According to Bill Kuo, the director of the COSMIC Project at UCAR, the GPS radio occultation technique used by the COSMIC system will not suffer from any of these problems and therefore will deliver a precise measurement. That is because the COSMIC receivers will measure changes in the phase and amplitude of radio signals transmitted by GPS satellites, not changes in radiant energy.

The vertical resolution of the receivers will vary from about 100 meters in the lower part of the troposphere to about 1 kilometer in the upper troposphere and stratosphere. By contrast, traditional weather satellites in geostationary or polar orbits use infrared or microwave measurements and have a vertical resolution of several kilometers. Poor vertical resolution has been an important factor in the uncertainties regarding initial conditions that have plagued global weather prediction models, says Kuo.

Weather balloons carrying instrument packages (radiosondes) are another traditional method of gathering atmospheric data. Currently, there are more than 850 stations taking radiosonde data around the globe. But the instrument packages provide data of varying quality, and although the balloons have good vertical resolution, their horizontal resolution is poor. What's more, radiosonde observations are generally available only over land, leaving about 70 percent of Earth's atmosphere, the part above the oceans, uncovered.

In contrast, COSMIC will be able to provide 2500 soundings, distributed uniformly around the globe, per day. Once the atmospheric data have been recorded by the satellites, the information is downloaded to the COSMIC ground stations in Fairbanks, Alaska, and Kiruna, Sweden, then sent to a data analysis and archiving center in Boulder, where it is processed and transmitted to global weather prediction centers.

The COSMIC system is the first collaboration between the United States and the Taiwanese National Space Organization. It has a price tag of US $100 million for the first two years of operations, with 80 percent of the mission being funded by Taiwan's National Science Council, the country's equivalent of the U.S. National Science Foundation. NSF is the lead U.S. sponsor of COSMIC.

COSMIC is the Taiwanese National Space Organization's third satellite project; it is called FORMOSAT-3 in Taiwan. The country became interested in working on the project after the success of GPS/MET, a prototype GPS satellite launched by UCAR in 1995. GPS/MET gathered data for two years, proving radio occultation would provide a basis for more accurate weather and climate predictions.

One important reason for Taiwan's involvement is its concern about typhoons, the most serious weather threat facing the island. "COSMIC's soundings will tell us more about the environment of these tropical storms and contribute greatly to improved storm forecasts," says Bor-Han Wu, a senior scientist with Taiwan's space program.

According to Anthes, who initiated the collaboration with Taiwan in 1997, "Taiwan wanted to become an international partner in Earth system sciences." Evidently it also sees COSMIC as a way to get into the global business of producing Earth satellites.

Orbital Sciences Corp., of Dulles, Va., built the prototype COSMIC satellite and the Minotaur launch vehicle. The remaining five COSMIC microsatellites were then produced as kits and sent to Taiwan for assembly, with Taiwanese manufacturers contributing some of the parts. Both Orbital Sciences and UCAR have technology-transfer agreements approved by the U.S. Department of State to let them carry out the collaboration with Taiwan on the COSMIC project.