Evaluating in-flight aviation icing indices using ARM observations

Description

Clouds containing liquid water at sub-freezing temperatures present a danger to aviation. The super-cooled liquid water can freeze on contact with the aircraft wings and nose, affecting the aerodynamics and instruments. As a result, forecasts of regions with a likely risk to aircraft are routinely provided by forecasters to help with regional flight planning. Incidents of icing on aircraft get reported as light, moderate or severe and consequently the forecasts are issued with the same 3-level intensity scale using an "icing index" that combines severity and likelihood. Potential verification of icing forecasts is complicated however, by regions of perceived risk being understandably avoided, hence leading to less incidents and fewer opportunities for a hit when calculated statistical scores. In addition, the severity is notoriously hard to verify. Although the severity of an icing event is undoubtedly affected by the local meteorology it is also very dependent on other factors. These include: the specific aircraft involved, the presence of de-icing equipment and the temperature of the fuselage, which depends on whether the aircraft is in the ascending or descending phase of flight. Here the two orthogonal aspects of an icing index: severity and likelihood are explicitly separated. Remote-sensing data from several ARM sites are used to flag the frequency of occurrence (or likelihood of encountering) super-cooled liquid cloud. A global weather forecasting model is then used to predict this frequency of occurrence as a function of height, for each site and for several month-long periods. Different methods of calculating likelihood of encountering super-cooled liquid cloud can be computed and compared, hence adding confidence to which formulation to use as part of a new icing index to be used by aviation forecasters in the future.