A fuzzy case-based system for weather prediction

Weather forecasting is a complex process that involves numerous specialized fields of expertise. The output from computationally intensive numerical weather prediction (NWP) models forms the starting point of the forecasting process. Expert forecasters have both a general knowledge of large-scale weather systems and specific knowledge about the idiosyncratic behavior of local scale weather phenomena. These expert forecasters, in effect, bridge the gap on the local scale between simple persistence forecasting and the output from the NWP models [1]. The types of forecasts commonly made by expert forecasters include terminal aerodrome forecasts (TAFs), public forecasts and marine forecasts. In addition to these, many specialized forecasts are produced for power companies, oilrigs and others. Because much of the knowledge and expertise used by weather forecasters is analogical the possibility of augmenting the forecasters experience with the detailed knowledge of similar weather situations from several prior decades of recorded weather data seems most promising. This is the possibility offered by applying case-based reasoning techniques to the problem of weather forecasting. Recently intelligent systems (IS) using artificial intelligence (AI) techniques have been used to forecast visibility, marine fog, precipitation, severe weather and other climatological conditions [e.g., 2, 3, 4, 5]. The IS modeling approach is complementary to NWP modeling that uses computationally intensive dynamical, thermodynamical and statistical algorithms to produce large scale (hemispherical) static forecasts. These large-scale static forecasts are not sufficiently stable on a small scale where