The accumulation of ice on aircraft prior to take off has long been recognized as one of the most significant safety hazards affecting the aviation industry today. As little as 0.08 mm of ice on a wing surface can increase drag and reduce airplane lift by 25%. Acutely aware of the impacts these icing hazards can have on aviation, the Federal Aviation Administration (FAA) began supporting ground de–icing research at the National Center for Atmospheric Research (NCAR)* in 1991. As a direct result of this FAA program, scientists at the Research Applications Program (the principal division of NCAR responsible for aviation weather projects) have developed a state–of–the art, integrated display system that depicts accurate, real–time nowcasts of snowfall rate, plus current temperature, humidity, wind speed and direction.
NCAR's Weather Support to De–Icing Decision Making System requires minimal training to operate, and no special meteorological knowledge to interpret. The display provides a maximum amount of weather information at a glance.
Designed specifically for airport decision–makers, graphic displays are strategically located at airline station control, dispatch, and deicing facilities, airline and city snow desks, and FAA air traffic manager positions. The snowfall and weather information are used by ground personnel conducting aircraft de–icing operations, airline station control managers coordinating flights, airport managers coordinating runway plowing activities, and air traffic controllers involved in gate–hold program planning. The information allows decision makers to anticipate both the onset and termination of snow at the airport and surrounding regions.
The principal sources of data for the system are regional area Doppler radars (National Weather Service WSR–88Ds and FAA Terminal Doppler Weather Radars), surface weather stations, and snow gauges situated within the terminal area which accurately measure the amount of water in the snow (i.e., the melted liquid–equivalent snowfall rate).
Research indicates that the icing hazard for aircraft directly corresponds to the amount of water in the snow, rather than visibility. It is the latter that has traditionally been used to determine de–icing and take off decisions. Results from field tests of de–icing fluids have identified the liquid–equivalent snowfall rate as the most important factor determining the holdover time (time until a fluid fails to protect against further ice build–up).
Aircraft ground icing danger is found to be greatest when the temperature is between 25 and 31 degrees Fahrenheit, and the wind speed is about 9 to 15 mph. After studying the weighing snowgauge data for five major take off crashes in the US attributable to inadequate de–icing/anti–icing prior to take off, NCAR scientists determined that all five accidents occurred when the liquid–equivalent precipitation rate reached the danger zone of a minimum of 0.08 inches/hour.
Analyses also indicated that visibility at the time of the accidents was suggestive of only light to moderate snow, indicating that the current National Weather Service snowfall rate categories, which are based on visibility alone, are not always adequate for effective ground de–icing operations. Since the de–icing fluid depends critically on liquid–equivalent snowfall rate, visibility alone can give pilots and ground crews a false sense of security.
NCAR's integrated display consists of one large plan view showing radar reflectivity of the storm in the main window. Geographic references, such as the terminal runways, taxiways, and airport concourses are overlaid on top of the gridded data. Yellow and green storm motion vectors are shown overlaid on the radar reflectivity of the storm, and indicate the expected 30 minute motion of the snowbands. Smaller windows on the right side of the same screen show current surface weather conditions from National Weather Service ASOS stations in the surrounding region in a simple, easy to read decoded form, WSDDM snowgauge and surface weather data updated every minute at the major airports, a time line graph of liquid equivalent snow intensity, temperature, wind speed and direction, and humidity for the last two hours updated every minute from specific locations within the terminal area. Also shown in the lower right is a 30 minute forecast of snowfall intensity (light, moderate, heavy) based on a nowcast of snowband motion and a real–time calibration between the snow intensity and radar return.
The time line in the lower right corner of the display indicates the liquid–equivalent snowfall accumulation, from the time since an airplane has been de–iced, to the time expected to elapse before take off. Use of the display during de–icing operations has been shown to reduce end–of–runway de–icing; a significant cost savings. The pink line indicates accumulation over the past 60 minutes. Current time is indicated by the vertical red line, and the snowfall is accumulated in a reverse sense, allowing users to easily estimate the amount of snow that fell from the current time to any chosen time in the past. The dashed pink line denotes predicted snowfall accumulation at user selected airports (LGA, EWR, or JFK in the example). The past one hour radar reflectivity history over the airport is shown as the solid yellow line, and the predicted thirty minute forecast as the dashed yellow line.
A prototype WSDDM system was successfully demonstrated at Denver International Airport (DIA) during the 1994–1995 winter season, at Chicago's O'Hare International Airport during the 1995–1996 winter season, and at Chicago O'Hare and New York LaGuardia during the 1996–97 winter season in cooperation with USAIR, Delta Airlines, Port Authority of New York and New Jersey, United Airlines, American Airlines, and FAA Air Traffic control managers. The WSDDM System was demonstrated during the 1997–1998 winter season at New York's La Guardia International Airport. The results from these demonstrations will be used by the FAA to determine how snowfall "nowcasts" (short–term forecasts) might be implemented at airports nationwide.
