The researchers based their simulation on a real asteroid known to be on course for a close encounter with Earth eight centuries from now. Steven Ward, a researcher at the Institute of Geophysics and Planetary Physics at UCSC, and Erik Asphaug, an associate professor of Earth sciences, report their findings in the June issue of the Geophysical Journal International.

March 16, 2880, is the day the asteroid known as 1950 DA, a huge rock two-thirds of a mile in diameter, is due to swing so close to Earth it could slam into the Atlantic Ocean at 38,000 miles per hour. The probability of a direct hit is pretty small, but over the long timescales of Earth's history, asteroids this size and larger have periodically hammered the planet, sometimes with calamitous effects. The so-called K/T impact, for example, ended the age of the dinosaurs 65 million years ago.

"From a geologic perspective, events like this have happened many times in the past. Asteroids the size of 1950 DA have probably struck the Earth about 600 times since the age of the dinosaurs," Ward said.

Ward and Asphaug's study is part of a general effort to conduct a rational assessment of asteroid impact hazards. Asphaug, who organized a NASA-sponsored scientific workshop on asteroids last year, noted that asteroid risks are interesting because the probabilities are so small while the potential consequences are enormous. Furthermore, the laws of orbital mechanics make it possible for scientists to predict an impact if they are able to detect the asteroid in advance.

"It's like knowing the exact time when Mount Shasta will erupt," Asphaug said. "The way to deal with any natural hazard is to improve our knowledge base, so we can turn the kind of human fear that gets played on in the movies into something that we have a handle on."

Although the probability of an impact from 1950 DA is only about 0.3 percent, it is the only asteroid yet detected that scientists cannot entirely dismiss as a threat. A team of scientists led by researchers at NASA's Jet Propulsion Laboratory reported on the probability of 1950 DA crossing paths with the Earth in the April 5, 2002, issue of the journal Science.

"It's a low threat, actually a bit lower than the threat of being hit by an as-yet-undiscovered asteroid in the same size range over the same period of time, but it provided a good representative scenario for us to analyze," Asphaug said.

For the simulation, the researchers chose an impact site consistent with the orientation of the Earth at the time of the predicted encounter: in the Atlantic Ocean about 360 miles from the U.S. coast. Ward summarized the results as follows:

The 60,000-megaton blast of the impact vaporizes the asteroid and blows a cavity in the ocean 11 miles across and all the way down to the seafloor, which is about 3 miles deep at that point. The blast even excavates some of the seafloor. Water then rushes back in to fill the cavity, and a ring of waves spreads out in all directions. The impact creates tsunami waves of all frequencies and wavelengths, with a peak wavelength about the same as the diameter of the cavity. Because lower-frequency waves travel faster than waves with higher frequencies, the initial impulse spreads out into a series of waves.

"In the movies they show one big wave, but you actually end up with dozens of waves. The first ones to arrive are pretty small, and they gradually increase in height, arriving at intervals of 3 or 4 minutes," Ward said.

The waves propagate all through the Atlantic Ocean and the Caribbean. The waves decay as they travel, so coastal areas closest to the impact get hit by the largest waves. Two hours after impact, 400-foot waves reach beaches from Cape Cod to Cape Hatteras, and by four hours after impact the entire East Coast has experienced waves at least 200 feet high, Ward said. It takes 8 hours for the waves to reach Europe, where they come ashore at heights of about 30 to 50 feet.

Computer simulations not only give scientists a better handle on the potential hazards of asteroid impacts, they can also help researchers interpret the geologic evidence of past events, Ward said. Geologists have found evidence of past asteroid impact tsunamis in the form of inland sediment deposits and disturbed sediment layers in the seafloor that correlate with craters, meteorite fragments, and other impact evidence. An important feature of Ward's simulation is that it enabled him to calculate the speed of the water flows created by the tsunami at the bottom of the ocean--more than 3 feet per second out to distances of several hundred miles from the impact.

"That's like a raging river, so as these waves cross the ocean they're going to stir up the seafloor, eroding sediments on the slopes of seamounts, and we may be able to identify more places where this has happened," Ward said.

He added that the waves may also destabilize undersea slopes, causing landslides that could trigger secondary tsunamis. Ward has also done computer simulations of tsunamis generated by submarine landslides. He showed, for example, that the collapse of an unstable volcanic slope in the Canary Islands could send a massive tsunami toward the U.S. East Coast.

A tsunami warning system has been established for the Pacific Ocean involving an international effort to evaluate earthquakes for their potential to generate tsunamis. Ward said that asteroid impact tsunamis could also be incorporated into such a system.

"Tsunamis travel fast, but the ocean is very big, so even if a small or moderate-sized asteroid comes out of nowhere you could still have several hours of advance warning before the tsunami reaches land," he said. "We have a pretty good handle on the size of the waves that would be generated if we can estimate the size of the asteroid."

Planetary scientists, meanwhile, are getting a better handle on the risks of asteroid impacts. A NASA-led campaign to detect large asteroids in near-Earth orbits is about half way toward its goal of detecting 90 percent of those larger than 1 kilometer in diameter (the size of 1950 DA) by 2008.

"Until we detect all the big ones and can predict their orbits, we could be struck without warning," said Asphaug. "With the ongoing search campaigns, we'll probably be able to sound the 'all clear' by 2030 for 90 percent of the impacts that could trigger a global catastrophe."

Rogue comets visiting the inner solar system for the first time, however, may never be detected very long in advance. Smaller asteroids that can still cause major tsunami damage may also go undetected.

"Those are risks we may just have to live with," Asphaug said.