NEWS NOTES — NEWS
Small earthquakes and tiny tremors originating deep in fault zones are the result of slow earthquakes at Earth’s surface, according to a new study. Such slow earthquakes may also be harbingers of the shallower, high-magnitude and highly destructive earthquakes that can occur in subduction zones, such as the December 2004 Sumatran earthquake.
“Non-volcanic tremor,” a weak seismic signal that was first identified in 2002 by the Japanese seismologist Kazushige Obara, was “the first new source of seismic waves to be discovered in half a century,” says Bill Ellsworth, a seismologist at the U.S. Geological Survey in Menlo Park, Calif. Tremor is different from more familiar earthquake seismic signals in that the signals do not include P- and S-waves, which seismologists use to pinpoint the epicenter of an earthquake. The signals can also last much longer, with individual bursts lasting as long as 15 minutes, compared with the few seconds of an earthquake. A series of small, low-frequency earthquakes, of magnitudes 1 and 2, also often occur along with tremor.
Since its discovery, Ellsworth says, tremor has intrigued seismologists because it is often associated with a type of silent, slow slip along a fault occurring 30 to 35 kilometers deep on the dipping slab in a subduction zone. Non-volcanic tremor has also been detected in the northwestern United States’ Cascadia subduction zone and under California’s San Andreas Fault. Scientists are particularly interested in whether these slow events, which can last up to a week or more but do not generate strong shaking, may signal stress building up that could ultimately lead to major seismic hazards.
“We’ve been working on the relationship between tremor, slow slip events and low-frequency earthquakes,” says David Shelly, a graduate student at Stanford University.
Observing a strong resemblance between the shapes of seismic waves from tremor and low-frequency earthquakes, Shelly and his colleagues set out to determine whether the two are separate phenomena, or are more closely related. The team took seismic waves from more than 650 low-frequency earthquakes recorded over three years by at least three stations near Shikoku, Japan, and then compared them with tremor signals, looking for matches in the waveform patterns.
The data showed that not only are low-frequency earthquakes and tremor related, but that tremor could be regarded as a “swarm” of those earthquakes, generated by small slip events, the team reported March 15 in Nature. “Tremor is just composed of a whole bunch of these low-frequency earthquakes put together,” Shelly says. “When you add up the slip on a fault, it may be equivalent to a magnitude-6 or -6.5 earthquake, but since it happens very slowly it doesn’t generate strong shaking and is imperceptible to people.”
The connection between these events could improve future earthquake probability predictions, Shelly says. A slow slip event does not mean that a major earthquake will occur at a given time, he says, but “when slip is happening on the deep part of these faults, it may load the stress on the shallower fault and bring it closer to failing. So it’s possible that during these events, the probability of getting a major earthquake shortly after on the same fault is increased.”
Shelly’s work “presents some very exciting results,” Ellsworth says, although he agrees that it is too soon to make probability estimates based on slow slip events without a better understanding of how earthquakes begin and propagate. For one thing, research like Shelly’s into slow slip events has shown that the mechanisms governing such deep earthquakes are “very different than the behavior of the fault system at shallower depths,” he says.
At the Nankai subduction zone in Japan, for example, an earthquake will occur about once every 100 years, “and run away to magnitude 8,” Ellsworth says. But “just another 30 kilometers deeper, on the same subduction interface, there are millions of tiny earthquakes being generated all the time, and none of them run away.”