There's a remote risk that the Kilauea volcano, one of the most active in the world, will slump into the ocean, triggering a gigantic tsunami, a wave up to 990 feet high, which could devastate coastlines around the Pacific from California to Australia.
Although no tsunami on that scale has been recorded during historic times, there is scientific evidence for mega-slumps and gigantic tsunamis in the Pacific within the past 100,000 years.
Sonar images of the ocean floor show landslides involving hundreds of square miles of rock. And geologists have found deposits of crushed coral, lumps of pumice stone and other wave-borne debris up to 1,000 feet above sea level in Hawaii, Australia and New Zealand which they say could only have been carried there by a tsunami.
Kilauea is bristling with high-tech monitoring equipment installed by geologists to detect changes in the shape of the mountain. The idea is to give advance warning of hazardous volcanic activity from new eruptions of gas and lava to catastrophic landslides.
The Global Positioning System, the navigational system operated by the Defense Department, plays a key role in monitoring Kilauea and other active parts of the Earth's crust. GPS is based on a constellation of two dozen satellites, each broadcasting its precise position and time by radio continuously to receivers all over the world.
Although the Pentagon restricts the accuracy of GPS for civilian users with mobile receivers to no better than 82.5 feet, geologists with specialized fixed receivers can detect movements of the Earth's crust to within a few millimeters.
Paul Segall and colleagues from Stanford University have been monitoring Kilauea intensively since 1990. They are looking particularly for the characteristic swelling of the ground that precedes a new volcanic eruption, as magma (molten rock) flows in.
Segall says the most dramatic change so far occurred on Jan. 30.
"The long-lived Pu'u O'o vent, which had been active for 14 years, stopped erupting and a new fissure opened. We saw the volcano widening for about eight hours beforehand," he says. "The flanks separated by 26 centimeters (10.4 inches) before it split."
At the same time, the moving magma set the ground humming with low frequency vibrations.
No one knows whether those events increase or decrease the chance of a catastrophic slump, which would occur if the southern flank of Kilauea -- a 720 square mile wedge of land resting on magma -- were to break off.
If the worst happened, the tsunami would take only 20 minutes to devastate the most densely populated parts of Hawaii. It would then race across the Pacific at the speed of a jumbo jet.
In the open ocean, the wave wouldn't look spectacular, but as the mass of water ran up against a shelving coastline it would slow down and increase rapidly in height. And in places where bays and estuaries added a funneling effect, the devastation could extend hundreds of feet above sea level.
Countries such as Japan and Australia would have several hours warning before the tsunami struck -- long enough to organize a partial evacuation from low-lying areas, but not enough to avoid catastrophic devastation and loss of life.
Although the threat of a devastating tsunami is greatest in the Pacific, there may be a remote risk in other oceans, too. The island of Reunion in the Indian Ocean is a possible slumping site, and some geologists fear that the steep west side of La Palma in the Canary Isles could collapse into the Atlantic, generating a tsunami that would devastate the coast of Florida.
If tests go well, scientists hope to install two instruments in July, said Eddie Bernard, director of the Pacific Marine Environmental Laboratory in Seattle. One would be off Alaska, south of Kodiak Island; the other west of Vancouver Island off Oregon.
Four other detector systems are planned in the next three or four years, pending federal funding, Bernard said in a telephone interview from Seattle.
The network will cover large gaps in the tsunami warning system and provide the first data on deep sea waves. It will span waters from the northern islands of Japan to Alaska, the northwest United States and Chile.
Bernard, former director of the Pacific Tsunami Warning Center at Ewa Beach, arrives today to take the equipment to sea on the University of Hawaii's research ship Moana Wave.
The vessel will leave tomorrow for a test site about 40 miles southwest of Honolulu. It will return Thursday.
The system works this way: A sensor placed on the ocean floor will measure tsunami waves and send the data to a surface buoy. It will be go via satellite from the buoy to the warning center.
Bernard said the National Oceanic and Atmospheric Administration, which runs his lab, budgeted about $800,000 this year for final engineering, testing and development.
Once all six buoys are installed, he said, it will cost about $800,000 a year to maintain them.
The system was designed to increase protection for the tsunami-vulnerable areas of Hawaii, California, Washington, Oregon, Alaska and Pacific rim countries and islands, Bernard said.
Since scientists have never had wave data from ocean depths, he pointed out, "We still have some ways to go when we start interpreting."
Michael Blackford, geophysicist in charge of the warning center, said, "It's like the beginning of a very long process to find out what the instrument can do for us. It's not a magic bullet here that's going to solve all our problems."
The Civil Defense procedure in Hawaii is to initiate evacuation three hours before a wave arrives, Blackford said.
The travel time for a wave from Alaska is 4-1/2 to 5 hours, he said. "So we're limited to 1-1/2 to 2 hours of tsunami time to gauge to even record something before we initiate evacuation."
If an earthquake should occur in Alaska and trigger a tsunami, Blackford said, "If you're lucky, the earthquake may be fairly close to the gauge and you get data right away. If you're not lucky, the gauge won't get data until the tsunami has progressed to the point where the state wants to start evacuating."