Study of 8.7-Magnitude Earthquake Lends New Insight into Post-Shaking Processes

PASADENA, Calif.—Although the magnitude 8.7 Nias-Simeulue earthquake of March 28, 2005, was technically an aftershock, the temblor nevertheless killed more than 2,000 people in an area that had been devastated just three months earlier by the December 2004, magnitude 9.1 earthquake. Now, data returned from instruments in the field provide constraints on the behavior of dangerous faults in subduction zones, fueling a new understanding of basic mechanics controlling slip on faults, and in turn, improved estimates of regional seismic risk.

In the June 30 issue of the journal Science, a team including Ya-Ju Hsu, Mark Simons, and others of the California Institute of Technology's new Tectonics Observatory and the University of California, San Diego, report that their analysis of Global Positioning System (GPS) data taken at the time of the earthquake and during the following 11 months provide insights into how fault slippage and aftershock production are related.

"In general, the largest earthquakes occur in subduction zones, such as those offshore of Indonesia, Japan, Alaska, Cascadia, and South America," says Hsu, a postdoctoral researcher at the Tectonics Observatory and lead author of the paper. "Of course, these earthquakes can be extremely damaging either directly, or by the resulting tsunami.

"Therefore, understanding what causes the rate of production of aftershocks is clearly important to earthquake physics and disaster response," Hsu adds.

The study finds that the regions on the fault surrounding the area that slipped during the 8.7 earthquake experienced accelerated rates of slip following the March shock. The region dividing the area that slipped during the earthquake, and that which has slipped after the earthquake, is clearly demarcated by a band of intense aftershocks.

A primary conclusion of the paper is that there is a strong relationship between the production of aftershocks and post-earthquake fault slip-in other words, the frequency and location of aftershocks in a subduction megathrust are related to the amount and location of fault slip in the months following the main earthquake. Hsu and her colleagues believe that the aftershocks are controlled by the rate of aseismic fault slip after the earthquake.

"One conjecture is that, if the aseismic fault slip occurs quickly, then lots of aftershocks are produced," says Simons, an associate professor of geophysics at Caltech. "But there are other arguments suggesting that both the aftershocks and the post-earthquake aseismic fault slip are caused by some third underlying process."

In any case, Simons and Hsu say that the study demonstrates that the placing of additional remote sensors in subduction zones leads to better modeling of earthquake hazards. In particular, the study shows that the rheology, or mechanical properties, of the region can be inferred from the accumulation of postseismic data.

A map of the region constructed from the GPS data reveals that certain areas slip in different manners than others because some parts of the fault seem to be more "sticky." Because of the nature of seismic waves, the manner in which the fault slips in the months following a large earthquake has huge implications for human habitation.

"An important question is how slip on a fault varies as a function of time," Simons explains. "The extent to which an area slips is related to the risk, because you have a finite budget. Whether all the stress is released during earthquakes or whether it creeps is important for us to know. We would be very happy if all faults slipped as a slow creep, although I guess seismologists would be out of work."

The fact that the Nias-Simeulue's postseismic slip following the December 28, 2004, earthquake can be modeled so intricately shows that other subduction zones can also be modeled, Hsu says. "In general, understanding the whole seismic cycle is very important. Most of the expected hazards of earthquakes occur in subduction zones."

The Tectonics Observatory is establishing a network of sensors in areas of active plate-boundary deformation such as Chile and Peru, the Kuril Islands off Japan, and Nepal. The observatory is supported by the Gordon and Betty Moore Foundation.

The other authors of the paper are Jean-Philippe Avouac, a professor of geology at Caltech and director of the Tectonics Observatory; Kerry Sieh, the Sharp Professor of Geology at Caltech; John Galetzka, a professional staff member at Caltech; Mohamed Chlieh, a postdoctoral scholar at Caltech; Danny Natawidjaja of the Indonesian Institute of Sciences; and Linette Prawfrodirdjo and Yehuda Bock, both of the University of California at San Diego's Institute of Geophysics and Planetary Sciences.