A Comparison Between Sounding Data and Model Data Used for Aviation Weather Hazards

The Integrated Meteorological System (IMETS) is a weather data system utilized by Air Force weather forecasters in support of Army Operations. On IMETS, the Atmospheric Sounding Program (ASP) derives forecasts for weather hazards such as icing, turbulence, and clouds. The ASP uses two data sources; conventional 1200 and 0000 UTC upper-air observations and output from a mesoscale model called the Battlescale Forecast Model (BFM). The BFM is typically run on a 500*500 km or smaller area and employs a horizontal resolution of 10 km with 16 terrain-following vertical levels. The model produces forecast output at 0-, 3-, 6-, 9-, 12-, 18-, and 24-h from the base time and places these data into a gridded meteorological database. From there, weather hazards are derived in a post-processing routine for each grid point. To forecast clear-air turbulence intensity, the Richardson number and Panofsky number are used along with a set of rules that check for consistency of the output. Icing type and intensity are derived at each sounding site or model grid point using the AFWA RAOB tool, with some modifications in the routine. Clouds are forecasted by using a routine which is a rule-based tool that uses season, temperature, time of day, station location, and relative humidity of the atmospheric layers to predict cloud amounts, ceilings, and cloud depth. This study discusses the results of a comparison between sounding data and model-derived data for point forecasting of turbulence, icing, and clouds. Results of this three-year study, employing pilot reports as a verification source, indicate that the mesoscale model forecasts turbulence more accurately than the upper-air observations, while upper-air observations proved to be a better data source than the model output for icing forecasts. Statistical analysis indicate no significant differences in verification of cloud forecasts using either data source.