
Past Geotimes coverage of Denali earthquake ("Alaska rumbles," December 2002)

Back to top

Geotimes Home | AGI Home | Information Services | Geoscience Education | Public Policy | Programs | Publications | Careers

© 2014 American Geological Institute. All rights reserved. Any copying, redistribution or retransmission of any of the contents of this service without the express written consent of the American Geological Institute is expressly prohibited. For all electronic copyright requests, visit: