Statistically-Based Method for Predicting Fog and Stratus Dissipation

摘要

The method is a success in producing forecasts for ceiling and visibility criteria that had never previously been examined. It is suggested that the 15 OWS incorporate this methodology into their operational forecasting routine. Ceiling forecasts at Dover AFB and McGuire AFB show improvements over conditional climatology ranging from 1-51% with an average improvement of 19.2% when verified against an independent data set. McGuire AFB visibility forecasts show an average improvement over conditional climatology of 3%. These findings are of particular importance to the Air Force in general and specifically to the 15th Operational Weather Squadron (15 OWS) who produces forecasts for these airfields. Demonstrating a method superior to conditional climatology is expected to provide improved forecasts and flight operations in this region. The two forecasts for Andrews AFB show relatively low mean square errors, but are unable to consistently improve on conditional climatology, demonstrating an average decrease in forecasting skill of 42%. Small samples of data could be the reason for the decrease in skill. The Dover visibility forecast also shows negative forecast skill, with an average decrease of 39%. The method is a success in producing forecasts for ceiling and visibility criteria that had never previously been examined. Further research into the forecasts could produce a powerful tool consistently able to defeat conditional climatology. It is suggested that the 15 OWS incorporate this methodology into their operational forecasting routine.