Seabed slide blamed for deadly tsunami.

A year after giant waves swept away 2,200 residents of Papua New Guinea, the disaster has claimed its final victim: the prevailing theory about what causes tsunamis.

Experts on these waves typically attribute them to undersea earthquakes, but evidence collected during marine surveys off the New Guinea coast implicates a submarine landslide or slump, reports the expedition team.

"There is no doubt that there is a shift--a sea change--in interpretation," says David R. Tappin, a coleader of the surveys and a marine geologist with the British Geological Survey in Nottingham.

"This really is one of those big paradigm shifts in science," says team member Philip Watts of Applied Fluids Engineering in Long Beach, Calif., who uses computer models to simulate tsunamis. "We suspect that a lot of the bigger, known tsunamis involved some landsliding."

The Papua New Guinea tsunami, a train of three monster waves, struck the north shore on July 17, 1998 (SN: 10/3/98, p. 221). Ever since then, researchers have struggled to explain how a moderate earthquake, of magnitude 7.1, could have heaved up a tsunami reaching 15 meters tall. Some speculated that the shaking caused an underwater sediment slide large enough to spawn the waves.

In January, a team of researchers boarded a Japanese ship to survey the seafloor. It was the first such intense study after a tsunami. In this expedition, they mapped the seabed and drilled samples of sediments. In late February, they used a robotic sub to photograph the seafloor.

The scientists described their findings at a July meeting of the International Union of Geodesy and Geophysics in Birmingham, England, and in the July 27 Eos.

The exploration focused on the continental slope, which plummets into a 4-kilometer-deep trench. This chasm marks where a piece of the Pacific Ocean floor crashes into New Guinea. As the Pacific tectonic plate scrapes beneath the island, it creates the deep trench and sparks frequent earthquakes.

The survey found that the extremely steep continental slope bears a thick carpet of sediments. In places, this coating has slid downhill in speedy landslides and slower-moving slumps.

On one dive, the researchers discovered a fresh, amphitheater-shaped scar, created when a giant chunk of sediment slumped downhill. "We know this slope failed sometime in the past," says Watts. "The mystery is, When did it go?"

The researchers also found a 15-km-long fault that showed evidence of recent movement. Faulting of the seafloor can generate tsunamis when one side jumps up and the other side drops, displacing water in the process.

Tappin, Watts, and others who think a slump caused the waves contend that the fault could not be the source. Eyewitness accounts indicate that the first wave struck shore about 20 minutes after the main shock of the earthquake, too long for the tsunami to have originated from subsea faulting during the quake. A slump, however, typically lags several minutes after an earthquake and could explain the delay.

When Watts modeled the tsunami, he obtained better results using an undersea slump than a sea-bottom quake. 'I'm convinced that the main part of the tsunami was generated by one giant slump," he says.

Further support comes from a 70-second-long rumble recorded in the middle of the Pacific soon after the earthquake. This sound lasted too long to have come from a small aftershock and may represent a seafloor slide, says Emile A. Okal of Northwestern University in Evanston, Ill.

Some survey participants, however, discount the slump theory. Harry Yeh of the University of Washington in Seattle argues that simulations of a subsea quake can explain the tsunami's size. The team found evidence for small slides but no obvious signs of a giant slump, he says.

Whatever the outcome of the debate, the recent tsunami is forcing researchers to consider slumps as potential sources of giant waves, says Eddie Bernard, coordinator of the National Tsunami Hazard Mitigation Program in Seattle. "I think the good thing that this has done is to open our eyes."

The disaster also suggests that relatively modest quakes, such as the kind that occasionally rock southern California, can trigger giant tsunamis by setting off slides. "This makes the hazard much more dangerous than the scientific community has perceived it in the past," says Bernard.