Scientists put a new spin on predicting earthquakes

Predicting earthquakes is one of those holy grails of science that seems will be out of reach forever. But every time a big one strikes and wreaks havoc and tragedy, the tendency is to turn to the experts and ask why couldn't there have been at least some warning?

Fault lines have been mapped out and modeled and are pretty well understood, and seismologists can say with a lot of confidence where earthquakes are going to happen, but the critical "when" has remains a guessing game.

But we could be getting one step closer to changing that.

A team of scientists took the data from every magnitude 7.0 or greater temblor since 1900 and found a pattern - every 32 years there has been an increase in the number of large quakes. And the only geophysical happenings that correlate with the phenomena are a slight slowing in the Earth's rotation - five years earlier.

By "slowing" they mean days may lose a couple of milliseconds but by their theory it is enough to trigger something that sets off more big earthquakes than usual.

The scientists are Roger Bilham of the University of Colorado and Rebecca Bendick of the University of Montana in Missoula and they reported their findings recently in the journal Geophysical Research Letters.

"The Earth offers us a 5-years heads up on future earthquakes, which is remarkable," Bilham told Science. "Of course that seems sort of crazy."

Bilham and Bendick began their research trying to see if earthquakes set each other off through a kind of synchronous oscillation, the way strings on musical instruments will automatically vibrate if their pitch is sounded, sort of. Faults in the Earth's crust, they reasoned, with their buildup and release of stress could be considered "really noisy, really crummy oscillators", as Bendick put it.

But as they combed through the data they found that earthquakes clustered in time, not in place, and the numbers of large earthquakes came at intervals of 32 years. So were the earthquakes talking to each other or were they getting a nudge from some external force?

Running down the laundry list of forces at work on the globe they found the match with the length of a day, which is known to go through decades-long cycles of speeding up and slowing down.

When day lengths slow down or speed up it also disrupts the Earth's magnetic field. Scientists think changes in the flow of the molten iron of the outer core may cause both, but they are not sure. As one seismologist put it: it's not like we can go down there and watch what's going on.

Does the slight slowdown basically cause the Earth's liquid mantle to "slosh" and rub the Earth's crust the wrong way? Or does the mantle stick to the crust and give it a tug?

And even if we could go down there to see, where would we start. Enormous earthquakes happen continually, most that don't affect people, but when the random nature of the event strikes places like Haiti or Sichuan, tragedy follows.

It is sort of chilling to note that the 2008 Sichuan 7.9 earthquake that killed upwards of 80,000 was the deadliest quake in China since the devastating Tangshan, Hebei, 7.5 quake that killed 242,419 in 1976 - 32 years earlier.

UC-Berkeley geophysicist Michael Magna calls the evidence for the link between shorter days and more earthquakes compelling. "I've worked on earthquakes triggered by seasonal variation, melting snow. [Bilham and Bendick's] correlation is much better than what I'm used to seeing."

At any rate, we'll all know soon. The Earth's rotation started to slow down four and a half years ago. By this theory, 2018 then should see five more major earthquakes - 17 to 20 - than usual.

Contact the writer at chrisdavis@chinadailyusa.com.