Dispersion model evaluation of PM2.5, NOX and SO2 from point and major line sources in Nova Scotia, Canada using AERMOD Gaussian plume air dispersion model

AERMOD was used to model the air dispersion of point and major line emissions of PM2.5 in Halifax and Pictou, NOX in Halifax and SO2 in Halifax, Sydney and Port Hawkesbury, Nova Scotia, Canada. Emission inventory data for 2004 were used in simulations within four, 50 km x 50 km, domains over annual, monthly and 1-hour averaging periods. Annual averaged surface concentration maps are reported. Modeled versus observed comparisons were made within each domain at the Government, National Air Pollution Surveillance (NAPS) monitoring sites (discrete receptors). Evaluation of the model was conducted on the annual, monthly and hourly results using a number of statistical methods that included R-2, fractional bias, normalized mean square error and the fraction of predictions within a factor of two of the observations. The AERMOD model evaluation showed that there was good agreement between the modeled and observed SO2 concentration for the annual and monthly comparison but less skill at estimating the hourly comparisons for SO2 in Halifax and Sydney. AERMOD showed poor model skill at predicting SO2 in Port Hawkesbury over the same averaging periods. The model evaluation for PM2.5 in Halifax, PM2.5 in Pictou and NOX in Halifax showed poor agreements and model skill. The surface concentrations from the point and major lines sources in all domains from all metrics were found to be well below the National Air Quality Standards. AERMOD has shown its utility as a suitable model for conducting dispersion modeling from point and line sources in Nova Scotia with good model skill for estimating annual and monthly SO2 concentrations in Halifax and Sydney. The study highlights the validity of using emission inventory data to estimate the surface impact of major point and line sources within domains containing complex terrain, differing land use types and with large variability within the annual meteorology.