Reoccurring low clouds have a major impact on air traffic at the San Francisco International Airport (SFO). Accurate, short-term forecasts of the time that the clouds will dissipate provide critical information for efficient air traffic management. Deterministic forecasts of the time of dissipation have been developed over the past few years by the SFO Marine Stratus Project. Recently-developed automated air traffic management tools require probabilistic forecasts. We shall describe some numerical methods, which have been used to derive probabilistic forecasts from these deterministic forecasts and to evaluate their comparative skill.

At issue is the possibility that there might be a significant difference in the skill of these forecast systems. Traditional skill measures for deterministic and probabilistic forecasts differ to the degree that they do not provide a mechanism for resolving this issue. In the course of these investigations, we have developed a common ground for making such comparisons. We shall present this technology and its application to the SFO forecast systems.