A 3-dimensional modeling and simulation study was performed, for the evolution and characterization of the smelter plume during winter field study around an industrial complex in Quebec, using the BLFMAPS-a Mesocscale Boundary Layer forecast and Air pollution prediction system. This system contains a mesoscale meteorological boundary layer forecast model (BLFM) and a set of air pollution transport, dispersion and deposition modules. The modeling system was utilized to simulate meteorology, air concentration, dry deposition and wet deposition of particulate during the field study period in February 2000. With the help of twice daily objectively analyzed weather data from Canadian Meteorological Center the BLFM will predict meteorological parameters for 24 hours on a fine spatial resolution of 5 km over 400 km x 400 km domain centered over the smelter complex in Quebec. The predicted meteorological variables along with the mixed layer depth and turbulent parameters as function of 3-dimensional space and time are used in the air pollution modules to predict hourly concentration and deposition as function of time in the 3 dimensional model domain. The emission rates for total particulate and lead particles were estimated from the 1999 Noranda smelter emissions and the data from aircraft monitored total particulate to the lead particle mass ratio. We also simulated lead particle concentration and deposition as function of two particle size bins (0.25 to 0.5um and 16 to 24 um). The aircraft data suggest that the lead particles exist in these two sizes. These numerical experiments will provide the sensitivity of the model to particle size, and density on the predicted concentration and deposition. Particle size dependent deposition showed some interesting pattern and phenomenon. Preliminary analysis of model results with the aircraft data for some selected flights suggests that the model-simulated values are quite reasonable.