Abstract
An urgent forecasting problem is attacked utilizing the hypothesis that parameters relating to operationally significant weather cannot be generalized in space and time, nor in terms of a “pure physics” abstracted and removed from geographical realities. Forecasting criteria are viewed as discoverable and valid only with respect to experience at specific times and places. The empirical approach discloses an anomaly wherein drier evening air sometimes represents more of a fog hazard than does more saturated air. It also suggests a fog forecasting tool not in general use: streamline analysis, with emphasis on the hyperbolic point preceding a cold front in low latitudes. An objective aid for Miami International Airport tailored to local realities yields, on independent data, higher skill scores than are shown on the corresponding regularly published subjective forecasts.
