CONTACT: Janet Basu, News Service (650) 723-7582;
AAAS '98 BACKGROUNDER: The following article appeared in the January 15, 1997, issue of Stanford Report. It describes the development of synthetic aperture radar (SAR) interferometry. Stanford Associate Professor Howard Zebker, a co-developer of that technology, will speak at the AAAS annual meeting in Philadelphia on Tuesday, Feb. 17.
Space-borne radar to "revolutionize" views of Earth's hazards
Once or twice a week geophysicist Paul Segall finds himself popping out of his chair to stride down the hall to Howard Zebker's office in the Mitchell Building. Segall says he's usually only halfway through the door before he blurts out the question: "Howard, if we had the data, could we do this?"
A lot of other scientists are asking the same question. Zebker, associate professor of geophysics and electrical engineering, is one of the inventors of a new technology that uses radar pictures taken from 500 miles or more up in space, to track tiny movements on the surface of Earth on a large scale the strain building up on a California earthquake fault, for example, or the flow of a glacier in Chile. The technology, called synthetic aperture radar (SAR) interferometry, can show the gentle pulsing of a volcano's magma dome, a quarter of an inch up, a quarter of an inch down, as molten material surges up for what might become an eruption.
Images like this have put visions in the heads of scientists like Segall, associate professor of geophysics, who already uses sophisticated technology to study the changing movements of earthquake faults and volcanoes. Zebker's radar technique will mean that they and other researchers can track larger and more complicated movements on Earth's surface, in subtle detail.
Segall says, "This will revolutionize the way we look at the Earth."
Zebker is more modest. He says, "This doesn't replace measurements that scientists take on the ground. Radar only shows the surface, and a lot of the interesting things about earthquakes and volcanoes require detecting what goes on underground. But this technique does allow us to measure some things we haven't been able to measure before."
Zebker received his doctorate from Stanford in 1984 and returned in 1995 as associate professor with a dual appointment. He had spent the intervening years at the NASA-Caltech Jet Propulsion Lab (JPL) in Pasadena, where a team of scientists was adapting a radar technique once used to map the surface of Venus. Instead of mapping the other planets, they pointed their synthetic aperture radar antennas toward Earth itself and used the data to make detailed, three-dimensional topographic maps.
Zebker was one of the first to show that SAR interferometry could also be used to find out how pieces of that topography move around.
Satellite-based motion detector
Now he and other scientists have a vision of a satellite continually circling the planet, using cloud- and darkness-piercing radar to snap pictures of regions where glaciers, earthquakes and volcanoes are active. Each snapshot is assembled by computer from scores of bounced-back radar waves as the satellite sweeps over an entire region: The effect is like using a panoramic camera to shoot pictures thousands of miles long and 60 miles wide.
The envisioned SAR satellite would cover Earth, returning every two weeks or so to snap the same region from a slightly different angle. Comparing pictures taken on one pass of the satellite with those taken on subsequent passes, vulcanologists will be able to detect changes as small as one-quarter inch in the remotest volcano possible warnings that it is rumbling to life. Earthquake analysts will see movement along the whole expanse of a fault zone. Antarctic ice specialists will be able to monitor the fluctuating volume of ice in the entire continental ice sheet essential data to track melting that could raise sea levels and change global climate patterns.
Right now, the United States has no civilian satellite with instruments dedicated to SAR interferometry. At least two versions of such a satellite have been proposed to NASA, and Zebker says he is confident that an SAR orbiter eventually will be launched. In the meantime, Japan, Canada and the European Space Agency have recently launched satellites with radar capabilities. The data that they collect are available to American scientists only on a limited basis; Zebker is one of the few with access to some of it, via the European group's ERS-1 satellite and Canada's Radarsat.
Zebker says one of his goals is to design software so that the data from a U.S.-launched SAR orbiter can easily be made available to scientists all over the world.
Zebker has earned several patents and a tall stack of NASA achievement awards for his contributions to remote radar sensing technology. Some of that work, with scientists Von R. Eshleman, Len Tyler and Richard Simpson of Stanford's Space, Telecommunications and Radioscience Laboratory, focused on the rings of Saturn. But primarily he worked with NASA radar imaging teams, in particular with JPL's Richard Goldstein. They developed ways to use radar to learn the speed of ocean currents and to make topographic maps.
