GPS steps forward with a small stumble.

GPS steps forward with a small stumble

Various groups of scientists report that with the new technology of the earth-space Global Positioning System (GPS), they are making giant strides in the field of measuring long distances with great accuracy. At the same time, one experiment seems to have uncovered a GPS pitfall that cannot yet be explained.

The GPS relies on a network of permanent and portable receivers on earth that monitor a series of microwave signals emitted by orbiting satellites. At present it can measure distances to within a few parts in 10.sup.8., which amounts to an error of a few centimeters in 1,000 kilometers, says William Melbourne of the Jet Propulsion Laboratory in Pasadena, Calif. Although other techniques, such as Very Long Baseline Interferometry, are more accurate, GPS receivers have the advantage of being relatively inexpensive, small and quite mobile.

Exploiting these features, Melbourne and a host of international colleagues recently finished the first stage of the largest GPS experiment to date -- an effort to measure the tectonic motion of crustal plates in South and Central America. The researchers placed temporary receivers on the mainland as well as on small islands in the Pacific -- a feat that would have been impossible with other techniques, says Melbourne. By comparing measurements taken in January with some planned for two years from now, the researchers hope to gauge how quickly the oceanic plates under the eastern Pacific are moving when they run into the continental plates and dive into the earth's interior.

The accuracy of GPS measurements has improved significantly since the initial tests five years ago, and the system has proved reliable in many experiments. But one recent GPS venture in California's Long Valley caldera yielded curiously inaccurate measurements, says Jet Propulsion Laboratory's Tim dixon. The caldera -- a volcanic crater -- is currently shifting as a pool of magma grows several kilometers beneath the surface. Researchers had hoped to use GPS to monitor the caldera movement with respect to a station 70 km away. But they got measurements 10 times less accurate than expected. For GPS, "This is the only fly in the ointment," Dixon says. He and others say it will be important to understand the problems at Long Valley before placing full confidence in the system.