But the grinding plates of the earth move in mysterious ways, and Friday the largest recorded earthquake in Japan’s history — a stunning magnitude 8.9 on the short list of most violent events since the dawn of seismology — hit about 230 miles northeast of Tokyo, generating a tsunami that within minutes socked the coast of Honshu, Japan’s largest island.
The epicenter of the earthquake was about 15 miles below the sea floor and about 80 miles east of the coastal city of Sendai. Tremors are common throughout Japan, and this one was near the Japan Trench, where the Pacific plate, the speediest of the earth’s major slabs of crust, dives beneath the islands of Japan in what’s called a subduction zone. There was a major tremor, magnitude 7.9, just two days ago — what now looks like a foreshock.
But although this is a seismic zone, part of the so-called Ring of Fire that lines much of the Pacific, until recently it wasn’t considered one of Japan’s most vulnerable areas. A 2009 paper by Japanese scientists discussed the possibility of a major earthquake in this part of Japan. Science does not instantly alter public policy, however.
The Japanese government has been prepared since the 1970s for the Tokai Earthquake, the idea of which emerged from the study of previous events along the Nankai Trough, another plate boundary that slides along the underside of Japan. This is among the most complex seismic zones in the world.
The plate boundary off the coast of Sendai had not had a “mega-quake” in the modern era. It may not have suffered a major rupture like this for more than 1,000 years. The closest analog may be a tremor recorded by monks in the year 869, according to Dave Applegate, a senior earthquake specialist at the U.S. Geological Survey.
“If anyone is in position to ride this out, it is the Japanese, and yet we see the scope of the devastation,” Applegate said.
Scientists said the event has reinforced a growing sense that the field of seismology needs to ditch some of its presumptions.
“This is a continuation in a sense of the cold shower that we got in Sumatra — these mega-earthquakes take place in places we do not expect them,” said Emile Okal, a Northwestern University geophysicist reached in Tahiti, where he was preparing to evacuate in advance of the tsunami generated by the quake. The huge Sumatra earthquake six years ago that generated the devastating tsunami along the rim of the Indian Ocean happened on what had been presumed to be a relatively quiescent stretch of a subduction zone.
“That means that on a global scale we should consider that all subduction zones are potential locations for such events,” Okal said.
Quakes aren’t predictable in time, space or intensity. Hazard maps give a good sense of where something is most likely to happen, and the theory of plate tectonics, developed largely since the 1960s, is considered a triumph of modern science. But there is an element of chaos in the way the stresses of the earth relieve themselves. And an earthquake in one place can increase strain on a fault some distance away.
“It’s really just a kind of guessing game, and Mother Nature never really puts up with those guessing games,” said seismologist Dave Wald of the U.S. Geological Survey’s National Earthquake Information Center in Golden, Colo.
Susan Hough, a USGS seismologist in Pasadena, Calif., noted that the recent earthquake in Christchurch, New Zealand, happened on an unmapped fault and caught scientists somewhat by surprise.
“We do tend to focus on the expected events. We’re going to get blindsided by unusual events. . . . But uncommon events happen,” Hough said. “The analog that’s worrisome is Boston. Put a 6.1 under Boston. You have all that un-reinforced masonry.”
Robert Geller, a geologist at the University of Tokyo, said by email: “The bottom line is that it’s not possible to identify in detail which specific areas are particularly dangerous. Also, quakes are not in any sense periodic. Unfortunately some earth scientists, including some government officials in both Japan and the U.S., persist in making highly area-specific risk forecasts and also using models based on periodicity of quakes.”