Quake Prediction Technique Shows Promise : Science File / An exploration of issues and trends affecting science and the environment

A string of loosely successful predictions of earthquakes in Greece, based on the monitoring of electrical signals in the ground, is stirring the interest of seismologists and engineers in many countries. An attempt to duplicate the process is under way in California.

The journal Science recently reported that over the last nine years, 10 of the 11 earthquakes of magnitude 5.8 or greater that occurred within the area being monitored had been predicted by Greek physicist Panayiotis Varotsos and his team.

The predictions, however, have been somewhat vague about time, place and magnitude. The scientists making them have frequently been able to place the epicenter of a coming earthquake within only 40 to 60 miles of its location. On other occasions, the date of the prediction has been off by three weeks or more.

Still, a number of scientists believe that progress has been made, and that the Varotsos team’s track record rules out coincidence.

*

Hiroo Kanamori, director of the Caltech Seismological Laboratory, who has been pessimistic about earthquake predictions, said he sees promise in the Greek work. “There is some very important physical mechanism behind it,” he said in a Times interview. “It could lead to the understanding of the earthquake process.”

The Greek scientists have been using electrodes in the ground spaced two to three miles apart to measure the changing voltage of electrical currents in the Earth’s crust, and interpreting the data to predict the earthquakes.

In California, changes were recorded of magnetism, not electrical currents, just before the 1989 Loma Prieta earthquake.

*

Antony Fraser-Smith, Stanford professor of electrical engineering, noted protracted fluctuations in the Earth’s magnetic field for a month before the Loma Prieta quake--reaching up to 600 times normal levels in the three to four hours before the quake--while he was conducting unrelated experiments within three miles of what turned out to be the epicenter.

Fraser-Smith observed that the bursts of electrical current recorded in Greece are different in character.

“They tend to have rather short bursts, or a series of bursts, but lasting only an hour or less, maybe 20 days to a month before the earthquake,” he said.

“The resulting predictions are not as specific as you’d like,” he said. “Varotsos leaves open the timing by two weeks, and the magnitude and location are a little ambiguous, but he actually does get all three roughly right and nobody else in the world is anywhere close to that at the moment, so I’m willing to give him a little credit.”

In October, scientists at UC Berkeley agreed in a meeting to pursue study of electrical currents and magnetism in an attempt to identify earthquake precursors in California. The work will be under the direction of H. Frank Morrison, a professor of geophysical engineering.

A Berkeley colleague of Morrison, seismologist Thomas McEvilly, said that although “there’s a lot of room for arguing about the precision” of the Greek predictions, “their measurements seem sound and it’s difficult to explain away as chance their relationships to earthquakes.”

“It really is a matter now of making measurements as best we can and see what we see,” he said.