When wings won’t fly

Posted: 30 November 2007 | Marcia K. Politovich, National Centre for Atmospheric Research, Boulder, CO | No comments yet

The wings of an aircraft are carefully designed to provide the lift needed to fly. The shape of the wing must be aerodynamically efficient and the surface should be smooth to allow air to flow effortlessly around it. However, prior to and during flight, atmospheric phenomena work to reshape and re-texture those wings. These phenomena create potentially hazardous icing conditions by which ice builds on the wings and degrades their performance. Aircraft icing researchers have applied their scientific and engineering expertise to develop new products, to allow the flying public to avoid the potentially hazardous consequences of icy wings.

The wings of an aircraft are carefully designed to provide the lift needed to fly. The shape of the wing must be aerodynamically efficient and the surface should be smooth to allow air to flow effortlessly around it. However, prior to and during flight, atmospheric phenomena work to reshape and re-texture those wings. These phenomena create potentially hazardous icing conditions by which ice builds on the wings and degrades their performance. Aircraft icing researchers have applied their scientific and engineering expertise to develop new products, to allow the flying public to avoid the potentially hazardous consequences of icy wings.

Inflight icing: cause

Supercooled liquid water is one of the wonders of the atmosphere – it is water that remains in its liquid state even when its temperature is below freezing. The physics are somewhat complicated, but think of it this way: in a drop, the water molecules are in constant movement, sliding all over one another, yet in a solid they are locked into a crystal lattice. To get from the liquid to the solid state they need an ‘example’ surface to show them how to form this lattice. This surface can be a small particle, referred to as an ‘ice nucleus’, or larger objects such as a twig on a mountaintop or an aircraft’s wing.

However, this wonder of the atmosphere is no friend to aviation. When an aircraft is exposed to supercooled liquid drops in clouds or precipitation, those drops adhere to its surface and freeze. While the whole aircraft can experience the phenomenon of aircraft icing, it is most critical on the wings’ surfaces since the wings provide most of the lift needed to become or stay airborne. Aerodynamicists go to great lengths to design wing shapes that are efficient in providing lift and reducing drag. The process of icing actually alters those shapes and adds unwanted texture to the surface, thus, icing is something that pilots generally want to avoid. In the US between 1982 and 2000, 17-49 accidents with icing as a cause or factor occurred annually, with a total of 819 fatalities.

The amount of ice, its location on the wing and its shape and texture, depend on the amount of liquid, the sizes of the drops and the outside air temperature. Other factors such as airspeed, wing shape, angle of attack and whether flaps are deployed, enter into the equation for inflight icing conditions. Thus the aviation community requires a variety of disciplines to understand, forecast, and prepare for this phenomenon. Meteorologists must know how to diagnose and forecast icing conditions from available weather information; engineers must design wing shapes (airfoils) and de-icing mechanisms to handle most icing conditions and flight crews must know how to respond appropriately to icing conditions.

Aircraft must undergo a certification process for flight into icing conditions. Even with advanced computer modelling tools and wind tunnels that simulate icing environments, candidate aircraft are still required to fly in natural icing conditions in real clouds. Test flights are conducted in a variety of situations within what are called the ‘icing envelopes’ – ranges of liquid water contents, drop sizes and temperatures that provide confidence that the aircraft can handle nearly all icing environments. ‘Nearly all’ is a key point here – Nature is always sure to provide a surprise. The October 1994 accident near Roselawn, Indiana, in which sixty-eight people lost their lives, is a good example. An ATR-72, flying in a holding pattern near the top of a cloud, suddenly banked sharply, recovered, banked again and entered a spin from which it did not recover. Detailed analyses of the weather conditions, flight data recorder information, combined with studies of ice accretion and resulting performance degradation, pointed to conditions outside the icing envelopes. The aircraft was indeed certified for flight in ‘nearly all’ icing conditions, but these conditions were likely to have included drops larger than those in the icing envelopes – supercooled large drops or SLD, which have diameters exceeding 50 micrometers. This serious accident initiated a series of international meetings and workshops, resulting in research to identify where SLD conditions are likely to exist, characterise their size ranges and liquid water contents and define new icing envelopes that will likely be implemented in the US in the next few years.

Avoiding icing

Many aircraft are not certified for flight into icing conditions and must avoid those areas. Even certified aircraft seek to avoid icing, as it demands more attention from the flight crew, activation of de-icing equipment and possible diversion due to heavy icing conditions. The Current Icing Product, CIP, is the product of research conducted at the National Centre for Atmospheric Research (NCAR) in Boulder, Colorado. CIP combines numerical weather model output with observations, using the underlying concept that no one model or observational platform provides everything the end-user needs to know about the potential icing hazard. Observations include satellite imagery, surface weather measurements, radar data and voice pilot reports of icing. The CIP algorithm incorporates a combination of decision-tree and fuzzy-logic methods to determine the weather scenario and interpret and combine information appropriate to that scenario. It has been evaluated year-round to ensure accuracy and reliability. CIP was recently approved by US government agencies for unrestricted operational use. This means that it has met certain standards not only for accuracy, but for unambiguous depiction of the icing hazard and suitability for pilots and dispatchers, as well as aviation meteorologists. It is supplemental to the icing AIRMETS (Airman’s Meteorological Bulletins) and provides greater detail than those official four-hour forecasts, with hourly outputs of 20-km horizontal and 1000-ft vertical resolution from the surface to FL030. The probability of encountering icing and its expected severity are depicted. Areas of potential SLD conditions are included.

