NewsWorld
SCIENCE

US seismologists have new approach to earthquake prediction

US seismologists say offshore faults sometimes signal a catastrophic rupture in advance, through series of smaller migrating tremors

PUBLISHED : Friday, 16 May, 2014, 10:55pm
UPDATED : Friday, 16 May, 2014, 10:55pm
 

For the past 20 years, getting an earthquake scientist to utter the word "prediction" has been about as tough as getting a biologist to endorse the existence of Bigfoot.

The quest to identify reliable precursors to ground shaking has turned up one dead end after another, from moon phases to radon gas and animal behaviour. Many seismologists are convinced that prediction is impossible because the factors that trigger quakes are so complex.

But a pair of professors at the University of California, Santa Cruz, say it's time to take another look.

In an article on Thursday in the journal Science, seismologists Emily Brodsky and Thorne Lay say that recent megaquakes in Japan and Chile suggest that offshore faults - like the one that runs along the coast of North America's Pacific Northwest - may chatter and creep before they rupture catastrophically.

Called subduction zones, such faults generate the world's most powerful earthquakes, along with deadly tsunami.

Brodsky said it was the monster quake and tsunami off Japan's Tohoku coast in 2011 that made her reconsider her ingrained cynicism about quake prediction.

"I've become a real convert," she said.

A series of small earthquakes started popping off on the seafloor 23 days before the main shock. Foreshocks are not unusual; at least half of major earthquakes have them.

The problem is that it has always been impossible to distinguish foreshocks from run-of-the-mill quakes that do not presage anything bigger.

But in the days before Japan's magnitude-9 quake, the small tremors were on the move, migrating along the fault towards the area that later broke with such force that it jolted the earth on its axis. A handful of instruments on the seafloor showed that the fault was slipping slowly during the foreshocks.

"That suggests that maybe the fault really was doing something different and, perhaps, even measurable, for about a month before a very major, magnitude-9 earthquake," Brodsky said.

A similar pattern of migrating foreshocks appears to have preceded the magnitude-8.1 quake that struck the northern coast of Chile on April 1. Seismologists around the world were glued to their computers, Brodsky said, because the small quakes occurred on a portion of the fault that had not ripped in 137 years.

"As it was happening, we were asking each other: What do you think? What do you think?"

But migrating foreshocks did not always mean a big quake was imminent, she added. Swarms off Central Chile in 1997 were not followed by anything major.

That is why other scientists remain sceptical about the prospects for prediction.

"I think they're being a bit too optimistic," said Dr John Vidale, director of the Pacific Northwest Seismic Network at the University of Washington.

After decades of research on earthquake prediction came to naught by the 1980s, the emphasis in the US shifted to what is called earthquake forecasting. Based on studies of individual faults, scientists calculate the probability of future quakes.

However, Japan is spending US$400 million to install more than 100 real-time sensor packages off the Tohoku coast to detect small quakes and measure seafloor motion.

"There's always a temptation to build these things after a large earthquake," Brodsky said, "and it really would be vastly more useful to do it before."

Geophysicist Dr Kelin Wang, of the Geological Survey of Canada, argues that scientists do not know enough yet to offer meaningful warnings, and that false alarms would be costly to society and confusing to the citizenry.

"The problem with earthquake prediction is in its practice," he wrote.

"I cannot think of a way to make short-term prediction useful and practical, at least not in the foreseeable future."

Share

Send to a friend

To forward this article using your default email client (e.g. Outlook), click here.

Enter multiple addresses separated by commas(,)

For unlimited access to:

SCMP.com SCMP Tablet Edition SCMP Mobile Edition 10-year news archive