Spatial/temporal characteristics of air quality prediction errors in the Bluesky modeling framework

Federal land management agencies and the Environmental Protection Agency conducted field trials in 2005 of an air quality prediction system known as BlueSky, developed by an interagency consortium in the Pacific Northwest. Of particular interest is the ability to predict the impact of wildland fire on air quality, including fine particulate concentrations at the surface. This case study focused on the spatial and temporal variability of PM2.5 prediction errors in the field trials on the Salmon-Challis National Forest along the Idaho-Montana border. We examined the variability of the prediction errors as a function of spatial and temporal averaging. We conclude by offering a framework for quantifying uncertainties in the particulate concentration predictions for the consideration of users of the BlueSky modeling framework.