CIP’s sister, FIP (Forecast Icing Product) is similar to CIP except that it relies on numerical weather model data to supply icing information. FIP provides 2 to 12-hour forecasts of icing probability, severity, and potential for SLD. It is currently in an experimental state, meaning that it should be used by those with meteorological training and strictly as a guidance product to supplement the AIRMETs.

The Aviation Digital Data Server (ADDS) is the primary presentation medium for viewing CIP and FIP output. CIP is displayed on operational ADDS (; FIP and Alaska versions of CIP and FIP are found on the experimental version ( This non-passworded website is a one-stop shop for many aviation weather products, not only icing, but turbulence, ceiling and visibility, and convective weather.

Safely on the ground

Aircraft can also encounter icing conditions when sitting, apparently safely, on the ground. When the temperature is below freezing, precipitation, snow, ice pellets, freezing rain and drizzle, can fall onto the wings and can adhere to them. This ice, like the ice accreted en route, can adversely affect the aircraft’s ability to fly. As little as 1/64th inch (or 0.40 mm) of ice can result in a severe reduction in lift. Takeoff is a critical time in flight, during which any degradation in lift can have serious consequences, thus, taking off with clean wings is critical. Aircraft are de-iced at the gate or a common de-icing location using such fluids as warmed, diluted ethylene glycol. Since there is usually some time, especially at large, busy airports, between removal of the ice at the gate or de-icing pad and takeoff, some means of preventing further ice accretion must be available to ensure safety. Extensive testing of anti-icing fluids for this type of protection is an ongoing research activity. Holdover timetables describe how long a plane can be ‘held over’ between fluid application and takeoff and relate precipitation rate and temperature to safe protection times for the different fluid types.

Determination of the precipitation rate in terms important to de-icing of aircraft is not straightforward. The National Weather Service estimates precipitation rates by visibility. For example: visibility of 5/16 to 5/8 miles is interpreted as moderate rain or snow. The holdover time is dependent however, on how much actual water contacts the wing and subsequently dilutes the anti-icing fluid. Since rain, snow, drizzle, ice pellets and other precipitation have a vast range of sizes and densities, there is not a good relation between the visibility-estimated precipitation rate and the rate that water is falling out of the sky – the liquid-equivalent precipitation rate.

A little WSDDM goes a long way

How to resolve this dilemma? The Weather Support for De-icing Decision Making (WSDDM) system was developed at NCAR to combine weather information with fluid holdover times. Gauges that measure the true precipitation rate, the liquid-equivalent precipitation rate, are placed around the airport grounds and their measurements are matched to radar data above them. The radar data, from the operational NEXRAD network, can thus be calibrated to provide a continuous map of precipitation rate. Movement of radar echoes can be tracked and extrapolated into the future, therefore they are able to provide air crews, guidance on the amount of precipitation expected during their taxi from gate or de-icing pad to takeoff point.

The main display provides real-time weather information, including radar and local surface weather observations. A radar-based display includes vectors for storm tracking, as well as radar echo reflectivity, velocity and derived precipitation rate and type fields. Local weather service and WSDDM weather stations provide updates of current temperature, wind, humidity, precipitation rate and accumulation every minute. One-hour forecasts of liquid-equivalent precipitation are shown for point locations around the airport.

The ‘checktime’ display feature uses real-time one-minute data, including liquid water equivalent rates, from WSDDM weather stations to provide current estimates of anti-icing fluid failure times using holdover timetable guidelines. Ground and air crews use this to check the time available between applying anti-icing fluids and takeoff.

The liquid equivalent display uses real-time one-minute data from WSDDM weather stations to provide a precipitation intensity (light, moderate, heavy) using liquid water equivalent measurements. This information is prominently displayed, to aid airline staff in making decisions about fluid types and holdover times.

This coming winter, WSDDM systems will be deployed at Denver (DEN), Pittsburgh (PIT), Chicago (ORD) and Minneapolis St. Paul (MSP). Additional sites are planned for the future.


CIP/FIP and WSDDM are examples of new, automated products available to the aviation end-user for avoiding hazardous icing conditions. CIP/FIP data are not only available in graphical form for use in flight planning, but as three- and four-dimensional gridded data fields which are suitable for datalinking to cockpit displays for the user to view discrete regions or different formats in a customised display. WSDDM represents the integration and presentation of data with products specially tailored to the needs of the user in making wise decisions about preflight de- and anti-icing. The door has opened on a new set of automated products. What will the future bring?

  • Higher-resolution products will allow tactical decision aids to be available in the cockpit.
  • Broader geographic coverage and harmonisation of inflight icing algorithms from different sources will ensure seamless information and displays over transcontinental air routes.
  • All data will be incorporated into a common, shared database to be made available in real time.
  • A team approach bringing researchers and users together early in the development process will ensure that ‘Decision Support Tools’ will supply the right information, in the right format, at the right time.

About the author

Dr. Politovich is a Councillor of the American Meteorological Society and member of the American Institute of Aeronautic and Astronautics’ Atmospheric and Space Environment Committee.

As head of the InFlight Icing Product Research Team, Dr. Politovich leads in-flight icing research efforts under the FAA-sponsored Aviation Weather Research Programme. In addition to coordinating activities under this programme, her contributions include analyses of weather conditions leading to icing, development of a meteorology-based icing severity index and the use of in situ and remote sensors to diagnose icing conditions. She served as Co-Operations Director for the four field efforts supporting basic atmospheric research for this programme.