Four years ago, two scientists warned that Southern California was entering a period when major earthquakes were likely to occur, and that the window of vulnerability would peak this month.
The scientists postulated that major quakes on certain types of faults in this region may be linked to a rare juxtaposition of the Earth, sun and moon, and the twin earthquakes that struck in the Imperial Valley near sunset on Monday and around sunrise on Tuesday would seem to support their theory.
So one would think the scientists would draw some intellectual satisfaction from this week's temblors, but at least one of them was considerably less than overjoyed.
"It's making us look good for the wrong reasons," UCLA geophysicist Leon Knopoff said. "We're not seers."
It's not that Knopoff now thinks he was wrong. It's just that he is still not sure he was right, and he doesn't want anybody to think that he thinks he really knew what was going to happen this month.
So in a sense, the Imperial Valley temblors, which struck exactly when his research indicated this region would be most susceptible to an earthquake, couldn't have come at a worse time.
However, some other scientists said Tuesday that the timing of the Imperial Valley quakes, as well as the Whittier earthquake that struck soon after sunrise on Oct. 1, will make them take another look at Knopoff's earlier conclusions.
Scientists have long sought to determine if the gravitational tug of the sun and the moon triggers earthquakes. Both bodies cause the powerful tides that force the oceans to rise and fall, and in this region of the Earth, even the solid ground lifts and drops by as much as half a foot a day.
But over the years, evidence generally has not supported any link between lunar and solar cycles, and scientists who tried to establish such links have often found themselves the subject of ridicule. And beneath the hostility, one scientist suggested, was the feeling that any such conclusions smacked of astrology.
Even Knopoff had once published an article debunking the concept.
But in a 1983 article in the British science journal Nature, Knopoff and astronomer Steven Kilston reported that, after analyzing major earthquakes in Southern California that occurred along north-south faults, they found some correlation between the positions of the sun and the moon and many of those earthquakes. They eliminated all smaller quakes from their study, because most smaller quakes are either foreshocks or aftershocks of larger temblors.
Kilston, who has since left the area and could not be reached for comment Tuesday, studied the records of all major earthquakes in this region back to 1857 and found that they seemed to occur in cycles of 18 to 19 years--including the killer quakes of 1857, 1933, 1952 and 1971.
As an astronomer, he knew that the moon wobbles in its orbit around the Earth by about five degrees, and that it takes 18.6 years for the moon to complete one "wobble." Thus, every 18.6 years, the moon reaches the northernmost point in its orbit around the Earth and could have a strengthened effect on northwest-trending faults.
He also found that 10 out of 13 major earthquakes had occurred near either sunrise or sunset, when the sun would exert its greatest influence on the surface.
When the moon is either full or new, it is aligned with the sun, thus combining lunar and solar gravitational influences on the Earth. And once every 18.6 years, the moon would be at its northernmost point, thus giving an extra tug on any fault running in a northwesterly direction, like the San Andreas, Kilston reasoned.
Four years ago the scientists concluded that the next time these circumstances would occur was this November--so the area could be ripe for an earthquake stronger than 6 on the Richter scale within three or four years on either side of that date.
Both men stressed during interviews with The Times in April, 1986, that they were not predicting an earthquake for November, 1987. That month would simply mark the peak of the window of vulnerability.
"We're not predicting an earthquake in November of 1987," Knopoff said flatly in the Times article on April 6, 1986.
However, some other news reports did not include such a disclaimer, he said Tuesday, and his life has not been the same since.
"This thing has been the bane of my existence," Knopoff said.
He said he has finally convinced most of his colleagues that he really had not predicted an earthquake for November, 1987, and then "as luck would have it," the area was hit with back-to-back quakes. One at sunset. The other at sunrise. Both on a fault system that runs northwesterly.
Knopoff conceded Tuesday that the Imperial Valley quakes--which were so close together that he would consider them "one event"--"adds further weight to the statistics" that led him to conclude that there is probably a correlation between lunar and solar cycles and major quakes.
Must Be Considered
Some other highly regarded seismologists said that the timing of the quakes cannot be dismissed lightly.
"I'm going to have to take another look at that," said Caltech's Clarence Allen, referring to the 1983 study by Knopoff and Kilston.
"It has to be taken seriously," noted UC Berkeley's Bruce Bolt. "The gravitational force is quite large."
"As they are well aware, the difficulty is that the data is not very long in the sense that we only have about 180 years of historical record, and only about 50 years of instrumented data. The uncertainties are large." No one is more aware of that than Knopoff, who sounded Tuesday a little like a man who wished he had kept silent.
The Imperial Valley quakes, he said, add up to "just another number in a weak statistic."
Just in case, mark May, 2006, on your calendar.