|A kilometer-high mountain dominates Lantau Island. Photo courtesy John Marwitz.|
The Hong Kong airport will have another distinction: one of the world's best warning systems for aviation weather hazards. High-tech instruments will look for signs of wind shear and turbulence lurking just off the airport's two runways. Sophisticated software will analyze the data and notify air traffic controllers of serious threats so that pilots can avoid trouble.
What?The Hong Kong project
Who?NCAR (Research Applications Program, Mesoscale and Microscale Meteorology Division, and Atmospheric Technology Division), as a subcontractor to Weather Information Technologies, Inc., in partnership with the Hong Kong University of Science and Technology, the University of Wyoming, and Coherent Technologies, Inc.
Why?To analyze threats to aviation from turbulence and wind at the new site and to develop a warning system
How?A comprehensive program with climatological analysis, a field study, computer modeling, software development, instrument placement, and real-world testing
Where?The site of the new Hong Kong International Airport, opening in 1998
When?1993 through 1997
Hong Kong turned to NCAR for the answers. From 1993 through 1997, a team of several dozen scientists, programmers, and engineers from NCAR and several other organizations designed a weather-warning strategy for the new airport. NCAR acted as the prime subcontractor for Weather Information Technologies, Inc., a for-profit company formed by the UCAR Foundation that brings NCAR's knowledge and technology into the private sector.
NCAR's breadth was tested by the Hong Kong project. In the 1970s and 1980s, its researchers had studied the problem of mountain-generated turbulence using fluid mechanics theory and computer models. Meanwhile, its Research Applications Program (RAP)--a blend of scientists, engineers, and programmers--had investigated microbursts and collaborated on a prototype radar- and anemometer-based warning system for U.S. airports. The Hong Kong government asked for more: not just a prototype but a working, ready-to-roll system to detect thunderstorm- and terrain-induced wind shear and turbulence, then warn pilots, air traffic managers, and airlines.
At least one problem had a ready-made solution. Hong Kong ordered one of the Doppler weather radars manufactured by Raytheon specifically to protect U.S. airports from microburst-related wind shear. Turbulence was a tougher issue, though. It often occurs in clear air, and it plays out on time and space scales smaller than those of microbursts and other weather hazards. No full-fledged airport warning system for turbulence had been developed anywhere.
The first order of business was to map out the large-scale weather and climate of Hong Kong with a meteorological survey that included five months of data collection in and near Lantau Island. The King Air, a veteran of the National Science Foundation/NCAR aircraft fleet, spent weeks on bumpy circuits sampling airflow around Lantau. With scientific and technical assistance from the University of Wyoming, the King Air helped find some of the strongest turbulence documented within a kilometer of ground level.
While turbulence was being measured by the King Air and by lidar (laser-based research radar, provided by the Colorado-based firm Coherent Technologies), other instruments collected wind data from atop Lantau Island and other points. The idea was to assemble a data set that would connect surrounding weather patterns to turbulence--a set of clues to help diagnose turbulence once the airport was up and running and the research aircraft and lidar had gone home.
Before these data were collected, the NCAR team had worried about a particularly nasty kind of turbulence. If the atmosphere is arranged in a certain temperature and wind structure, the air flowing over a mountain can shoot upward and amplify, then crash back downward with erratically high winds that are difficult to predict. At other times, the flow is torn into smaller, weaker eddies as it runs up, around, or over an obstacle. Although it can be strong, the second kind of turbulence (labeled mechanical) isn't as troublesome to forecast as is the first. NCAR's meteorological report hinted that mechanical turbulence would be the main concern, but a more detailed look was needed.
Here is where NCAR's computer-modeling expertise came into play. "People had studied flow over rough terrain for years, but we didn't yet have the resolution we needed to study it fully," says Terry Clark. For over a decade, Clark and colleagues in NCAR's Mesoscale and Microscale Meteorology Division have refined a model that uses concentric nests, like the lenses in a telephoto camera, to zoom in on and simulate weather processes that play out over the space of a kilometer or less. Progressively faster computers have enabled the resolution in Clark's model to get sharper.
