Scientists think they may have found a precursor to the kind of earthquake that sent a devastating tsunami crashing into wide areas along the northern Indian Ocean, killing at least 200,000 people. If they are right, it may be possible to warn coastal residents a few years before such an earthquake strikes, thus allowing time for protective measures to be taken.
But if they are wrong, they will join a large collection of scientists who have tried and failed to figure out when an earthquake is about to hit.
The technique applies to subduction zones, like the Pacific Northwest and parts of Alaska, where the oceanic plate is being pushed under the continental plate. Those regions are capable of generating what scientists call "megathrust" earthquakes, frequently resulting in tsunamis.
The technique won't work in seismically active regions like Southern California, where subduction is not the reason faults move. But it may prove to be a lifesaver in many areas of the globe where subduction rules the day.
Nowhere is that need any greater than along the northwest coastline of North America, which many experts believe is past due for a catastrophic quake.
"We think we're on to something," says Jere Lipps, professor of integrative biology at the University of California, Berkeley.
But, he adds quickly, it needs to be confirmed by other researchers.
Lipps and a team of researchers have found evidence that in the years preceding a megathrust earthquake, the coastal area subsides, or sinks, just about a foot. It's subtle, but it shows up clearly in the life and death of tiny microorganisms that dwell right along the high tide line.
Their most compelling evidence was found in Alaska, where a 9.2 magnitude quake struck on Good Friday, 1964, devastating much of Anchorage and a number of other communities and sending a tsunami that killed dozens along the west coast.
The researchers found that between five and 15 years before that deadly quake struck, the coastline near Anchorage sank about a foot.
The subsidence was so gradual that "even people who lived there didn't notice," says David B. Scott, director of the Center for Environmental and Marine Geology at Dalhousie University in Halifax, Nova Scotia.
A report on the research will be published in the May/June issue of the Geological Society of America Bulletin.
Scott has spent 30 years studying tiny, shell-like creatures called foraminifera and thecamoebians, that are "about the size of a pinhead, or smaller," says Lipps, who like Scott is a micropaleontologist. These organisms live in the freshwater mud along the coast, and are very vulnerable to salt water.
Scott and Lipps, along with Andrea Hawkes of Dalhousie and Rod Combellick of the Alaska Division of Geological & Geophysical Surveys, collected 12-foot deep cores from Alaska and Oregon to see if the biological record could tell them anything about events leading up to major quakes.
The Alaska quake was particularly important because scientists know exactly when it occurred. The other quakes are historical, including one in Alaska 1,800 years ago and four in Oregon 3,000; 1,840; 1,670 and 300 years ago. But those dates are not as precise as the Alaska quake, "which we know down to the minute it happened," Scott says.
Thus it is possible to reconstruct the timetable for changes leading up to the 1964 quake with considerable confidence.
When they examined the cores collected in Alaska they found that microorganisms, and the fauna, changed dramatically less than 15 years before that quake struck due to a relatively quick change in the water and mud along the high tide line. Thecamoebians, which live only in fresh water, disappeared, apparently because of the infusion of salt water as the coast subsided. Other changes in fauna and microorganism populations told the same story.
Looking farther, they found the same pattern preceding the other quakes. First the subsidence, and then the quake just a few years later.
But why subsidence before a quake?
"As the oceanic plate dives under the continental plate it pulls the continental plate down slightly and eventually it pulls it down so much it snaps," Scott says. "But what this precursor seems to be is sort of the bending before the snap. It bends a little, maybe for five or 10 years, and then it snaps."
The researchers think their findings could lead to an early warning system for subduction zone earthquakes.
"It's not a prediction," Lipps notes, because even if subsidence does occur, there's no way to know precisely when the quake is going to hit. But this research suggests it won't be long, probably less than 15 years, and possibly within just a couple of years.
Instruments, like tilt meters that reveal when the surface has tilted slightly, could provide a warning in time to do something about it. City planners, for example, are more likely to require reinforcement of public buildings if they have reason to believe a major quake will hit within a few years. They are less likely to take expensive precautions if the best scientists can do is warn them that quakes have happened in the past, and should be expected within the next few hundred years.
If it's coming soon, "you better get your stuff together," Lipps says.
Of course, all of this will require independent verification, and even if it turns out to be on the money it won't help some areas. The notorious San Andreas Fault in California moves laterally, for example, so there's no reason to believe subsidence occurs there before a big quake.
But some experts believe the Great Quake that is waiting in the wings won't be in California. It will most likely be along the Cascadia Subduction Zone, which stretches from Northern California to British Columbia. That one could kill thousands.
A few years warning won't stop it, of course. But at least it would provide the incentive to do all the things residents of that area should be doing anyway. Batten down the hatches.
Lee Dye's column appears weekly on ABCNEWS.com. A former science writer for the Los Angeles Times, he now lives in Juneau, Alaska.