Seismic Doubleheader: Seismologist Shows Deep Earthquakes Come In Pairs

The second earthquake generally has the same magnitude and follows the first within one day. These earthquake doubleheaders are seldom found for earthquakes at shallow depths, such as along the San Andreas Fault. These results may imply that earthquakes deeper than 300 miles do not represent brittle slip along a fault, as do shallow earthquakes that can be studied more readily.

The results were published in the August 24 issue of Science. The research was supported by the National Science Foundation. "We noticed some deep earthquake seismograms were identical, showing precisely the same pattern of wiggles each time. Our research then showed that these earthquakes were occurring in the same spot repeatedly," he said. "Because these seismograms were so similar we could tell that part of the fault was slipping over and over again."

The cause of earthquakes deeper than 50 miles has been controversial for several decades. One theory on the mechanism of deep earthquakes is called transformational faulting. In this model, earthquakes occur when material at high pressure deep in the earth undergoes a phase transformation, similar to when carbon is converted to artificial diamonds in a laboratory. Previous studies had suggested that if deep earthquakes were due to transformational faulting, they should not recur at the same location. But Wiens' observation shows that deep earthquakes often recur at the same location, making this theory less likely.

Wiens said the find supports the ductile shear zone model. In this idea, deep earthquakes are very sensitive to the temperature along a slipping zone. During a deep earthquake, a lot of heat is produced by viscous dissipation, which is similar to friction. This heat then increases the chances of further earthquakes at the same location. "The data argues against phase transformation," Wiens said. "In phase transformation, after one earthquake happened the materials would have transformed and you couldn't have an earthquake at the same spot. We favor the ductile shear zone model where the heat generated by the first earthquake seems to make the situation favorable for future earthquakes in a very short time."

Wiens and Washington University undergraduate student Nathaniel O. Snider used data collected six years ago from an array of 11 seismographs in the Tonga and Fiji islands, They identified three groups of nearly identical earthquakes, each one composed of between 10 and 30 earthquakes. Wiens and his colleagues recently installed a 26-seismograph array for future data.

"The repeating earthquake seismograms allow us to locate the earthquakes to accuracies of less than 1 mile, even at depths of 400 miles", Wiens said. "This will allow us to study deep earthquakes by mapping out the individual fault zones, much as geologists have mapped out many small fault zones near the earth's surface to understand the San Andreas Fault system."