Discovery of Slow Quakes Stirs Scientific Tremors

Most people think of an earthquake as a fearsome release of earth energy measured in seconds--still long enough at its most severe to topple the tallest buildings or remake a mountain range. But in research made public Wednesday, scientists at the Carnegie Institution of Washington show how some earthquakes can last hours, even days.

Indeed, they present evidence of a magnitude 4.8 earthquake along the San Andreas fault in 1992 that lasted a week--unnoticed by the people who live along the famous fault line.

The earthquake, which occurred near San Juan Bautista in Central California, was a hundred times slower than any previously detected--so slow that it did not even generate conventional seismic waves.

The idea of a slow earthquake is a contradiction in terms--like jumbo shrimp. And the significance of this newly discovered phenomenon already is being debated.

Some scientists suspect that this kind of slow-motion underground turmoil actually may trigger the chain of events that leads to more conventional--and destructive--earthquakes.

Other experts, however, suggest the opposite: Slow earthquakes may prevent large temblors by harmlessly venting the powerful forces that build up as the Earth’s massive tectonic plates grind against or slide underneath one another.

In areas like Southern California and Japan, experts warn of an earthquake “deficit” because there have been too few temblors to relieve the pent-up seismic strain generated by plate motion.

Quake researchers and emergency planning officials have argued over whether Southern California should brace itself for a series of devastating major earthquakes in the next few decades, which would release the titanic energy building up in faults crisscrossing the region.

But the new finding suggests to some that perceptible, slow earthquakes may be a previously unsuspected safety valve.

“It is a very interesting finding that has an important implication,” said Hanoo Kanamori, director of Caltech’s seismological laboratory. “It shows there is a place in the crust where slippage can be slow” and Earth’s potentially destructive energy can be released safely.

Researchers have been tantalized for years by hints of such unusual earth tremors. A slow earthquake was detected just before a 9.5-magnitude temblor in Chile in 1960. A second was detected after a magnitude 7 earthquake southwest of Tokyo in 1978. And traces of a third slow earthquake were recorded before a 1992 7.2-magnitude quake in Nicaragua.

But none of these were definitively linked to those larger quakes, nor were they as distinctive or as slow as the San Andreas event reported in the current issue of the journal Nature. The slow quake near Tokyo lasted two hours and generated aftershocks as large as magnitude 5.8; the Nicaraguan event lasted three minutes--about three times longer than normal.

Slow earthquakes almost always elude detection, even with growing networks of sensitive seismographs and satellite sensors. So scientists usually must infer them after the fact through more indirect means.

Researchers cannot use seismometers to detect them because such quakes usually do not generate seismic waves. The ground movement at the surface is often so small that even sophisticated measuring techniques, like the satellite-based Global Positioning System, can miss it.

The slow earthquake on the San Andreas was detected serendipitously by two strain meters that had been installed in bore holes near the fault in 1984. Scientists wanted to monitor an area where the rocks on either side of the fault line often “creep” past each other without the need of an earthquake to push them along.

In December 1992, the instruments picked up signals generated as 11 square miles of the fault surface slowly ruptured over a week to 10 days, the researchers reported in Nature. The sides of the fault moved about an inch, which would be equivalent to the movement cause by a magnitude 4.8 earthquake, the researchers said.

“We had never seen this sort of thing before,” said Alan Linde, a staff scientist in the Carnegie department of terrestrial magnetism, who conducted the study. “We observed these huge, slow changes [in the instrument readings].”

Although the slow quake did not show up on seismographs, it did generate a series of more conventional aftershocks, ranging up to magnitude 3.7.

To double-check their readings, Linde and his colleagues monitored the site for three more years. The temblor was much deeper and much more powerful than any known creep event, putting it in a category all its own, the researchers said.