UCAR > Communications > Staff Notes > February 1997 Search

Where There's Ice, There's RAP

Winter is the busy season for scientists in the Research Applications Program who deal with ice-related aviation hazards. Here are updates on two projects now under way in RAP. Both projects are sponsored by NSF through an interagency agreement in response to requirements and funding by the Federal Aviation Administration (FAA) Aviation Weather Research Program.

(Photo courtesy United Airlines.)

New help for deicing decisions: Airlines test FAA/NCAR system

Two of the nation's busiest and snowiest airports, New York's LaGuardia and Chicago's O'Hare, are test sites this winter for a new data-gathering and display system to aid airlines in making deicing decisions. Since early January, the system has been providing snowfall "nowcasts" up to 30 minutes in advance for participating airlines to help reduce takeoff delays, increase safety, and save money on deicing procedures.

Ice buildup on aircraft waiting to depart can be a serious safety hazard. As little as 0.8 millimeter of ice on the upper wing surface increases drag and reduces airplane lift by 25%. With deicing fluids ranging from $2 to $4 a gallon, battling ice buildup can cost airlines tens of thousands of dollars on a single snowy day, in addition to the expense of flight cancellations and delays. The new system's half-hour forecasts could mean big savings for airlines through more effective deicing practices and fewer cancellations.

The FAA will evaluate the Weather Support to Deicing Decision Making (WSDDM) at both airports through user surveys and cost/benefit analyses. If successful, it will become a standard feature for those airlines willing to pay for its operation at airports regularly besieged by winter weather.

At La Guardia, Delta and USAir are participating from early January through March. American and United, who helped test a prototype of the system last year at O'Hare with encouraging results, are participating there again this winter, from mid-January through April.

Chuck Wade and Roy Rasmussen

Roy Rasmussen, head of NCAR's deicing program, comments, "Passengers can get anxious about safety and unpredictable delays in bad weather. I hope that providing the most up-to-date snowfall information will result in safer winter flying and greater confidence for the public." According to Roy, four or five major storms at each site this year would be enough to demonstrate the system's usefulness.

During the demonstration, surface weather stations, snow-weighing gauges, and Doppler radars are measuring snowfall accumulation, temperature, humidity, wind speed and direction, and the water content of snow. The data are being processed instantly and displayed graphically on video monitors at both the Delta control tower and the Marine Air Terminal at La Guardia, the New York Traffic Control (TRACON) office in Westbury, USAir's operations center in Pittsburgh, and the Delta operations center in Atlanta. At O'Hare, monitors are being installed at United's station control, tower, and flight operations center.

The monitors show bands of snow detected by the National Weather Service (NWS) WSR-88D radar network as they move toward or away from the airport. Data from special snow-weighing gauges strategically placed in and near the airports are displayed as a simple graph showing the water content of snow at various locations--a key factor in deicing decisions. The resulting short-term nowcasts (0-30 minutes) based on these and other meteorological data are expected to aid airport officials, including ground personnel deicing the planes, airline station control managers coordinating flights, airport managers in charge of plowing the runways, and air traffic controllers deciding how long to hold planes at gates.

The new technology is a direct result of scientific research. Roy found that the potential of snow to form ice on an airplane's wings and fuselage corresponds to the amount of water in the snow rather than to reduced visibility during snowfall. Visibility is often used by the National Weather Service to estimate whether snowfall is light, moderate, or heavy, and it has been adapted by the aviation industry as a factor in deicing and takeoff decisions. In studying a number of takeoff crashes due to icing (see list below), Roy found that visibility at the time of the accidents varied widely. He determined that large, dry snowflakes were less of a threat than small, heavy flakes holding more water, though the former reduce visibility to a greater degree. The snow-weighing gauges used in this winter's test at O'Hare and LaGuardia will measure the actual liquid content of the snow.

"Pilots have already become more aware that visibility can be misleading when it comes to aircraft icing," says Rasmussen. "Now we can give them quantitative measurements indicating the real potential of snow to form ice on aircraft."

NCAR has placed two snow-weighing gauges at La Guardia, two at John F. Kennedy Airport, and one at Newark Airport nearby in New Jersey. In the Chicago area, gauges will be placed at O'Hare, at Midway Airport, in the city of Wilmette, and at the College of Du Page (southwest of O'Hare). •Anatta

See last March's Staff Notes Monthly for a report on RAP's snow-weighing research at the Marshall site.

Takeoff accidents in which wing-surface ice contaminations were a contributing factor

Month/YearAircraft TypeLocationIcing Conditions
12/68DC-9Sioux City, Iowa*Light, freezing drizzle; fog
11/78DC-9Newark, N.J.*Snow, fog
2/79Nord 262Clarksburg, Wyo.Frozen snow
2/80Bristol 253Boston, Mass.*Light snow, fog
1/82B0737Washington, D.C.*Snow
2/85DC-9Philadelphia, Pa.*Light, freezing drizzle, ice pellets
12/85DC-8Gander, NewfoundlandLight, freezing drizzle
11/87DC-9Denver, Colo.*Snow
3/89F-28Dryden, OntarioSnow
3/92F-28New York, N.Y.*Snow
*Data from snow-weighing gauges available

Forecasters guide pilots into and out of the ice

As investigators seek clues to the 9 January crash of Comair Flight 3272 near Detroit Metropolitan Airport, atmospheric researchers are launching an experiment to study one possible culprit in the accident: large-droplet icing. RAP is testing innovative forecasting methods to help pilots steer clear of dangerous in-flight icing conditions. The NCAR team is collaborating with colleagues at the NASA-Lewis Research Center in Cleveland, Ohio, and the NWS Aviation Weather Center (AWC) in Kansas City, Missouri.

