COLUMN ONE : Predicting Japan’s Big One : Much is riding on the country’s effort to forecast a monster earthquake that could hit at any time. The program could save thousands of lives.

At least once a week, the ground shakes beneath Yoshimitsu Okada’s feet in a graphic reminder of his profound responsibility.

Okada’s job, quite simply, is to make sure that when an anticipated monster quake strikes in the area south of Tokyo, it doesn’t arrive unannounced. As chief of earthquake prediction for the National Research Center for Disaster Prevention here in “science city” north of Tokyo, Okada heads a team of 20 researchers who hope to detect subtle changes in the Earth’s surface, signifying that the great quake is about to strike.

“We are expecting to succeed,” Okada said in an interview at this unique city, which is Japan’s center for scientific research in a wide range of disciplines.

Japan’s effort is aimed at predicting a catastrophic quake that all the evidence suggests is due any time now. If Okada’s team succeeds, it will mean there is hope that some other earthquakes can be predicted in the future. If the effort fails, it will mean that even great earthquakes may be unpredictable.

For now, however, all the efforts here are concentrated on predicting a single earthquake. Much is riding on Japan’s effort, because the program could literally save thousands of lives if people can be evacuated from several crowded cities just before the quake strikes.

The earthquake is expected to measure at least 8 in magnitude, comparable to the Great Kanto Earthquake that leveled Tokyo in 1923 and the San Francisco earthquake that destroyed that city in 1906. It is expected to hit in the Tokai area south of Tokyo along the densely populated and heavily traveled corridor that links Japan’s capital city with the popular resorts to the south.

There are many programs around the world aimed at predicting earthquakes, including California’s celebrated Parkfield experiment, but they are purely research projects. Okada’s program is something quite different. It is not an experiment. It is an operational program designed to predict a quake in time to evacuate cities before it strikes.

Only China claims to have succeeded at that in the past. Chinese scientists predicted an earthquake in time to save thousands of lives in 1975, based largely on the involvement of 100,000 volunteers who reported evidence ranging from small temblors to unusual animal behavior. Buildings were evacuated just hours before a large quake hit, destroying a city that had been home to half a million people. There were very few deaths.

But one year later, a great quake struck another area of China without warning, leveling the city of Tangshan. China has said 250,000 people died in that disaster, but experts in Japan and the United States believe that the actual death toll may have been nearly a million.

China’s success in predicting the first quake proved how valuable an earthquake prediction program can be. It’s failure to predict the second shows how difficult it is to make it work.

Despite the hope that the Tokai program offers, and the considerable expense and effort that Japan is investing in the project, many experts believe that when the great earthquake finally comes, it will not reveal its intentions in advance.

“I don’t believe that they can forecast the earthquake,” said Etsuzo Shima, professor emeritus of the University of Tokyo and one of the most highly respected seismologists in Japan.

“They are looking for precursors, especially very small earthquakes, before the big earthquake, but in Japan, we have many earthquakes that do not have foreshocks,” added Shima, now an adviser to the research institute of the Kajima Corp., one of the largest construction firms in Japan.

Others agree.

“I have very strong doubts,” said Hirokazu Iemura, a professor in Kyoto University’s earthquake engineering laboratory and one of the leading earthquake experts in Japan. “The more we learn, the more sophisticated becomes the earthquake process, which only God knows.”

But if the project fails, it won’t be for lack of effort.

It is narrowly focused, and the region where the earthquake is expected to strike is probably the most heavily instrumented spot in the world, with the possible exception of California’s Parkfield experiment.

This is the logical place to try it. Japan is probably the world’s best natural laboratory for earthquake research, because quakes occur here far more frequently than in most areas of the globe. Japan records about 1,000 earthquakes every month, “mostly so small you can’t feel them,” Okada said, but his research center detects about 50 earthquakes a year of 5.5 or greater.

As luck would have it, Tsukuba--Japan’s science city--sits atop the most active seismic region in the country. Four or five earthquakes of at least 5.5 magnitude hit right below the city every year.

But it is a large quake that concerns Okada.

“The target of our earthquake prediction is restricted to one earthquake, the Tokai earthquake,” Okada said. “It’s impossible to predict all the earthquakes, so we have only one target, the forthcoming Tokai earthquake, which will be a magnitude 8 class.”

The Tokai region is centered about 100 miles south of Tokyo. While the population there is less dense than the Tokyo-Yokohama area, dozens of cities with many thousands of residents would be severely impacted by such an earthquake.

“It (Tokai) includes the main railways and all principal routes” to the south, Okada said. “The cities will have very serious damage.”

Okada believes that he can succeed because he thinks there will be some changes in the area before the great quake strikes. The concept is analogous to watching a pencil under a microscope as it is bent toward the breaking point. Before breaking, the paint on the surface of the pencil should crack, and small splinters should appear in the wood as the pencil begins to snap.

“We would expect some change just before the earthquake, something different from the ordinary trend,” he said. “Of course, we have no experience of that. We are only expecting.”

Nearly every major university and research institute in Japan is involved in the project.

