On March 11, the seafloor 130 kilometers off Japan's eastern coast slipped more than 20 meters beneath the crust that makes up the Pacific plate, pulling the island nation as much as 4.3 meters closer to California and its coast 66 centimeters down. In fact, the first geologic sensors on the seafloor, which happen to lie near the center of the Tohoku-oki quake, as it is now formally called based on the closest regions of the island nation to the quake's epicenter offshore, registered a shift of some 24 meters east-southeast and an uplift of three meters at that point.

"Japan has historically suffered damage from huge earthquakes, whose focal regions usually lie beneath the seafloor," says hydrographer Mariko Sato of the Japanese Coast Guard, whose team compiled the seafloor data, published online May 19 in Science. "We installed seafloor stations to monitor crustal movements offshore directly."

That seafloor data, however, was not available to two other analyses of the earthquake, also published this week online in Science. Geophysicist Mark Simons of the California Institute of Technology and his colleagues developed a computer model of the several minute–long temblor based on more than 1,200 surface global positioning sensors as well as 12 tsunami buoys that also matches the Sato data. "It turns out our model fits it almost perfectly," Simons notes.

The earthquake—estimated at magnitude 9.0 on the Richter scale—occurred in a total area much smaller than previous large earthquakes, such as the 8.8 Chilean earthquake last year, arguing that the slippage was much greater for the Japan quake, one of the four most powerful earthquakes on record. "We suspect, but we have no clue, that there was one region that was particularly stuck—an old seamount or chunk," Simons explains. "That roughness made it so stuck that stress had to build up for a long time before it decided to fail."

In fact, it appears that the historical precedent for this earthquake occurred in July of the year A.D. 869—dubbed the Jogan earthquake—based on paleo-tsunami findings as well as documents from that time, meaning that such large slips occur on millennial timescales whereas smaller quakes characterize the creeping subduction of roughly eight centimeters per year that occurs normally. That helps explain why such a large earthquake was unexpected in the region, resulting in catastrophic consequences that included more than 24,000 people dead or missing and fuel meltdowns in three reactors at the Fukushima Daiichi nuclear power plant on the coast.

"The area to the south [which is closer to Tokyo] is no different," Simons says, in that a major earthquake would be unexpected but not impossible—and now a powerful temblor has occurred in the area immediately adjacent to it. That may suggest a so-called "unzippering" series of quakes, in Simons' words, like those in Sumatra after the 2004 Indian Ocean earthquake and tsunami, that propagated over time along the fault. "It has the potential for a large event, and it has the potential to just creep" with smaller events, Simons explains. "We shouldn't be complacent."

The other analysis, led by Satoshi Ide of the University of Tokyo, argues that the deep subterranean slip was much smaller, whereas the shallower movement was "too large, and probably released more than 100 percent of the accumulated stress," Ide says. That large movement—releasing energy equal to 9.1 million gigajoules, or roughly the amount of energy contained in 1.5 billion barrels of oil—explains the massive tsunami that devastated the northeastern Japanese coast. "[The] tsunami was large, but not unexpectedly larger than the estimation using seismic waves," Ide adds.

Yet, this model of the quake does not match up well with the information from the ocean floor sensors—incorporating that data into future computer simulations should give a better picture of what actually happened during the massive tectonic event. That does not necessarily mean, however, better predictions in the future. Simons, for one, believes that earthquake forecasting is a fool's errand. "Earthquakes are just completely unpredictable for a lot of basic physical reasons," he says. "These rare and extreme events are so rare and extreme that the only way to understand them, even if we're only interested in our backyard, is to understand them wherever they occur."