Frank McDonough and Marcia Politovich

From 13 January through 20 March, NASA-Lewis is flying a Twin Otter turboprop airplane equipped with measurement probes straight into treacherous icing zones identified by the RAP team. Scientists Ben Bernstein, Frank McDonough, and Marcia Politovich want to know whether their methods for forecasting these danger zones match up with the actual conditions pilots encounter in flight. NASA wants flight information to compare real-world ice shapes with those occurring in their experimental wind tunnel in Cleveland. "What's new and unique about this experiment is that it's specifically designed for large-droplet icing," says Marcia.

How does large-droplet icing bring down a plane?

The crash in 1994 of American Eagle Flight 4184 near Roselawn, Indiana, drew attention to the problem of large-droplet in-flight icing, especially for smaller aircraft. Large aircraft usually climb above dangerous conditions quickly. Smaller planes, which typically fly at lower altitudes, are at greater risk. Researchers studying the atmospheric conditions at Roselawn determined that relatively warm temperatures at cloud top (-5 degrees C) may have combined with certain wind conditions to produce large drizzle drops of supercooled water (water that is still liquid even though its temperature is below 0 degrees C). Supercooled droplets usually freeze when they hit the front of an airplane wing, and deicing devices on the wing are designed to remove them. However, some large droplets can flow back over the wing before freezing, or hit the wing aft of the deicing equipment, creating crusty ridges of ice that can disrupt airflow and destabilize some aircraft.

The experiment

The NCAR icing forecast team uses weather observations and output from numerical weather forecast models, combined with radar and satellite data, to target large-droplet icing conditions. The Twin Otter then heads into that area, taking precise measurements of the atmospheric conditions and the plane's response, as well as photographs and video images of any ice that forms. Since even experienced research pilots need direction out of dangerous situations, the NCAR forecasts include escape routes.

Ben Bernstein

Ben and Frank are also launching weather balloons in the Cleveland area from NCAR's instrumented weather van (see sidebar). The balloons carry sensing packages designed and built by Atek, Inc., of Boulder. The Atek sensors convey cloud structure data by measuring supercooled liquid as the balloons ascend through clouds.

Last year, at the request of the FAA, Ben created an algorithm, or mathematical problem-solving procedure, to automate freezing-drizzle advisories for areas smaller than those covered in general icing advisories. Due in large part to cooperative research by the AWC and NCAR, the areas of the advisories have become more specific over the past few years.

The research flights provide a good test of the new algorithm, which Ben calls the "stovepipe" because it uses data from observations on the ground to characterize what's happening in a stovepipe-shaped column of air at higher elevations. The experiment is also testing a new algorithm, developed by Jotharim Vivekanandan, that detects supercooled water droplets in cloud tops by examining infrared and visible-light readings from the NOAA Geostationary Operational Environmental Satellite-8 satellite.

NASA Lewis scientists will use flight data to improve icing simulation tools such as their icing research tunnel and their computer code for ice accretion. Working with AWC forecasters, the NCAR team will use the experiment's results to streamline pilot weather advisories issued at Kansas City.

Another step will be instructional materials for operational forecasters and pilots. Training modules, which include interactive CD-ROMs and Internet access, are being developed by the Cooperative Program for Operational Meteorology, Education and Training. They should reach forecasters and pilots by mid-1998. •Zhenya Gallon

An interesting day

Ben Bernstein and Frank McDonough were on their way to Cleveland in RAP's mobile cross-chain Loran atmospheric sounding system (CLASS) van on 9 January, the day of the Comair crash in nearby suburban Detroit. As Ben tells it, "Frank and I were driving on the Ohio Turnpike. When we got near Toledo, we were getting moderate freezing drizzle, which is a good indication of icing conditions aloft. We were having defroster problems and the engine was giving us some trouble. We pulled into a truck stop to fix the problem, and that's when the van's cell phone rang." The call was from the cockpit of the NASA Twin Otter, which has a satellite phone system on board. The crew had taken off from Cleveland that morning on a test flight and were picking up good measurements on a run over northeast Ohio.

"They said they were getting large-droplet icing near Cleveland and were interested in doing a second flight that afternoon. They asked us where they should go, and we said, 'Toledo,'" recalls Ben. He and Frank had a good mental picture of what was going on from looking at data before they left Colorado and at hotels along the way. In addition, there was the freezing drizzle hitting their windshield.

Meanwhile, they figured out how to make the defroster work without overtaxing the engine, and started heading eastward on the two-hour drive to Cleveland. They wanted to be there when the Twin Otter took off on its afternoon run. The freezing drizzle turned to snow as they headed east, but Ben and Frank made it to Cleveland that afternoon.

The Twin Otter, however, was stuck on the ground. Every plane files a flight plan before takeoff that includes alternate airports for landing in bad weather. There were no alternate airports with acceptable conditions near Toledo in the hours before the Comair commuter plane flew through the same area on its way to Detroit Metropolitan Airport. "As much as they wanted to go, they couldn't," says Ben. If they had, "they would have sampled that case very close to the location and time of the crash." Documenting such an accident with high-quality, in-flight data would have been "rare, perhaps unprecedented," he adds.

As luck would have it, NOAA's GOES-8 satellite was down that day. Despite missed opportunities, the NCAR team has data from GOES-9 and from the morning flight that may prove useful to crash investigators at the National Transportation Safety Board. The investigation will take time, and in-flight icing may not be implicated. Nevertheless, as Marcia Politovich put it, for the in-flight icing experiment team, 9 January was "an interesting day." •Zhenya Gallon

In this issue...
Other issues of Staff Notes Monthly


Prepared by Jacque Marshall, jacque@ucar.edu