Satellite catches earthquake in act.

Last spring, geologist Robert E. Crippen compiled a wish list describing his dream quake. Crippen, who works at NASA's Jet Propulsion Laboratory in Pasadena, Calif., was convinced that satellite pictures of Earth's surface could show an earthquake in motion, provided the jolt met certain criteria. The quake should be big - enough to cause at least five meters of motion across the fault; otherwise, the satellite imagers would miss the action. It should occur in an open area without much ground-obscuring vegetation or cloud cover. And it would be nice if the epicenter lay close to home, although preferably in a desert area where relatively few people would be hurt.

As if by design, the Landers earthquake that rocked the Mojave desert on June 28 met all of Crippen's criteria, enabling him to make the first film of fault motion from space. Like timelapse photography, Crippen's film consists of photographs taken at different times by a camera positioned in the same spot. In this case, the camera sits on the French SPOT satellite, which orbits 830 kilometers above the Earth's surface. Crippen obtained two images: one taken on July 27, 1991, and another taken when the satellite passed over the same spot on July 25, 1992, just a month after the Landers jolt. By shifting back and forth between these two images, he produced a video that shows movement across the fault.

During the quake, land northeast of the fault slid past land southwest of the fault, like two trains running in opposite directions. The two parcels of land slipped past each other by up to six meters, an amount actually smaller than the 10-meter resolution of the satellite sensors. But the earthquake moved the land surface enough to alter individual image pixels, making the motion discernible during the video.

The SPOT shots even show evidence of ground cracks along the fault. Individually, the cracks measure only a few centimeters across, so a single crack can't appear on the images. But so many fissures laced the ground along the fault that they collectively show up as a dark line in the satellite photo taken after the earthquake (right). The prequake image (left) does not show any fault cracks.