Published online 30 September 2009 | Nature | doi:10.1038/news.2009.964


Past quakes cause future shocks

Seismic waves from earthquakes might make distant fault lines more slippery.

San Andreas faultThe Indian Ocean earthquake of 2004 may have weakened the San Andreas fault.David Parker / Science Photo Library

Distant earthquakes may increase the risk of subsequent earthquakes around the globe, a team of US seismologists reports.

Examining 22 years of seismic data from a section of the San Andreas fault near the small central Californian town of Parkfield — an earthquake hotspot — the researchers found a substantial upsurge in the frequency of small earthquakes in 2005. These occurred following the Sumatra–Andaman earthquake in late 2004, which caused the deadly Indian Ocean tsunami. A similar increase occurred in the mid-1990s, after a magnitude-7.3 earthquake in 1992 in the Californian desert, hundreds of kilometres away from Parkfield.

It has long been known that earthquakes can trigger new quakes from afar. A magnitude-7.9 tremor that struck Alaska in 2002, for example, caused small earthquakes as far away as Wyoming and California. But these earthquakes occurred within hours of the triggering event, presumably caused when the vibrations induced other faults to give way. Many of the ones at Parkfield, by contrast, occurred months later.

The only way that could have happened is if a distant earthquake somehow weakened the San Andreas fault, says the study's first author, Taka'aki Taira, a seismologist at the University of California, Berkeley. "Weakening the fault means the fault can store less stress before [it] fails," Taira explains.

Slippery slope

In addition to observing more small earthquakes, Taira and his colleagues also saw changes in a phenomenon called seismic scattering, in which incoming seismic waves are reflected in multiple directions, like sunlight from a wind-rippled pond. Such changes, the team believes, are associated with the movement of groundwater deep within the fault zone. The water would, in essence, lubricate the fault, making it weaker and more likely to move for as long as the water remained. Their research is published this week in Nature1.

“We've been wondering for a long time about why, sometimes, earthquakes occur in bunches.”

Seth Stein
Northwestern University

How exactly this process occurs has yet to be determined. "We need lab experiments to see how fluids migrate after the ground is shaken," Taira says.

This means that accurate earthquake forecasting is not yet a possibility. "We still have to do many things," says Taira. Nevertheless, the team's finding is an important step in the process of determining earthquake risk. "We only looked in Parkfield, but we speculate that this happens at many places in the world," Taira adds.

The researchers also counted the number of large earthquakes — classed as magnitude 8.0 or greater — in the years following the Sumatra–Andaman earthquake. They found that worldwide, the three years from 2005 to 2007 saw more large earthquakes than any comparable period since at least 1900 — a strong indication that other faults have been affected, Taira says.

Quake conundrum

Future studies of the fault may be aided by the completion of a monitoring station called SAFOD (San Andreas Fault Observatory at Depth), where geophysicists have drilled a 3-kilometre-deep borehole directly through the fault and installed earthquake detectors. SAFOD, which became operational in 2007, can detect much smaller earthquakes than can be measured from the surface. "In the future, it would be a way of looking for these correlations using many more, and smaller, events," says Mark Zoback, a geophysicist at Stanford University, California, who is a principal investigator on the SAFOD project.


The findings of Taira and his colleagues may also help to resolve long-standing geological puzzles, says Seth Stein, a geophysicist at Northwestern University in Evanston, Illinois. One such puzzle is the New Madrid fault zone near St Louis, Missouri, which has produced several powerful earthquakes in the past 2,000 years, but has also had long interludes of inactivity. "We've been wondering for a long time about why, sometimes, earthquakes occur in bunches," he says. "The only thing we could come up with was if the fault got weaker with time."

Stein isn't suggesting that the earthquakes at New Madrid are triggered by distant Sumatra-style events. But the Parkfield study is the first to observe a fault getting weaker, he notes. "That's very intriguing because faults getting stronger and weaker with time could explain a lot of very interesting things." 

  • References

    1. Taira, T., Silver, P. G., Niu, F. & Nadeau, R. M. Nature 461, 636-639 (2009). | Article
Commenting is now closed.