Clark was able to pinpoint zones of turbulence that sheared off downwind of Lantau Island and cascaded within several kilometers of the approach and takeoff regions for aircraft. In his model, virtually all of the turbulence near Chek Lap Kok appeared to be mechanical, which agreed with the observations collected and analyzed. "We breathed a sigh of relief when we saw this," says Peter Neilley, who led the meteorological analysis.
The next step--and one of the biggest--was to determine how to convert the wealth of incoming weather data into actual warnings of wind shear and turbulence. For this task, Larry Cornman made use of an increasingly popular technique called multivalued logic. To explain it, he uses the analogy of someone sorting balls into "black" and "white" categories. What happens when a grey ball comes along? "You can't deal with this in typical logic. You have to throw it out as bad data, when it may be good data and you just didn't include enough categories."
The grey balls in Hong Kong were weather situations that fell between the cracks of the definitions for complex phenomena. For example, a given storm system might have a weaker-than-usual radar indication of turbulence. If there are several other indices that also are weakly favorable, then turbulence could be expected, even though no single index exceeds a black-and-white threshhold.
Another part of the Hong Kong solution is the fifth-generation version of the Pennsylvania State University/NCAR mesoscale computer model, a long-time research tool now being adapted for operational forecasting in several countries. The model will issue high-resolution forecasts every six hours for wind, temperature, and moisture at various heights in the Hong Kong airport vicinity, providing valuable input for turbulence outlooks.
With engineers and programmers working full bore, NCAR delivered two meteorological reports, two feasibility studies, and a complete system design to the Hong Kong government by mid-1996. Testing followed, along with collaboration with the Hong Kong Royal Observatory. NCAR and UCAR are now looking at other Pacific Rim airports as possible avenues in which to apply their expertise.
In the meantime, the technology-transfer loop will be circulating in the opposite direction. Scientific results and theories derived from the Hong Kong project are already feeding back into new research at the Hong Kong University of Science and Technology, U.S. universities, and NCAR. Clark expects to use his model to study mechanical turbulence further in other situations, such as near U.S. airports. "This has been one of the pioneering modeling studies of turbulence at this scale. We got interesting data on topics that are really uncharted."
William Mahoney, who followed Richard Wagoner as technical manager of the project in 1994, sums up the endeavor: "It's been creative and stressful and everything else."
Taking high-tech meteorological training abroad
Putting winter weather on ice at O'Hare
UOP's Cooperative Program for Operational Meteorology, Education and
Training (COMET®) was established to provide professional
development--training in the latest techniques, tools, and
concepts--for U.S. weather forecasters, hydrologists, and other
atmospheric scientists. For instance, a highly successful series of
interactive modules on CD-ROM and laser disc allows forecasters to
master a single topic (thunderstorm initiation, ocean wave analysis,
and the like) from their home sites.|
From a trailer near Marshall, Colorado, meteorologists from NCAR's
Research Applications Program kept tabs on a variety of snow-measuring
equipment. The project was funded by the Federal Aviation
Administration in conjunction with field tests in Illinois of a
software package to warn airlines of flight-delaying snowfall. United
Airlines, American Airlines, and the city of Chicago took part in a
demonstration of RAP's Ground Deicing/Snowfall Display and Prediction
System at O'Hare International Airport during the winter of 1995-96.
The test followed three years of system refinement at Denver-area
airports. Weather radar, although it has limitations in detecting
snowfall intensity, is proving useful to airlines that have to make
time-critical decisions in winter weather conditions. The system tested
at O'Hare consists of a network of snow gauges and weather stations
around the city and data from a radar located southwest of the airport.
Many thousands of dollars in unnecessary flight delays may have been
Winter 1996-97 Deicing Field Projects|