The purpose of this study was to improve aircraft icing forecast algorithms in complex terrain below 10,000 feet (MSL). The area of study is the vicinity of Mt. Washington, New Hampshire, and data was gathered during the period from December 2000 through April 2001. Summit surface conditions on Mt. Washington were collected using standard hourly observations and two sensors mounted on the summit tower to collect continuous icing intensity data. During this period, the 00Z and 12Z ETA model forecasts and nearby radiosonde data were archived to provide input data for a decision tree algorithm outlined in Air Weather Service Pamphlet (AWSP) 105-56. Air Force Weather Agency and Aviation Weather Center icing forecasts were also archived for comparison purposes.

It is hypothesized the enhanced moistening and adiabatic cooling that results as air is lifted over the topography creates a greater frequency and intensity of icing events. The accuracy of each icing algorithm was calculated using summit icing conditions as verification. Initial results showed the algorithms tended to underpredicted the amount of icing that occurred on Mt. Washington. Further analysis was done to determine if modification of an algorithm would result in more improved icing forecasts.