The devastating earthquake of March 2011 has left its mark on science, shaking up widely accepted seismic theory and forcing forecasters to reassess the chance of major quakes from a soft seabed fault line.
The devastating earthquake of March 2011 has left its mark on science, shaking up widely accepted seismic theory and forcing forecasters to reassess the chance of major quakes from a soft seabed fault line.
The tremor rapidly became a massive magnitude-9.0 quake when a small initial source area rapidly widened along an ocean trench, releasing energy from other small stress zones and larger unstressed sections, collectively a zone where seismologists believed major earthquakes could not occur.
Furthermore, the two plates slipped 50 meters, the largest measurement on record.
"I had never before seen a geological fault line slipping as much as 50 meters," said James Mori, a professor of seismology at Kyoto University's Disaster Prevention Research Institute. "That is what took place near the Japan Trench, some 200 kilometers off the coast of Miyagi Prefecture."
Even the significantly larger magnitude-9.5 Chilean earthquake of 1960 is believed to have involved only about 20 meters of fault slippage. The 50-meter movement off Japan deformed the crust and lifted the seabed, spawning a tsunami.
The quake is now known in Japan as the Great East Japan Earthquake. Until it occurred, accepted theory held that large earthquakes could only occur at depths of 20 to 50 kilometers along the boundaries between tectonic plates.
At that depth, rigid plates butt up against each other, accumulate great strain and release it when they shift positions during a seismic event. It was believed big earthquakes were impossible at shallower or deeper depths.
The unusually large slippage took place near the Japan Trench, a plate boundary 7 km beneath the ocean surface, where an oceanic plate begins to dive beneath a continental plate.
It was believed that soft sediments deposited in the trench functioned as a lubricant, oiling the joint and preventing the plates from coupling—sticking—to one another. With reduced friction, the two faces would slide continuously and generate no major seismic events. Scientists believed under these conditions only a certain kind of earthquake was possible, one involving slow slippage along the trench.
There are two possible explanations for why the earthquake was so large. The explanations involve not only the deeper plate boundary but its shallower parts too.
The first theory says the plates had in fact been stuck to one another near the trench and were accumulating strain all along, but that this went unnoticed because of weaknesses in the seismic and geodetic monitoring network.
The second says some mechanism was in play that allowed a sudden large movement despite only a small accumulation of strain.
Until 2011, few seismologists paid attention to this latter hypothesis.
In April and May 2012, Mori was aboard the deep-sea drilling ship Chikyu, which bored into the seabed near where the large slippage occurred.
"We must look right there before we do anything else," Mori recalled telling himself.
Researchers are still working on detailed analysis, but according to initial findings the seabed near the Japan Trench was "liquefied and slushy," said fellow expedition participant Kohtaro Ujiie, an associate professor of geodynamics at the University of Tsukuba.
It did not look like a location where the two plates could have been stuck together and storing up high strain, Ujiie said.
Nevertheless, he believes a large slippage could have been possible.
He said the movement may have originated in a small section under strain, where the two plates have stuck to each other, and then propagated outwards to other portions. Even though those new sections were unstrained, they shifted along with the rest.
During the March 2011 earthquake, the fault initially slid along deeper parts adjacent to the Japanese Islands. Then the huge slippage occurred near the more distant Japan Trench.
In seabed samples drilled by the Chikyu, Ujiie and his coworkers discovered a soft clay mineral called smectite. If water had been present too, frictional heat from slippage would have caused that water to expand and perhaps to facilitate further movement along the fault, Ujiie said.
Gaku Kimura, a professor of structural geology at the University of Tokyo, agreed that the plate boundary may have contained water.
In that vicinity the seabed is covered with thick layers of mud and sediments such as plankton shells.
The sediments contain a mineral called opal, which becomes quartz when the subducting slab pulls it under and heats it. That reaction releases water, which could have rendered the plate boundary more friction-free and helped the slippage to spread, Kimura said.
In the past, researchers have been recording repeated magnitude-7 earthquakes at certain points off the coast of Miyagi Prefecture.
It was believed that some mechanism was applying a brake to propagation and therefore helped to contain seismic activity within a tight area around the trigger point.
But the magnitude-9.0 earthquake involved simultaneous movement of all those magnitude-7 seismic trigger zones as well as areas nearby. Apparently the brakes did not function, and a wide area was suddenly on the move.
Why, then, did the brakes fail?
One focus of recent attention is the slip rate, or the speed at which the two sides slide against each other. Experiments have shown that high-speed slippage of two rocks causes lower friction, which itself further facilitates movement.
Bunichiro Shibazaki, a chief research scientist at the Building Research Institute, used that phenomenon when he modeled earthquake cycles over a period of 3,000 years. He found that when very high-speed slippage took place, friction dropped and all magnitude-7 seismic sources and their surroundings moved together, creating one giant quake.
His simulations predicted that magnitude-7 events would recur every few decades and a magnitude-9 event every 900 years.
But Shibazaki's simulations are based on a number of assumptions, including on just how much friction is present. The balance is fine, and slight adjustments in those values alter the size and frequency of virtual earthquakes, reminding Shibazaki how difficult it is to forecast giant seismic events.
Is there any reason why the March 2011 earthquake should have occurred exclusively off Miyagi Prefecture? Or can similar earthquakes occur just about anywhere with similar underlying faults?
The scientists' quest for answers has only just begun.