These tremors may signal an increased likelihood of earthquakes on the San Andreas fault. Scientists have long expected California's next big earthquake—of magnitude 8 or higher—to occur on the San Andreas fault.

"Because these tremors occur directly below the epicentral region of the 1857 quake, and because this part of the fault is locked and could rupture again soon, it is possible that increases in tremor activity may signal times of increased likelihood for a large event in the area," said Robert Nadeau of the Berkeley Seismological Laboratory in California

Nadeau is the lead author of the study, which is described in tomorrow's issue of the journal Science.

Chaotic Shaking 

The San Andreas Fault, the boundary between the North American and Pacific tectonic plates, runs 800 miles (1,300 kilometers) along the California coast and extends 10 miles (16 kilometers) down into the Earth. It is like a master fault in an intricate network of smaller faults that branch from the San Andreas and join it.

The scientists detected the tremors at a depth of 12.5 to 25 miles (20 to 40 kilometers) below the San Andreas fault near Cholame in central California, some 15 miles (25 kilometers) southeast of the town of Parkfield.

"The tremors are shaking of the ground that differ from earthquakes in that they last for up to 20 minutes, compared to earthquakes which last for less than 30 seconds," Nadeau said. "Unlike earthquakes, the tremor shaking is chaotic."

The discovery marks the first time such deep, non-volcanic tremors have been reported on the San Andreas fault, suggesting that the deformation causing earthquakes may have deeper origins than previously thought.

"Analyzing these tremors may help us to better understand deeper processes associated with earthquakes, therefore improving our knowledge of them, including our forecasting capabilities," said Andres Chavarria, a seismologist at Duke University in Durham, North Carolina.

Seismic Signals

California's two largest earthquakes—Fort Tejon in 1857 and San Francisco in 1906—occurred on the San Andreas fault, and scientists expect the next "big one" to occur on this fault.

"It certainly is the big dog," Nadeau said.

In recent decades, seismologists have taken particular interest in the Parkfield segment of the San Andreas Fault because of the moderate-size earthquakes (around magnitude 6) that occur there at fairly regular intervals.

Seismologists have installed a dense network of instruments in the Parkfield area to learn more about the earthquakes. Researchers have placed seismometers in bore holes 100 to 200 yards (90 to 180 meters) deep, away from noise sources that typically mask seismic signals from small earthquakes.

"The high sensitivity and low noise level of seismic signals recorded by this network were also extremely helpful for the tremor analysis we performed," Nadeau said.

On September 28, nine months after the scientists ended their analysis, another magnitude 6 earthquake struck the area close to the tremor region.

"Many kinds of instruments were able to record [this event], providing an invaluable source of data for earthquake studies," Chavarria, the Duke University seismologist, said.

Predicting Quakes

The discovery in 2002 of non-volcanic tremors in Japan represented a major advance in seismology. Fluids from diving ocean bottom rocks were believed to be critical for generating the tremors in the subduction zones of Japan and the North American Pacific Northwest.

However, the San Andreas Fault does not subduct. As a result, no source of fluids from the rock of a descending tectonic plate is available.

"Consequently, no one expected tremors to occur beneath the San Andreas Fault and nobody really looked for them systematically," Nadeau said.

Indeed, the tremors may have been occurring all along, while scientists simply failed to detect them. The new findings may stimulate the search for tremors more generally throughout California. Researchers hope that tremors will prove to be a useful tool for forecasting larger earthquakes.

"We still have only a rudimentary understanding of the process that creates the tremor and its significance for earthquakes," said William Ellsworth, chief scientist for the Earthquake Hazards Team at the U.S. Geological Survey in Menlo Park, California.

"This discovery opens an important new window into the workings of the San Andreas Fault that will undoubtedly become a major research focus in the coming years," he added.