Predicting the Quake

Since man first appeared on this planet, he has been at the mercy of earthquakes, which over the ages have devastated wide areas and killed millions of people. While scientists may never learn to prevent quakes, they may soon be able to forecast them accurately, giving inhabitants a chance to flee a threatened area in time. Early this month, for the first time, scientists predicted an earthquake--and then felt it rumble beneath their feet right on schedule.

The earth tremor, which occurred in the Blue Mountain Lake region of the Adirondack Mountains of New York State, was forecast by Yash Aggarwal, 33, a seismologist at Columbia University's Lamont-Doherty Geological Observatory. Aggarwal and another Lament scientist, Lynn Sykes, began to study the Blue Mountain Lake area two years ago, intrigued by the fact that in a generally calm region it experienced frequent small tremors. In mid-July, when two moderate quakes jolted the area, Aggarwal and colleagues from Lamont set up seven portable seismographs in addition to a permanent station already in place. For two weeks, Aggarwal drove 100 miles every day to check the instrument readings. His routine paid off. On Aug. 1 he telephoned Sykes, head of the seismology group at Lament, and predicted that a quake of magnitude 2.5 or more should be coming "in a couple of days." Two days later, the earth dutifully complied with an earthquake that registered about 2.5 on the Richter scale. "I was so excited," recalls Aggarwal, an Indian who was born in Kenya, "that I nearly drove into a tree."

Geophysicist Gordon Greene, of the U.S. Geological Survey, was equally enthusiastic. He had been with Sykes when Aggarwal phoned in his forecast and had driven to Blue Mountain Lake just in time for the quake. "If you can do this three times," he told Aggarwal, "you will all be famous."

The prediction technique was devised independently by the Lamont researchers (TIME, Feb. 12) and Stanford University Scientist Amos Nur. It is based upon a sudden cracking and expansion of rock along a fault zone in the earth when stresses reach a critical point. This cracking creates many tiny cavities in the water-saturated rock. That slows the passage of P (pressure) waves, which travel faster through liquid-filled cracks. Another kind of seismic wave, the S (shear) wave, however, is less affected by the newly opened cracks; thus the usual ratio of P-to S-wave velocity drops sharply. Then, as ground water gradually seeps into the new cracks, the ratio returns to normal. But the water increases pressure within the rock, causing one wall of the fault to slide along the other. It is this slippage that creates the shock. In a paper submitted to Science, Sykes, Aggarwal and Christopher Scholz assert that in the Blue Mountain region the seismic-wave phenomenon occurs before every sizable earth tremor.

The P-and S-wave phenomenon on which the Lamont technique is based was first reported in 1969 by Soviet scientists working in the Garm region of southern Russia. It was later shown to have occurred before earthquakes in Ja pan and the disastrous quake near Los Angeles that destroyed half a billion dollars worth of property in 1971. Whether the same prequake behavior occurs along key segments of the San Andreas fault region has not yet been determined; San Andreas rock formations are different from those at Blue Mountain. Nonetheless, the Lamont scientists suspect that the phenomenon may be widespread, and that, combined with other geological clues, it will eventually enable seismologists everywhere to call the time, place and size of many quakes. "An earthquake just can't occur without warning," says Aggarwal. "It's too big a thing."

This file is automatically generated by a robot program, so reader's discretion is required.