Scientists at Cambridge University say they've worked out why the tsunami that devastated Japan in March 2011 was so much larger than expected.
They say it was made more severe by a huge collapse of soft material on the sea bed, resulting in a far greater movement of water than would have been caused by the earthquake alone.
Tsunamis occur when a fault surface ruptures and the shape of the sea floor changes, displacing the water above it. In the case of the Japan earthquake, the fault is the plate boundary, allowing the Pacific sea floor to slide beneath Japan.
"As the plates move against each other, the rocks on their boundaries slowly bend under the pressure, until they eventually crack and slide on faults. When they do, there is an upwards and outwards movement that takes just a few seconds: a movement of 10 metres is a large earthquake and out at sea this causes a tsunami," says Professor James Jackson.
"But data from the Japanese earthquake show a movement of more than 60 metres. Rocks can bend - but they cannot bend to that extent and, anyway, the rocks that moved were sloppy sediments with little strength. This suggests that something else was taking place to increase the movement several fold."
By examining data gathered before, during and after the Japanese tsunami, the scientists have concluded that the squeezing together of two plates in the earth’s crust not only led to a fracture, but also caused a massive collapse of the debris that had built up on the sea bed as tectonic movements scraped loose sediment into an unstable wedge.
"When the wedge of material collapsed, the leading edge split off and shot forward a bit like a pip shooting out of a giant pair of tongs," says Jackson.
"In essence, what happened was a release of both the elastic energy stored in the rocks and the gravitational energy contained in the wedge-shaped build-up of debris."
The research throws a light on other unusually large tsunamis that have long puzzled scientists, including those that struck Nicaragua in 1992, Sumatra in 2004, and Java in 2006.
"These events share a number of unusual features, including large displacements, suggesting that they resulted partly from the collapse of debris," says Jackson.
"We hope that our research represents a step forward in understanding how large tsunamis occur and in what circumstances they are likely to happen."