Forecasting icing severity is one of the more difficult jobs in aviation weather. One has to forecast a poorly-defined condition (current severity definitions are based on an airplane and pilot's response to the icing environment), using inadequate information (no direct measurement of cloud liquid water content). However, aircraft operating requirements force the issue and we must find a way to do this satisfactorily.

This paper will consider three approaches. The first is to base an icing severity index on the amount of supercooled liquid water present. Considerable work has been done in this area by meteorologists and aerodynamicists, whereby the amount of liquid in the atmosphere has been related to the accretion on an airframe. However, even the most recent versions of operational weather models are not capable of providing the accuracy in liquid water content needed to implement a scheme at this time. It's a difficult task to ask of even the most sophisticated cloud physics parameterization package.

The second approach is to compare current icing PIREPs to atmospheric conditions, then extrapolate those into the future. This is one way that many aviation weather forecasters use to create their icing severity forecasts. Considering that at any given time icing covers on the order of 10% of the total airspace volume over the CONUS, we are actually dealing with a relatively infrequent event and interesting statistical methods can be brought to bear on the problem. Many locations and times of day lack PIREPs, and, different airplanes and pilots will report icing severities differently. Both factors contribute additional elements of uncertainty to the problem.

The third approach we will consider is to develop weather scenarios conducive to higher (and lower) liquid water contents in clouds. The Integrated Icing Diagnosis and Forecast Algorithms developed at NCAR use this approach to determine where regions of supercooled liquid water, and thus icing environments, are likely to occur. Extension of this methodology, based on sound physical principles and an experience based gained from over a decade of forecasting for icing field efforts, seems to be a reasonable approach.

The aviation user community wants and needs information on areas of moderate or greater icing, and of severe icing conditions. NCAR is working on inclusion of icing severity forecasts in their automated icing diagnosis and forecast products that meet these requirements, and progress on this daunting task will be presented.