Study of Earthquake Zone Helps Set the Stage for Drilling Project

In a first-of-its-kind study, seismologists have used tiny "microearthquakes" along a section of California's notorious San Andreas Fault to create unusual images of the contorted geology scientists will face as they continue drilling deeper into the fault zone to construct a major earthquake "observatory."

A chain of 32 seismometers recorded the small earthquakes at underground locations along a 7,100-foot-deep vertical drill hole. This eight-inch-diameter pilot hole was excavated last year about 1.1 miles southwest of the San Andreas Fault to monitor earthquake activity and assess the area's underground environment before drilling the main hole. After more vertical drilling at the same location next summer, the main hole will be angled off towards the northeast to pierce the fault zone itself.

In a paper in the Friday, Dec. 5, 2003 issue of the research journal "Science," researchers from Duke and the United States Geological Survey (USGS) described how they used seismic signals and computer analysis to derive outlines of what may be secondary faults, and perhaps fluid filled cracks, in subterranean locations between the main fault and the pilot hole.

The scientists created those color-coded images by tracing the complex paths earthquake waves took after scattering off the possible underground structures. The scientists' goal was to pinpoint not the earthquake sources but rather features between those sources and the detecting seismometers.

"The seismic waves that we recorded are telling us a story about what's happening within the Earth," said Andres Chavarria, a senior graduate student in seismology at Duke's Nicholas School of the Environment and Earth Sciences, who is first author of the "Science" paper.

"This is the first seismic study to employ an array of deeply placed seismometers and nearby microearthquakes to study a major active fault zone," added Duke seismology professor Peter Malin, another author of the study. "It's an important piece of information that has to be taken into account for understanding the mechanics of the fault zone at depth," Chavarria said. "Our results show that this part of the San Andreas Fault is a more complex zone than previously thought."

Other authors included Duke research scientist Elyon Shalev and Rufus Catchings of the USGS in Menlo Park, Calif. Catchings was a visiting scholar working with the Duke team when the main ideas behind the study were conceived. The study was supported by the National Science Foundation (NSF).

The San Andreas Fault extends about 800 miles on land from near the Mexican border to north of San Francisco Bay and is identified as the source of numerous devastating earthquakes, including the one that decimated San Francisco in 1906. These quakes are believed to be caused by sudden releases of stress between two huge blocks of the Earth's crust, the westernmost of which is slowly grinding northwestward past the other.

In parts of the San Andreas, the western block slips quietly by the eastern block in what scientists call an "aseismic creep." But other earthquake-producing sections remain locked up for long periods before dramatically letting go.

The San Andreas Fault Observatory at Depth (SAFOD) is being jointly managed by scientists from Stanford University and the USGS and is receiving more than $20 million in federal support as part of NSF's new EarthScope initiative (see www.earthscope.org). SAFOD will be located near the tiny town of Parkfield, Calif., which bills itself "the earthquake capitol of the world" because of its unusual history of repeating earthquakes of significant intensity.

Co-author Malin installed in the SAFOD pilot hole seismometers he helped design, each instrument built to detect vibrations in three directions and also withstand the 220 degree Fahrenheit temperatures and 300 atmosphere pressures possible in the pilot deep hole's reaches.

Because the instruments are below ground and thus shielded from surface noise, they can detect the very small microearthquakes in the fault's vicinity that occur much more frequently than larger ones.

The new "Science" paper describes the authors' analyses of "secondary signals" emanating from 43 such microearthquakes, plus 11 additional surface test explosions. The microearthquakes and the human-produced shocks each originated within about five miles of the pilot hole. The paper proposes that that the secondary signals represent energy from the disturbances that was then scattered-off underground structures before arriving at the pilot hole's detecting seismometers.

Chavarria compared the analytical technique to medical computerized tomography -- CAT scans - which creates internal anatomical images by tracing the paths of x-rays passing through the body at different angles. Another analogy might be ultrasound, a technology that uses computers and sound waves to produce such images. "In seismology we do something similar, analyzing seismic signals to give us a kind of tomography of the Earth," he said.

The scattering zones as imaged by the analysis "may represent secondary faults in the San Andreas Fault zone," the authors wrote in "Science." These results imply that continued drilling towards the fault zone for the SAFOD project "will encounter at least one significant structure, quite likely another fault, before reaching it," the paper added.

There is also evidence to corroborate results of a 1997 study by Malin and others that suggested the possibility of fluid-filled cracks in the fault's main branch, the authors wrote.

Duke senior Charles Stopford worked with the scientists last summer in California to test new pilot hole instruments. Two other Duke graduate students, Henrique Tono and Carolina Pungiluppi, have also done work at the site.