Predicting Quakes: a Shaky Art

By Frederic Golden

Why seismologists usually cannot tell when shocks mil strike

On a bitterly cold day in February 1975, the Communist Party chiefs in Haicheng, a city in northeastern China, convened in emergency session. They had just learned from seismologists that various signs, including changes in the water level of wells, pointed to an imminent earthquake. Convinced the threat was real, they issued an order to evacuate all homes and hospitals, close down shops and factories and turn off gas mains and furnaces.

That night, while thousands huddled in open fields or watched special film showings in the city square, the earth began to rumble. A quake measuring 7.3 on the Richter scale devastated Haicheng, collapsing buildings, turning bridges into grotesquely twisted heaps and tearing up roads. The death toll could easily have run into six figures. Thanks to the forewarning, it was probably fewer than 300.

Scientists hailed that prediction as one more indication that quake forecasting could become as reliable as weather forecasting. Such optimism has proved premature. True, other quakes have been predicted in China and elsewhere. But more often than not they have struck without specific warning, as in Italy last week, and in Algeria last October. Only a year and a half after the Haicheng temblor, an 8.2 quake near Tangshan, 90 miles southeast of Peking, caught seismologists by surprise and killed as many as 650,000. Says Polish-born Volcanologist Haroun Tazieff: "At the present level of research, nature almost always surprises man."

Even so, the earth's behavior is becoming much less mysterious. As late as 1750, the Bishop of London told his flock that two recent quakes were warnings from an angry deity. Today, scientists prefer another explanation, an all-encompassing view of the earth known as the theory of plate tectonics. It holds that the planet's surface consists of a dozen or so restless plates, each about 70 miles thick. Their movements explain volcanoes, the rise of mountains and the drift of continents. They account for quakes as well, most of which seem to occur where the great plates meet--at the so-called Ring of Fire, for instance, the tremor-and volcano-prone region that rims the Pacific Ocean and outlines the Pacific plate.

Driven by as yet unfathomed forces deep within the earth, the plates move no more than a few inches a year, bumping, jostling and grinding against each other with incredible force. These movements can have enormous consequences. Opposing plates often lock, so that great stresses begin to build up. When the pressures become so large that they exceed even the strength of the rock, the earth fractures, frequently along "faults" where earlier breaks have occurred. Like a wound spring suddenly uncoiling, the earth releases its stored energy in shocks that may be felt far beyond the fracture. This is an earthquake.

At the turn of the century, some scientists proposed measuring the buildup of stresses along known fault lines to determine when the rock might be nearing the breaking point. More recently, they have begun to look for other signals of an impending quake, such as changes in the ability of the rock to convey sound waves, local alterations in the earth's surface tilt and magnetic field, and increases in the release from the ground of radioactive radon gas, a phenomenon associated with the fissuring of rock under pressure. Experts have even begun to take seriously reports that animals behave skittishly before quakes.

While such data have been used to foretell tremors accurately, they can also be misleading. Because the earth's geology varies greatly, rock in one place might behave in a much different way than rock does elsewhere. Also, monitoring seismic data requires networks of field stations that can automatically pipe information to central analysis points, something few countries can afford. In Italy and indeed throughout the Mediterranean region, predictions are complicated by yet another factor: the area is the meeting place of not just two plates but three or more. Italian Geophysicist Forese Wezel describes the region as "a terrible geological crossroads."

The commonly accepted view of the region's geology is that the plate bearing the African continent is generally pushing north and trying to slip under the Eurasian plate. That movement, under way for millions of years, created the Alps and has helped to stoke such volcanoes as Sicily's Mount Etna. The devastating 7.2 quake that leveled the Algerian city of El Asnam nearly two months ago, killing more than 2,500, occurred almost precisely at one of the points where the African and Eurasian plates are believed to be thrusting against each other. But the Italian peninsula is also being wrenched by other forces within this broad pressure pattern. It is being pinched from the west by the Eurasian plate and from the east by a subdivision of the African plate known as the Apulian. Presumably, strains from these opposing forces caused the Italian quake. Says Florence-born Seismologist Leonardo Seeber, now at Columbia University's Lamont-Doherty Geological Observatory: "In Italy we are even worse off, in that we understand very little about the tectonics there."

To their credit, Italian seismologists, who operate a national network of 54 seismic stations, had warned that a major quake might hit sometime after 1978.

They based their prediction on what scientists call the gap theory: that a quake can be expected in any part of a seismically active region where one has not occurred in the past 50 years or so. Understandably, the forecast was so vague that it was ignored and no preparations were made for the disaster that was, in deed, impending.

By Frederic Golden. Reported by Philip Faflick/New York

With reporting by Philip Faflick

This file is automatically generated by a robot program, so viewer discretion is required.