Comparing estimates from the R-LINE near road dispersion model using model-derived and observation-derived meteorology

Observed meteorological conditions, usually measured at airports or weather monitoring stations, have long provided the only source of meteorology for many Gaussian air pollution dispersion models. This introduces uncertainty and limitations in numerical model estimates, especially for locations of interest far removed from these monitoring stations. Hence, it is advantageous to employ predicted meteorology from a prognostic meteorological model as a substitute. The objective of this study was to compare estimates from the R-LINE near road dispersion model at three inland sites and one coastal site in Connecticut using observation-derived (weather station) and model-derived (Weather Research and Forecasting Model) meteorology. Both the graphical and statistical comparisons indicated less pronounced discrepancies in model estimations in the time periods generally characterized by unstable atmospheric conditions than those characterized by stable atmospheric conditions. There were also more pronounced differences at larger distances from roadways. Comparison of the estimated surface characteristic variables using both the observation-derived and model-derived meteorology displayed similar diurnal trends.