Continental Wobble Preceded Japan Mega-quake, Groundbreaking Study Says

A similar signal of micro-creep may have happened in Chile too before the last giant quake there, in 2010, but the system can’t be used to predict quakes any time soon

Ruth Schuster
Ruth Schuster
Send in e-mailSend in e-mail
A scene of devastation wrought by the magnitude 9.0 earthquake in Japan, March 2011.
A scene of devastation with a broken mannikan's hand, wrought by the magnitude 9.0 earthquake in Japan, March 2011. Credit: AP
Ruth Schuster
Ruth Schuster

Continental drift is the theory of the formation of the continents as we know them. The continental plates of rock are moving relative to one another: pulling apart, colliding or moving side by side at different speeds. We think of their movement as being more or less constant creep, with each plate moving in one direction vis-à-vis the adjacent plate/s. Now an international team of researchers has discovered continental wobble – movement back and forth – of the land masses months before the 2011 mega-quake in Japan, and possibly ahead of the 2010 giant quake in Chile too.

During a span of months, the land masses of Japan shifted eastward, then westward, then eastward again before the strongest earthquake in the country’s recorded history, researchers from the GFZ German Research Centre for Geosciences, Potsdam, National Seismological Center in Santiago, Chile, and Ohio State University report in Nature. Measured at magnitude 9 on the Richter scale, the March 2011 quake in Japan killed more than 15,500 people.

A similar tectonic wobble may also have happened ahead of the magnitude 8.8 earthquake that hit Maule, Chile, in February 2010, says the team, though the paucity of data makes the conclusion tentative.

The Japan wobble study is based on signals from over 1,000 Global Navigation Satellite Systems in the country that tracked plate motions before, during and after major seismic events in general.

Both the Japanese and Chilean quakes were in subduction zones: places where plates of the Earth’s crust slowly collide. In the case of denser oceanic crust crashing into lighter continental crust, typically the oceanic plate “slides” below the other and descends into the mantle. If two roughly equally dense continental plates collide, they may crumple up against one another with neither subducting, forming great mountain ranges.  

The aftermath of the mega-quake in Japan, March 2011.
The aftermath of the mega-quake in Japan, March 2011.Credit: AP

The islands comprising Japan sit on or near the boundary of four tectonic plates: the Eurasian, Pacific, North American and Filipino. Chile sits on a triple junction of the South American, Nazca and Antarctic plates, and is atypical in that a mid-oceanic ridge is sliding under the South American plate.

The wobble the researchers observed in Japan and possibly Chile ahead of the giant subduction-zone quakes were markedly different from the steady, cyclical movements the Earth’s land masses continuously make, the team says.

Mark you, the Japanese plates’ twitches eastward, westward and then eastward again were wee: 4 to 8 millimeters in surface displacement, not movements perceptible to life on land. But they lasted five to seven months and spanned thousands of kilometers, the team says.

“What happened in Japan was an enormous but very slow wobble – something never observed before,” stated Michael Bevis, a co-author of the paper and professor of earth sciences at Ohio State University.

Continental wobble could therefore be a harbinger of trouble to come. But it is early days to determine that all, or even most, mega-quakes are preceded by the phenomenon.

“We don’t know because we don’t have enough data,” Bevis said. But the findings suggest it bears watching for wobble when assessing seismic risk in subduction zones, which include Alaska, Sumatra and the Andes, he said.

The aftermath of the mega-quake in Chile, February 2010.
The aftermath of the mega-quake in Chile, February 2010.Credit: REUTERS

Tru’a believer

Earthquake prediction is nonexistent at this point, though efforts are being made by geologists the world wide. We can say that a given area is more prone, but not when a quake might happen, precisely where or how bad it might be. Historical statistics only take us so far: For instance, based on past frequency, geologists warn that a major quake in Israel is overdue, and in southern Israel it’s very overdue. But they can’t say it will happen tomorrow, in a year or even in our lifetime.

Early warning systems for earthquakes are something else: they transmit alerts when an earthquake begins. The idea is that if a quake is detected really fast at the epicenter, people at a distance can be alerted and will have precious minutes or at least seconds to take protective measures before the quake reaches them – such as racing out of the building, sheltering under a hopefully robust table or in a “safe room” or stairwell. Israel, for one, is working on an early warning system called Tru’a.

It bears adding that one knotty conundrum for early warning designers is the threshold beyond which a warning is transmitted.

Recent big-data research discovered 2 million quakes in southern California over 10 years, from 2008 to 2017, mostly micro-quakes too small to be discerned by the locals. The quakes averaged 495 a day. If you set the bar for the early warning system too low, you get too many warnings and people would just ignore them.

What could have caused the wobble? The geologists modeled the surface displacement back-and-forth-and back again ahead of the 2011 Tohoku-oki earthquake and think it may indicate an initial “slow slip” in which two adjacent oceanic plates beneath Japan “silently” and slowly slipped along; then came a sudden, powerful downward thrust that drove the Pacific plate beneath Japan. And that caused the mega-quake, which in turn caused tsunami waves over 40 meters (131 feet) high at some spots, that devastated the coast and led to the Fukushima nuclear disaster. That quake moved the island of Honshu 2.4 meters east.

So could the detection of wobble portend a major quake? Maybe, although much research remains to be done to see if this is some sort of rule. But in any case, right now the answer is no, for a more prosaic reason. This study was done using more than 1,000 GPS stations distributed throughout Japan, the team explains. It isn’t that the world’s subduction zones are all studded with GPS stations. Our best immediate defense, therefore, remains appropriate building standards – and adhering to them.