The massive earthquake that struck under the Indian Ocean southwest of Sumatra on April 11, 2012, came as a surprise to seismologists and left them scrambling to figure out exactly what had happened. Analysis of the seismic waves generated during the event has now revealed a complicated faulting process unlike anything seen before.

"Nobody was anticipating an earthquake of this size and type, and the complexity of the faulting surprised everybody I've spoken to about this," said Thorne Lay, professor of Earth and planetary sciences at the University of California, Santa Cruz.

During the magnitude 8.7 earthquake, the Indo-Australian plate - a major tectonic plate that includes Australia and the surrounding ocean--ruptured over a complex network of at least four faults lying at right angles to one another.

According to Lay, the energy released on each fault individually was about magnitude 8, adding up to a total event magnitude of 8.7 (a point higher than the 8.6 value initially reported for the quake).

The initial shock was followed two hours later by a magnitude 8.2 aftershock on yet another fault to the south.

Most great earthquakes (magnitude 8 and above) occur at the edges of plates in subduction zones, where one plate is diving under the adjoining plate and motion along the fault causes vertical displacement of the surface.

The April earthquake involved horizontal ("strike-slip") motion on a series of faults in the middle of the plate.

It was both the largest strike-slip earthquake and the largest intraplate earthquake ever recorded. The faults broke through the upper part of the plate and appear to have slipped as much as 35 to 40 meters during the quake, Lay said.

"What we're seeing here is the Indo-Australian plate fragmenting into two separate plates," he said. "We've seen local fragmentation on a small scale, but this is an unprecedented opportunity for us to witness it on the scale of a giant tectonic plate."

The fragmentation is caused by stresses within the plate resulting from its collision with Asia in the northwest, which slows down the western part of the plate, while the other side continues moving steadily north, sliding under Sumatra to the northeast.

The Indian subplate will eventually separate from the Australian subplate, but exactly where the plate boundary will form is not yet clear, Lay said.

He added that the process of forming a new plate boundary will take millions of years and is likely to require hundreds if not thousands of earthquakes like the one in April. "This was a huge earthquake, but it's going to happen again and again to make a through-going fracture that separates the plates," he said.

The good news is that this is not a particularly hazardous type of earthquake, because the horizontal slip does not displace the water above the fault enough to generate a big tsunami.

"It doesn't present a lot of societal hazard, so that's good," Lay said. "It is humbling, though, that this event was well beyond anything we could anticipate."

Lay, UCSC graduate student Han Yue, and University of Utah geologist Keith Koper have published a detailed analysis of the fault ruptures in Nature (published online September 26). Another team reported a similar pattern of faults in a recent paper in Science, but without the quantitative details provided in the new paper, which is accompanied by two other papers addressing different aspects of this unusual earthquake.