From 1994 through 1999 fatal airplane crashes in the U.S. averaged 399 per year. Over the same period, the average number of fatal motor vehicle crashes was 37,530 per year with 583 per year related to fog. Improved fog forecasts could help to reduce the rate of fatal vehicle crashes, but before fog forecasts can be effective for en-route travelers, the forecasts must be location- and time-based with greater highway specificity.

This paper explores the feasibility of using artificial intelligence methods to aid in fog forecasting for surface transportation. The lack of weather observations, especially visibility, along highways makes forecast initialization and verification difficult. An evaluation of the use of data collected from airport observations to support highway fog forecasting applications is presented. This work is in conjunction with ongoing research into effective decision support systems for improved surface transportation weather decision-making.