Features of The Integrated Display:
Aviation-based winter weather research at RAL, sponsored by the Federal Aviation Administration, has focused on developing two new systems in support of Ground Deicing operations. The Liquid Water Equivalent (LWE) system combines a Hotplate and GEONOR snow gauge; a Vaisala PWD–22 precipitation type sensor; a Campbell freezing rain sensor; a Vaisala WXT wind, temperature, and humidity sensor; and a Decagon Leaf Wetness Sensor to estimate a real–time liquid water equivalent precipitation rate. This rate is a critical component of the Checktime System, a UCAR patented technology for aircraft ground deicing operations, that determines when deicing/anti-icing fluids applied to aircraft are close to failure based on temperature measurements and precipitation rates that are updated every minute from the LWE system. Checktime is aircraft independent and only requires the end user to know the time that the aircraft was deiced. Demonstrations of the LWE and Checktime systems are conducted at Denver International Airpot, Chicago's O'Hare International Airport, and the Clevel-Hopkins International Airport. LWE's ability to detect ice pellets has been tested at St. John's, Newfoundland, Canada, the site of the climatological maximum of freezing precipitation and ice pellets in North America.
When an aircraft experiences winter precipitation conditions on the ground, it is required to undergo a de/anti-icing procedure involving glycol-based fluids before takeoff. In order to determine the fluid holdover time (the length of time the fluid will provide protection to the aircraft), the pilot needs to know the time deicing began, the precipitation intensity and the ambient air temperature. This information is then used with a look up chart that will tell them how long they can expect their fluid to provide protection from snow or ice building up on the aircraft surfaces under the given conditions.
There are, however, several problems with this procedure. First, the pilot’s look up chart relies on visibility to determine the intensity of the falling snow. Rasmussen et al, 1999, showed how visibility can lead to incorrect estimates of snowfall intensity and demonstrated the need for intensities to be based on Liquid Water Equivalent (LWE) measurements of snowfall rate. Secondly, this method assumes that the snowfall rate and ambient air temperature will remain constant until takeoff. In reality, snowfall rates and air temperatures are highly variable, especially at airports prone to snow squalls and lake effect snow events. Lastly, the task of looking up a holdover time means pilots have an additional task to add to their pre-flight checklists, which can further add to weather-induced departure delays.
To address this issue, an automated holdover time determination algorithm was developed by NCAR known as the Checktime algorithm. Checktime addresses the above issues since it’s based on a system that uses a measured LWE precipitation rate instead of visibility to determine snowfall intensity. This LWE system was developed by NCAR, and updates every minute using the real-time data provided by sensors from a locally deployed LWE instrumentation site. Measurements taken by the LWE system include temperature, pressure, humidity, wind speed and direction, and precipitation type and rate. In addition to using the rate, wind speed, precipitation type and air temperature measurements from the LWE system, Checktime also incorporates the regression algorithms published by APS Aviation (APS) for each type of de/anti-icing fluid to give accurate estimates of holdover times for a given fluid.
Unlike holdover times, which are set in the future, Checktime provides a wall clock time in the past (typically referred to simply as the Checktime), which estimates the length of time a given fluid would fail by incorporating together current and past weather conditions. The Checktime algorithm begins with the current time and integrates the LWE rate backwards in time, minute by minute, until it determines sufficient precipitation has fallen for the fluid to exceed its protection capability. Precipitation rate, wind speed, temperature and precipitation type are all inputs into the Checktime algorithm. Fluid type and concentration can be selected by the end-user and, using the regression equations relating holdover times to precipitation rate and ambient temperature developed by APS, Checktime then produces a time in the past. A pilot only needs to know the time their plane was de/anti-iced and as long as that time remains more recent than the Checktime, they know their fluid is still providing protection. This gives Checktime the unique advantage that it is aircraft independent and the only information the pilot needs to know is the time their aircraft was de/anti-iced. Additionally, because Checktime is incorporating real-time snowfall rates, it provides a more accurate estimate of fluid failure because it does not assume a constant snowfall rate or ambient air temperature. Rasmussen et al, 2009 showed that in some cases, assuming a constant rate and temperature, holdover times can be almost twice as long as Checktime leading to conditions that may be unsafe for aircraft departures. To further ensure a margin of safety, Checktime was developed to incorporate a small conservative bias in the reported values such that when compared to actual fluid failures, Checktime should give shorter holdover times.
A common question asked by pilots is “If Checktime is based in the past, and holdover times are based in the future, how are these two related?” The following scenario can answer this question. A pilot is told s/he was anti-iced using Kilfrost ABC-S Plus fluid at 08:45, the snowfall rate is moderate and the air temperature is -8.0°C. The lookup chart tells the pilot s/he has a holdover time of 35 minutes. Using these same numbers and fluid type, Checktime gives a time of 08:10 (35 minutes into the past). If the snowfall rates and the ambient air temperature remain constant for the next 35 minutes, the pilot will hit their holdover time at the same time Checktime gives a value of 08:45. Thus, holdover time (35 minutes) and the difference of 35 minutes between the Checktime (now 8:45 since we’ve gone 35 minutes ahead in time) and the current time (now 9:20) would be the same.
To demonstrate the advantage of a Checktime-based system, the same scenario above can be used, however, instead of snowfall rate and temperature remaining constant, the snowfall rate now increases after fluid application and the temperature slowly decreases. Using the lookup chart, the pilot is still given a holdover time of 35 minutes at the time of appliation, but Checktime incorporates the increasing snowfall rate and decreasing temperature and comes up with a Checktime of 08:25, now only 20 minutes in the past. The pilot will now get to the Checktime time quicker than the holdover time from their chart since the fluid will now have failed quicker than the holdover time indicates. Checktime thus alerts the pilot to these changing conditions and automatically corrects the estimated holdover time to account for these changes.