UCAR > Communications > Highlights > 1996 Search


What do you do with a satellite navigation signal that's made noisy by temperature and moisture?

Listen to the noise.

The get-together at Michael Exner's house in Boulder on a cool, clear April evening in 1996 was much like any other birthday party. Several dozen scientists and colleagues were on hand, trading stories and making toasts. But the guest of honor couldn't be present. It was flying 700 kilometers (400 miles) overhead, circling the globe every 100 minutes.

One year earlier, the MicroLab 1 satellite had taken off from Vandenberg Air Force Base in California within the nose cone of a Pegasus rocket. The modest payload aboard the satellite included a shoebox-sized instrument. Its mission was to intercept signals from an array of 24 Global Positioning System (GPS) satellites deployed 20,000 kilometers (12,000 miles) above the earth by the U.S. military. The goal: to measure how much these signals were being distorted by noise from the earth's atmosphere--and to use that information to provide urgently needed global data on temperature, pressure, and moisture.



GPS/MET: meteorological applications of the Global Positioning System network of defense satellites


Principal investigators: Randolph Ware, Michael Exner, and Christian Rocken, UNAVCO; Benjamin Herman, University of Arizona; William Kuo, NCAR Mesoscale and Microscale Meteorology Division; Thomas Meehan, Jet Propulsion Laboratory (JPL)

Mike Exner, GPS/MET program manager. Photo by Carlye Calvin.


To gain inexpensive vertical profiles of temperature and moisture across the globe with high spatial and temporal resolution


By intercepting GPS signals with a satellite-based receiver and inferring the deviations in each signal's straight-line path caused by temperature and moisture gradients


Research and development based at UOP's GPS/MET Program Office and UNAVCO, with participation and support from NSF, NCAR, National Oceanic and Atmospheric Administration, National Aeronautics and Space Administration, Federal Aviation Administration, University of Arizona, and JPL; launch carried out at Vandenberg Air Force Base, California


First data collected in April 1995, following several years of development; data processing and system refinement ongoing

Meteorological Applications of Global Positioning System

By the satellite's first birthday party, it had taken its first steps toward fulfilling the dream of its creators. Hundreds of meteorological profiles from near ground level to 60 kilometers (36 miles) had been sent to earth. The potential for global coverage was second to none. "I don't think there's any question," says Exner, "but that these data will have very high value for weather prediction and climate research."

It was a nationwide, interdisciplinary team that put together this experiment, dubbed GPS/MET for meteorological applications of GPS. Its nerve center is at UOP, headed by co-principal investigator Exner and spun off from the University NAVSTAR Consortium. (UNAVCO has been based at UOP since 1991.) Some of the know-how for GPS/MET came from UNAVCO's long-successful program, funded by the National Science Foundation, of sensing earthquakes and other geologic motions as small as a millimeter via GPS signals.

Like the Internet, GPS is a technology originally developed for U.S. defense goals that has spawned an unforeseen array of civilian applications. From precisely known points in space, the GPS satellites issue signals. If a receiving device catches four or more signals from these locations, it can calculate its own location in space or on earth. For example, some automobiles can now intercept GPS signals and overlay the car's position onto digitized road maps so that drivers can see exactly where they are.

No matter how well the GPS machinery is engineered, the signals still have to pass through the ever-changing atmosphere. That poses a problem for navigation, but an opportunity for atmospheric science. As GPS signals pass through the atmosphere, they are affected by changes in density, a function of pressure, temperature, and moisture. The biggest changes occur near the ground, where moisture is most prevalent. Water vapor has an especially pronounced effect in bending the GPS signal path, which causes a measurable delay. The amount of water vapor above a given point on earth can vary more than tenfold, depending on local weather.

Meteorologists soon began to recognize the gold mine of data inherent in this "wet delay." By comparing the arrival times of signals sent to a ground-based receiver over a range of atmospheric conditions, scientists could infer the amount of water vapor present in a given column of air.

Even greater potential lay in the notion of using receivers in space. Researchers at the Jet Propulsion Laboratory (JPL) and Stanford University had studied the atmospheres of Mars, Venus, and Jupiter since the 1970s through radio occultation. In this technique, planetary-probe transmitters on a planet's far side sent signals to earth that brushed by the planet being studied so closely that they were bent and slowed as they passed in and out of the planet's atmosphere. The signal delay produced by the occultation was used to extract data on pressure and temperature.

Could the same approach work for our own atmosphere? In the early 1990s, Exner, UNAVCO director Randolph Ware, and a number of colleagues put together a plan to answer that question. If a receiver could be deployed in space, they figured, it could intercept a GPS signal that was occulted by the earth's atmosphere. With GPS transmitters stationed in space around the globe, occultations--and the atmospheric data they hold--could be gathered worldwide. The general concept had been suggested by JPL scientists as early as 1988 but had never been tested.