“There’s a total of about 200 experiments” in the Tokai region, Okada said.

The region is blanketed with extremely sensitive seismometers that detect even the faintest temblor. But since earthquakes occur here far below the ground, unlike those in California that frequently rupture the surface, three research wells have been drilled two miles deep, penetrating the bedrock that underlies the region. Sensitive meters to measure strain in the earth’s crust, as well as any tilting due to seismic activity, have been placed at the bottom of the wells.

Expecting the ground to move gradually as strain builds toward the breaking point, 10 antennas have been scattered around the area. Orbiting satellites measure any movement between the antennas down to 1/250th of an inch.

Information from the sensors is automatically relayed to the research center here, as well as to the headquarters of the Japan Meteorological Agency in downtown Tokyo, which is responsible for sounding the alert.

At the first sign of a “sudden change” in the data, Michio Takahashi will pick up his telephone and call the six seismologists who sit on a special committee. The seismologists will rush to the agency and evaluate the data. If it appears that the earthquake is imminent, Takahashi will call the office of Japan’s prime minister, and an announcement will be made.

“All of that will take only two hours,” he said. “We plan to predict this earthquake.”

And what if an alert is called and nothing happens?

“You mean, if we strike out?” Takahashi asked.

At least, he said, they will have tried.

“It is better to swing where the ball isn’t than to get a strike without swinging,” he added.

Kyoto University’s Iemura sees a different scenario, however. Iemura believes that the first time the seismologists see a significant change they will not be sure enough that the great quake is imminent.

So instead of sounding a public alert, authorities will quietly move to reduce hazards, he said.

“They can reduce the water level in the dams,” he said. “They can shut down the atomic power plants and say they are down for maintenance. They can reduce the oil level in the fuel tanks. They can do all that without opening the information to the public.”

In the several years since the program began, it has never been necessary to do any of that because there has never been a hint that the great quake is near.

But Okada and others are convinced that it is only a matter of time.

“We have several kinds of convincing evidence that show the earthquake will attack this region,” he said.

The evidence also suggests that when it does hit, it will do so with gusto.

Tokai, like the rest of Japan, is on the Eurasia Plate, one of about a dozen major tectonic plates that drift around the surface of the Earth.

At Tokai, the Eurasia Plate is riding up over the Philippine Plate, which is being subducted into the Earth’s mantle. That is the process that created the volcanoes that became the islands of Japan, and it accounts for the volcanic and seismic activity throughout the region.

The Tokai area includes a sharply pointed peninsula that juts into the sea, and it is that peninsula that has Japanese scientists particularly worried. As the Philippine Plate slides under the Eurasia Plate, it pulls the peninsula down “several millimeters per year,” Okada said.

That process is called “crustal deformation,” and when the big quake finally comes, it is expected to release the strain that has been accumulating in the peninsula for 140 years.

“The next earthquake must allow several meters of rebound to recover the present crustal deformation,” Okada said.

That means the peninsula will suddenly leap up several meters in a process that will send deadly shock waves across the heart of Japan.

Furthermore, Okada said, the historical record clearly shows that a major quake is due at Tokai.

“In 1707, a very big earthquake destroyed all this region,” he said. “In 1854, we had two earthquakes within two days.”

The stage was set for the coming disaster in 1944 and 1946 when major quakes hit south of Tokai, destroying more than 100,000 homes.

Only the Tokai area was spared, he said. That means the strain that caused the 8.3 and 8.1 quakes in the ‘40s is still trapped in the Tokai region.

“Since it has already passed 140 years (since the last Tokai temblor) we must anticipate this earthquake,” he added.

Yet despite the overwhelming evidence that the Tokai quake is inevitable, some scientists in Japan believe that other areas are even further past due. Some contend an area closer to Tokyo is likely to erupt before Tokai.

That leaves open the possibility that the next great quake to hit Japan may not be predicted because most of the effort is concentrated in the wrong place.

But sooner or later, Tokai will erupt with a fury. And after all of this, the great quake may still come as a surprise.

That will mean, as Iemura put it, that, indeed, only God knows.

Imperiled Coastline 1 The Philippine Plate, one of the huge slabs of the Earth’s crust that drifts around the planet, is dipping beneath the Eurasia Plate south of Tokyo. 2. As the Philippine Plate “subducts” under the Eurasia Plate, it is dragging down a peninsula along the coast, instruments placed there reveal. Strain is accumulating in the peninsula as it is forced down. 3. An earthquake is expected to release the strain, causing the peninsula to jump up several yards. That, in turn, is expected to cause some buildings to topple. the sudden movement could also generate a tsunami that would send a giant oceanic wave crashing across parts of Japan. Earthquake Program Japan has the most ambitious program in the world aimed at predicting a major earthquake in time to evacuate several large cities. The program is based on the assumption that the region south of Tokyo will undergo some kind of geological changes before the quake hits. That could include foreshocks or a significant increase in strain in the rock formations in that area. Not all scientists believe the effort will succeed, but if it does it will offer some hope that many major quakes might be predicted in advance.