Geologists Wonder If The Northwest Is Up Next For A Giant Earthquake

http://www.oregonlive.com

---

Since the magnitude 9.1 earthquake in Sumatra in 2004, five more giant earthquakes have struck the continents ringing the Pacific Ocean. And some experts speculate that the planet has entered a period of increased seismic activity that could trigger giant earthquakes in vulnerable regions including the Pacific Northwest.

A somewhat reassuring new study suggests otherwise.

University of California researchers examined the timing of earthquakes worldwide from 1900 and found no evidence of a domino effect in which one great earthquake triggers others on distant continents. It could be random chance.

"We don't want people to assume that our conclusion means the ongoing risk is small," says study co-author Peter Shearer, a professor of geophysics at the University of California San Diego. "There is a significant risk of big earthquakes in all subduction zones." It's just that the run of very large earthquakes most likely does nothing to change the risk in distant locations, Shearer says.

But the data is less than convincing to some experts, who think it's possible that short-term risk of a great earthquake has increased in the Northwest.

"There is a possibility that maybe we should be worried about this," says Oregon State University geologist Chris Goldfinger. "I definitely do not dismiss it."

Thorne Lay, a professor of geophysics at the University of California Santa Cruz, says scientists don't understand the physical processes well enough to rule out long-distance triggering of great earthquakes based on observations to date.

"One would really need to have good data in the location of each great event to see whether earlier remote large events had changed earthquake timing locally," he says. "Unfortunately, careful work has not been done in all regions of recent big events or the data are very poor due to lack of nearby seismic networks."

The likelihood that a magnitude-9 earthquake will hit the Northwest within 50 years is about 10 to 15 percent, according to the most widely cited estimate. From Northern California to British Columbia, an ocean-spanning slab called the Juan de Fuca Plate is pushing beneath the North American plate in the Cascadia subduction zone. Geologists have found evidence of 19 massive earthquakes on this fault in the past 10,000 years.

The most recent struck in 1700. It drove a 30-foot wall of water over low-lying coastal areas. The tsunami crossed the Pacific and flattened houses along Japan's east coast.

View full sizePeter Shearer & Philip StarkGlobal earthquake magnitudes (A) and yearly rates of quakes at differing magnitudes (B-D) from 1900 to the present.

Global records since 1900 show a cluster of very large earthquakes between 1950 and 1965 followed by 38-year pause in magnitude 8.5 or greater events until the surge that began in 2004. The number of events in both clusters was too few to stand out as more than random variation, Shearer and co-author Philip Stark conclude in a paper published in the Proceedings of the National Academy of Sciences.

"For the very large events, occurring so infrequently, we would have to have a much longer earthquake catalog to evaluate whether they truly are clustered or not," Shearer says. To get around this problem, the researchers looked for significant changes in the global frequency of smaller earthquakes, reasoning that any geological process that increases the rate of large earthquakes should also increase the rate of smaller earthquakes. They found no clustering of smaller earthquakes.

"That suggests that it's more likely that the surplus of magnitude 9 events was just due to random chance, and not due to some process," Shearer says. And, he says, "there isn't a plausible physical mechanism that would link an earthquake in South America, say, to an earthquake in Japan. The events are just too far apart."

While Goldfinger agrees geologists know of no process that could drive a global increase in great earthquakes, he says that doesn't mean such a process doesn't exist. To prove or disprove clustering of great earthquakes, Goldfinger says scientists have to look back further than 100 years of recorded earthquakes.

"Cycles of earthquakes may take many thousands of years to play out, and therefore the 100 years of seismology we have simply doesn't address the long term energy trends at all," he says.

Geophysicist Emily Brodsky, a professor at the University of California Santa Cruz, said: "If we can identify a mechanism of interaction between these earthquakes, then the story would change dramatically. Thus far, no such mechanism has been definitively shown and so we are stuck with the statistics for the time being."

Goldfinger and others have reconstructed a 10,000-year history of major quakes along the Cascadia subduction zone by examining the remnants of undersea landslides. History suggests that magnitude-9 quakes strike in the northernmost section about once every 260 years during a cluster, and clusters are separated by gaps of 1,000 years of lesser activity. On the southern end, smaller quakes -- roughly magnitude 8 -- recur at regular intervals rather than clustering.

In southern Oregon and Northern California, Goldfinger and colleagues calculate a probability of 37 percent that an earthquake will strike within 50 years. In the northern segment, Goldfinger's group puts the odds at 10 to 15 percent during the next 50 years.

Whether the recent upturn in great quakes changes these probabilities, "We just don't have any idea at this stage," Goldfinger says. "I can't say yes, and I can't say no."