Image: courtesy of HOWARD ZEBKER

RADAR INTERFEROMETRY. This technique uses synthetic aperture radar mapping satellites to form detailed images of geological surfacesand to reveal centimeter-size changes in the earth's crust.

When Howard Zebker talks about volcanoes, he reminds his audience of the ones they might have drawn back in grade school: big cones on the flat surface of the earth, with chimneys in the middle to carry up magma from round puddles at the bottom called chambers. That abstract image is about right, he says, save for a few key differences: magma chambers arent necessarily round, and they actually change shape.

Zebker, an associate professor of geophysics and electrical engineering at Stanford University, and his colleagues know this because they have documented how volcanoes themselves change shapeslowly expanding at the surface and then collapsing back down in what seem like great, heaving sighs. Last December at the American Geophysical Unions annual fall meeting in San Francisco, Zebker and one of his graduate students presented their work on shape-shifting volcanoes in the Galpagos Island chain.

Using data from the ERS-1 and ERS-2 satellites, both operated by the European Space Agency, the researchers collected images of the topography of several volcanoes over a 10-year period. By comparing the changes in elevation of the entire surfaces of these volcanoes, they were able to measure how they rose and fell over that time. Sometimes the magnitude of change Zebker and his colleagues observed amounted to only several centimeters, but at other times, the researchers found up to several meters of deformation.

Though it's tempting to imagine, the volcanoes are not really breathing in any sense of the word. Instead, as hot rock enters and leaves a volcanos magma chamber, its walls shift. These changes, in turn, make the cold rock sitting on topthe actual cone of the volcanorise and fall. But Zebker and his co-workers took their observations a step further, and used theoretical equations that describe the pressure of melted rock inside the volcanoes' magma chambers to determine the depth and the shape of the actual magma chambers.

Instead of idealized spherical magma chambers, the scientists found different shapes, including "an oblong ellipsoidal kind of thing, that looks a little bit like an egg that's been squished down in the middle," Zebker said in a recent interview, surrounded by humming computers in his office. "What we're measuring, really, is the change in volume of the structure." The changing topography of the volcanoes further describes how their inner magma chambers expand and contract.

Tracking these kinds of changes in a volcano's magma chamber could lead to better predictions of volcanic activity, the scientists say. According to the U.S. Geological Survey, there are about 550 historically active volcanoes on the earth, of which about 60 erupt each year. Volcanic activity can spawn earthquakes or result in the release of steam or other gases before an eruption even occurs.

To create better predictions, however, scientists must first obtain more detailed and more frequent measurements of volcano metamorphoses. The ERS satellites used during the past decade traveled in orbits that brought them back to the same spot in the sky every 501 orbits. "What we acquired was a series of images, in principle, every 35 days over a seven- or eight-year period," Zebker explains. In fact, the radar equipment wasn't always turned on at the time the satellites flew over the Galpagosfor reasons of electrical load or priority given to other research teams using the satellites.

Ten-day observations would be more useful for making predictions, says Peter Mouginis-Mark, chief scientist at the Maui Pacific Disaster Center, who also reviews Zebker's work. "The more frequent the observations, obviously the better time resolution one obtains," he says. Some changeseven surprisingly large shifts like the lava domes that preceded the 1983 Mount Saint Helens eruptionhappen so quickly that the satellites would have missed them.

Also, radar interferometry, the observational tool currently used, cannot "see" through vegetation, which may cover the side of a volcano. So for now, ground movement measurements are lost in those ecosystems. In order to improve their observations, Zebker and Paul Segal, also of Stanford University, proposed launching something like an array of dedicated satellites, with specific observational capabilities that could document movement more often and at smaller scalesand even monitor only certain types of materials.

Researchers still do not fully understand the reasons behind the changing flow of magma in and out of magma chambers, causing volcanoes to go up and down, or to erupt. "We're not very good at predicting what it is yet," Zebker says, "and that's the focus of the research: to better characterize the processes so we can try and identify whats happening at any given time."