This second earthquake was not at a plate boundary and was not directly caused by subduction -- the moving of one plate beneath the other. Rather, it was a normal faulting event, where the Pacific plate stretched, bent and broke.

While Japan and the Kamchatka Peninsula are active earthquake areas, the region of the Kuril Islands where the large November earthquake occurred, had not had a large earthquake since 1915 and researchers are unsure of the exact nature of that event.

Working with Hiroo Kanamori, the John E. and Hazel S. Smits professor of geophysics, emeritus, California Institute of Technology, and Thorne Lay, professor of Earth & planetary sciences, University of California, Santa Cruz, the Penn State researcher looked at the sequence of seismic activity that link these two earthquakes into a doublet.

"Such large doublet earthquakes, though rare, could be an underestimated hazard," says Ammon. "We are also interested in what these events tell us about how earthquakes interact, how the stresses and interactions allow one earthquake to trigger another."

Looking at the seismic record, the researchers found a series of smaller, foreshock earthquakes beginning about 45 days before Nov. 15. On Nov. 15, there was the magnitude 8.3 earthquake on the plate boundary, the largest event of 2006.

"Within minutes of the Nov. 15 earthquake, seismic activity began on the Pacific plate in the area where the January earthquake would take place," says Ammon. "This large second earthquake generated a larger amplitude of shaking in the frequency range that affects human-made structures than the first earthquake."

Usually, aftershocks from a large earthquake are at least one order of magnitude less than the main event and taper off rapidly. In this case, the events within the Pacific plate east of the plate boundary did not taper off, and the second event that occurred in January was about the same size as the first earthquake.

Earthquakes at plate boundaries in subduction zones occur when the plate that is going under – being subducted – gets temporarily stuck and causes compression in the plate away from the edge. Tension builds and when the plate overcomes the friction holding it, it moves downward, slipping under the top plate and causing an earthquake. According to the researchers, the second earthquake that occurred on the Pacific plate happened because of bending experienced by the pacific plate that occurs before it subducts beneath the upper plate. As the front edge of the plate slipped, the plate east of the November earthquake bent, cracked and broke in January.

Like pie crust, when the Earth's crust bends, small cracks begin to appear – these were the small shocks that began immediately after the first earthquake – but when the bending becomes severe, a larger region of the crust breaks – creating the second, very large event.

In the United States, subduction zones exist only in the Pacific Northwest, Alaska and the area around Puerto Rico. The researchers note, "Triggering of a large outer rise rupture with strong high-frequency shaking constitutes an important potential seismic hazard that needs to be considered in other regions."

The National Science Foundation and the U.S. Geological Survey funded this research.