When dealing with forest fires, smoke managers need to identify smoke sensitive areas (hospitals, communities, main roads, …) and to find and use appropriate methods to reduce smoke impact in those areas. In order to mitigate the impacts of smoke from forest fires to public health and safety, it is critical to better understand smoke dispersion behavior. For that purpose, smoke dispersion models are important tool. But, modeling of smoke dispersion and transport from wild fires is much harder than modeling of smoke/pollution form a point pollution sources (such as power plant stack).
Dispersion models are usually using emission information and meteorological data as an input data. The outputs are usually concentration and trajectory information. In my study, I’m using dispersion model called FLEXPART (FLEXible PARTicle dispersion model), a Lagrangian transport and dispersion model, to look at smoke dispersion from prescribed forest fires and to investigate dispersion model results under different set of meteorological and emission conditions. As a meteorological input data for my dispersion model, I’m using ARPS-CANOPY outputs. ARPS-CANOPY is a modified version of ARPS model (the Advanced Regional Prediction System), where processes within a forest canopy are considered. Emission inputs used in dispersion modeling are Fire Emission Production Simulator (FEPS) outputs. FEPS is a system for simulation of fuel consumption and emissions. I’m also using advance statistical tool called emulator to help me to investigate all possible combinations of different parameters used in the model simulations, as well as probability assign to those parameters.