Patterns of deep, prolonged tremors newly revealed beneath the San Andreas fault zone may offer scientists a way to foretell earthquake activity there.
The small tremors don’t produce typical seismic vibrations that indicate a sudden slip along a fault, says Robert M. Nadeau, a seismologist at the University of California, Berkeley. Instead, the deep tremors gradually rumble to life. During a 3-year period that ended in December 2003, instruments detected 110 such events along a 30-kilometer-long stretch of fault centered just southeast of Cholame, in south-central California.
The tremor episodes, all of which were too weak to be felt by people, originated at depths of 20 to 40 km and lasted from 4 to 20 minutes, says Nadeau. In contrast, sudden slips along the San Andreas fault typically occur at depths of less than 13 km and last less than a minute.
Nadeau and his Berkeley colleague David Dolenc described the new tremor findings last month at the fall meeting of the American Geophysical Union in San Francisco. Their findings are also slated to appear in the Jan. 21 Science.
In the past 2 years, scientists have observed similar long-duration tremors along subduction zones at the edges of the Pacific Ocean. In these regions, the convergence of tectonic plates pushes one plate—usually, dense ocean crust topped with water-soaked sediment—beneath another one.
That action injects large amounts of moisture into Earth’s mantle, and the resulting fluids may strongly affect tremor activity, says Nadeau. However, he adds, there’s no evidence of moisture acting along the San Andreas fault, whose two sides scrape horizontally past one another at Earth’s surface.
After Nadeau’s presentation in San Francisco, Stephen H. Kirby of the U.S. Geological Survey in Menlo Park, Calif., commented that the San Andreas fault about 30 million years ago was a subduction zone. Kirby noted then that his analyses suggest that the crust then carried down into Earth’s mantle could still be releasing water.
The San Andreas’ deep tremors took place at varying intervals during the 3-year observation period, says Nadeau. Changes in their rate of occurrence seem to take place a couple of months before changes in the rate of shallower earthquakes, he notes. Indeed, Nadeau suggests that a sudden flurry of deep tremors near Cholame in early September 2004 may have presaged the magnitude-6.6 quake that struck about 40 km away, near Parkfield, Calif., on Sept. 28.