After its 1995 launch and a few days in orbit, the GPS/MET receiver came through. In its first full day of operation, the system successfully profiled the atmosphere above nearly 100 points from Greenland to Australia. The broad coverage brought cheer to the GPS/MET team, since it verified a key strength of the system. Radiosondes (instrument packages borne by balloon) have profiled the earth's atmosphere daily for more than 50 years, with hundreds launched across the globe every 12 hours. But few radiosondes can be deployed above the deep ocean or polar regions. "The value of GPS/MET," says Exner, "is that it gives you much of the information available from radiosondes, but with global coverage, and at a much lower cost per sounding." Moreover, weather balloons typically burst at heights of 25-30 kilometers (15-18 miles), while GPS/MET has obtained data from twice that high.

Other observing systems, such as rockets and spaceborne microwave sensors, each have their own limitations as well, making GPS/MET a useful complement to the existing mélange. Exner points out, "With a well-designed mix of observing systems, we could get more information for less money."

To assess their accuracy, the preliminary GPS/MET data were compared to readings from standard global analyses, rocket launches, radiosondes, and a microwave limb sounder from the Upper Atmosphere Research Satellite. Compared with the best available temperature analyses between 5 and 40 kilometers (3-25 miles), GPS/MET data differed less than 1degreesC on average. Below 5 km, it became more difficult for GPS/MET to separate the effects of temperature and moisture on the signal.

However, if more GPS/MET receivers are launched, Exner foresees an improved ability to track and decipher signals closer to the earth's surface. Even now, computer models of the atmosphere are able to benefit from the data. Xiaolei Zou uses NCAR's primary mesoscale model to study midlatitude low-pressure systems and other phenomena. When she inserted the information from simulated GPS/MET signals during early tests, she found that it improved the model's overall picture of the atmosphere. Tests using actual data are now under way. The upper-level data from GPS/MET could be a boon in the model's handling of wind perturbations at high altitude that can work their way down and trigger the formation of storm systems days later.

Scientists across the world have been using GPS/MET data since the first promising signals arrived. There are now more than 200 investigators examining the data at universities and research centers in over a dozen nations. Their interests range from short-term weather forecasting to long-term climate monitoring.

High-altitude researchers may get a special boost. GPS/MET can gauge the number of electrons at different heights within the sparse atmosphere above 100 kilometers (60 miles). These measurements could prove vital as background data when, as Exner puts it, "the sun has a temper tantrum." Solar storms inject large numbers of high-energy particles into the upper atmosphere, jeopardizing power grids on earth and communications satellites in space. The GPS/MET data on electron density could provide baseline data for other satellite-based warning systems and help identify the best frequencies for various radio communications uses. By the turn of the century, billions of bytes of data could be telling us where our atmosphere is headed--a valuable spin-off from the signals that now tell us where we are.

Weather data via the Internet

Radar processing goes miniature

For over a decade, UCAR's Unidata program has channeled, via satellite, a range of weather data--forecasts, computer-model output, satellite pictures, and more--to universities across the country. Now this service has migrated to the Internet and become part of a fruitful experiment in decentralized information transfer. The Internet Data Distribution (IDD) system allows a university to "subscribe" to specific data streams: data are delivered to subscribers as soon as they are available, without the mediation of a data center. Unidata created and maintains the software that allows this and organizes the routing of much of the data to a large number of universities.

The system depends on the active participation of the universities themselves, many of which have assumed responsibility for relaying data to others. However, any site can use the Unidata software to create its own distribution system on the Internet, delivering data of interest immediately to its own subscribers. The IDD is an entirely new approach to providing data, one that is compatible with the Internet's likely evolution toward offering point-to-multipoint transmission as a basic service.

Unidata Internet Data Distribution (IDD)

"We took a big step in technology in a short time with very limited support and very great success." That's how NCAR's Mitchell Randall characterizes the project he spearheaded, PIRAQ (PC Integrated Radar Acquisition). He and colleague Eric Loew developed a book-sized card that turns a personal computer into a complete radar receiver/data acquisition system. The inexpensive PIRAQ card opens the door to vastly greater flexibility in radar deployment by compressing the space needed for signal processing from bookshelf-sized units. The programmability allows each user to tailor the configuration as needed. Since it is small and inexpensive, it is ideal for many new research applications. For instance, the PIRAQ card allows the data from weather-avoidance radars in aircraft nose cones to be used for scientific research. Without altering the cockpit weather-avoidance displays, the PIRAQ system samples and processes the radar echoes and provides a high-resolution Doppler weather display. A PIRAQ-based, mobile Doppler radar system was used with success in the Verification of the Origins of Rotation in Tornadoes experiment in Oklahoma in the spring of 1995.

Elsewhere in this publication...


For more information, contact Milli Butterworth, butterwo@ucar.edu.
Prepared for the web by Jacque Marshall
Last revised: Mon Apr 10 13:23:27 MDT 2000