A Modified Fog Detection Algorithm Developed at the Miami Weather Forecast Office

Limited visibility due to fog poses a significant danger to travelers. Accurate fog forecasts provide travelers with ample time to make preparations for low visibility conditions. Previous to this study, the Miami’s NWS Weather Forecast Office used a combination of factors to forecast fog, mainly moisture and wind speeds derived from forecast and observed soundings. However, a quantitative approach based on research results has not been used yet in South Florida. Due to its subtropical climate that combines marine, Everglades, and urban conditions, traditional fog forecasting techniques are not directly applicable to South Florida. Here we perform a retrospective analysis of the past three fog seasons in South Florida (from November 1 to April 30) using METAR data, radiosonde profiles, and in some cases satellite imagery. The reanalysis applies the United Parcel Service (UPS) Airlines technique based on crossover temperature and modified Richardson number, to all cases. The results demonstrate that a combination technique that uses the crossover temperature in conjunction with a 15-knot maximum threshold of 925 mb winds for fog formation yields the most accurate fog formation predictor at the study location. The crossover temperature is defined as the minimum dew point observed during the warmest daytime heating hours. Fog is forecast when the shelter temperature is expected to cool to a few degrees below the crossover temperature, rather than a few degrees below the dew point. This combination technique successfully predicted 65% of all fog events compared to 38% when the crossover temperature was used in conjunction with the modified Richardson number. Possibly the subtropical location limits the frequency of turbulent mixing due to synoptic features, which may explain why the 925 mb winds outperform the modified Richardson number used in the UPS airlines technique. 1. Motivation of Work Reduced of visibility due to fog can pose a significant hazard to travelers. In fact, a 1995-2000 study by the National Transportation Safety Board found that 63% of all weather-related fatal aircraft accidents were due to Low Instrument Flight Rules (LIFR)/fog situations (Pearson 2002). Fog is also notorious for multi-vehicle accidents for land travelers, with the capability to injure or kill dozens of people (Croft 2002). While the American Meteorological Society Glossary defines fog as reducing visibility below 1km, for the purpose of this paper fog will be defined as a reduction in visibility below 2 miles (3.7 km) due to a surface layer condensation. The site chosen for this study is Tamiami Airport, located in the suburbs of Southwestern Miami-Dade County. Due to the nature of the site, the primary fog analyzed is radiation fog. While the conditions favorable for fog formation, such as clear skies, high humidity in the boundary layer, low turbulence, and a surface temperature inversion, are well known, formation is very sensitive and an inaccurate forecast for any one of them. Even a slight change in weather conditions, can produce a faulty fog forecast. Here we analyze similarities among fog events over the past 3 years and assess the accuracy of two fog forecasting techniques: the Crossover Technique (Baker et al. 2002) and the Modified Richardson Number (Baker et al.