The U.S. Army Research Laboratory (ARL) has developed a mesoscale model; the Battlescale Forecast Model (BFM), which is a hydrostatic model that employs a horizontal resolution of 10 km with 16 terrain-following vertical levels to a top of 7000 m above the highest elevation on the model domain. Since today's weapons and sensors are more sensitive to weather than in the past, many meteorological variables have been derived in the post-processing software package of the BFM. These parameters include such forecast elements as turbulence, icing, clouds, surface visibility, fog, thunderstorm probability, and precipitation parameters. Since ARL receives 15-km MM5 output from the U.S. Air Force Weather Agency (AFWA), it was decided to utilize the post-processing routines from the BFM on these MM5 data to investigate the results of two different modeling systems. The study was conducted during the cold-weather seasons from 1999 to 2001 and emphasized several of the variables in a wide range of meteorological conditions. Results of this project show that model biases play a significant role in the output of aviation hazards. As an example, icing output can give clues to moisture biases in a model, and in this case the BFM underforecasts icing in the mid-levels while the MM5 overforecasts icing in that layer. Cloud forecasts show that the skill is highest in the lowest levels, where model resolution is highest and skill scores are significantly reduced with increasing height. This project and the results will prove that common software is not ideal for aviation hazards and there should be an effort to derive weather hazards based on the model trends and model physics.