PASADENA—For geophysicists, the 7.3–magnitude Landers earthquake of June 28, 1992 has yielded much in terms of understanding the basic mechanisms of seismic events. A new study appearing in this week's Science provides a new model to explain why the ground near the fault gradually shifted the first few years after the main shock. The work could be used in the future for the analysis of earthquake hazard.
In the Science article, Jishu Deng, a postdoctoral researcher at the California Institute of Technology, and his coauthors attribute the postseismic deformation to a viscous flow in the lower crust. Experts have known for some time that such slow motions around faults can occur, and in fact were quite aware of the effect near the Emerson fault on which the Landers earthquake was centered. But no one knew whether the ground was moving in small, quirky steps or slowly flowing like a viscous liquid.
Analyzing existing data from various satellites, Deng speculates that viscous flow must be the case, even though the "afterslip model" has for some time been the preferred explanation. Deng believes the "viscoelastic model" is preferable because the satellite data shows both a horizontal motion along the Emerson fault over about three or four years, as well as a vertical motion. While the viscoelastic model is not completely new, previous studies have been unable to distinguish between the viscoelastic and afterslip models. The Landers earthquake, however, provides the first opportunity to determine which mechanism is indeed at work.
Specifically, the area just west of the north–south fault has continued to move northward since the initial rupture. On the day of the earthquake, the fault slippage was measured to be about five to six meters along the fault line. But the GPS satellites show that the displacement has gradually expanded another 10 centimeters or so.
This continued slippage can be explained by the prevailing theory of postseismic slippage, but an additional result calls for a new theory: according to information gained from the Interferometric Synethetic Aperture Radar satellite (the ERS-1), the ground to the west of the fault has also sunk by about 28 millimeters, while ground east of the fault has risen slightly. And because the afterslip model cannot explain this motion, Deng shows that the effect must be the result of viscous flow.
"So we think the fault is not slipping," says Deng, who came to Caltech after earning his doctorate at Columbia University. "It must be in a flow." Deng further says the new information could be used in the future to assess the seismic hazard in specific locales. "Our new calculations will lead to a new generation of stress evolution models and help people understand how stress builds up and releases in seismic areas."
The other authors of the paper are Michael Gurnis and Hiroo Kanamori, both professors of geophysics at Caltech; and Egill Hauksson, senior research associate in geophysics at Caltech.