Measurements never before available
In the late 1980s, Zebker published several scientific papers on data-processing techniques that combined both of those capabilities. Instead of measuring the large and relatively fast-moving changes of an ocean current, he turned his attention toward large areas on land that move a little bit at a time.
His first widely noticed demonstration of this technique was done with several radar images taken of the same plots of farmland in California's Imperial Valley. On a day before the farmers irrigated the land, the valley was uniformly flat. When some plots of land were irrigated, the surface of the soil was expanded by the water underneath, an almost imperceptible boost in altitude of less than half an inch. In a radar image, the irrigated plots stood out like square, fluffed pillows.
"Most people did not take much notice of that paper," Zebker recalls. But a few scientists, including Segall at Stanford, were intrigued. "I thought, if he can show a change that small in farmland, what can he do with something we're interested in?" Segall said.
Zebker says that glaciologists also were quick to pay attention. They study ice movements the way Segall studies earthquake faults and volcanoes: using instruments and markers placed at strategic points so they can track the ice inching forward. In both cases, the scientists want to see the way Earth's surface moves, not over the eons but over the days, weeks and years that matter to people worried about hazards.
Segall works with his students to deploy networks of Global Positioning System (GPS) instruments on the flanks of Kilauea volcano in Hawaii and along the San Andreas earthquake fault system in the Bay Area. Each instrument uses data from a different type of satellite in the GPS network to precisely pinpoint changes in its location. The GPS data allow the researchers to track changes in the shape of Kilauea as it squirts out lava along its flanks, and to record the movements of Loma Prieta mountain as it is jostled by the San Andreas and other faults nearby.
Tests of SAR interferometry quickly showed that it could detect the same kinds of changes, in places where an expert on ice or on lava dared not set foot. The radar also could take pictures in the foulest weather. While Segall and his students can see movements at the few spots they placed their instruments, with SAR they can track changes all across the surface. And ice specialists found that they could track events so large that they might not have detected them with ground-based instruments for example, the breakup of a giant ice floe in Antarctica in 1992.
"With previous methods we could only tell what was happening after we went back and analyzed our data," Segall says. "This is so visual that you can almost see it moving."
Zebker says that for scientists, the most exciting uses of this technology so far have been French and American studies of SAR pictures taken before and after the magnitude 7.3 Landers earthquake in Southern California in 1992. The study showed not only how the surface near the Landers fault changed due to the main earthquake's shaking, but also how strain relaxed along a network of related faults in the months after the quake. No other method had been able to show continuous changes in a web of faults on such a large scale.
"These are phenomena we would not have been able to record before," said Zebker.
In the future, he says, he can envision an SAR orbiter continuously collecting data of all the seismically active regions of Earth. In addition to the instruments they have deployed in spots known to pose a threat to humans, scientists would scan the SAR data for movement along unknown faults or remote and sleeping volcanoes. They would find out when and where to rush their instruments for closer study on the ground. And local populations would get early warning of potential dangers.
Zebker continues work to improve SAR interferometry. He is working on techniques to minimize distortions to radar data caused by water vapor in the air, and by distortions in Earth's ionosphere. He is looking for ways that ecologists might take advantage of SAR to calculate the biomass of vast forests, and to keep track of the losses as forests are cut and cleared.
As one of the scientists clamoring to add Zebker's technique to his own, Segall says that it may have many uses none of them has envisioned yet. "We'll be able to ask questions we haven't even thought of asking before," he says.
Last fall, when hot magma from a volcano melted the underside of a glacier in Iceland, 3 cubic kilometers of hot water surged in chambers underground for weeks before bursting out in a steaming, muddy flood. Icelandic officials predicted the flood but could not say when and where it would occur.
"I was teaching an Earth hazards class, and we wondered if we could map the water running around under the glacier [with this technique]," Segall says. It was another reason to go to Zebker and ask, "If we had the data . . ."
Zebker says SAR probably could have been used to help the Icelanders predict where the water would burst forth. But at least for this eruption, no radar pictures were available in time.
For additional information, see Howard Zebker's home page at http://ee.stanford.edu/~zebker/.