Development and Testing of a High-Resolution Model for Tropospheric Sulfate Driven by Observation-Derived Meteorology.

A high-resolution three-dimensional Eulerian transport and transformation model was developed to simulate concentrations of tropospheric sulfate for specific times and locations; it was applied over the North Atlantic and adjacent continental regions during October and November, 1986. The model represents emissions of anthropogenic SO_2 and sulfate and of biogenic sulfur, horizontal and vertical transport, gas-phase oxidation of SO_2 and dimethylsulfide, aqueous-phase oxidation of SO _2, and wet and dry deposition of SO _2, sulfate and methanesulfonic acid (MSA). The meteorological driver is the 6-hour output from the forecast model of the European Centre for Medium-Range Weather Forecasts. Results, examined in detail for October 15 and October 22 at 6Z, are related to weather patterns, and exhibit rich temporal and spatial structure. The characteristic (1/e) temporal autocorrelation over the central North Atlantic averages 13 hours; 95% of the values were 25 hours or less. The characteristic distance of spatial autocorrelation depends on direction and averages 1600 km; with low and high 10 ^{rm th} percentile values of 400 km and 1700 km. Daily average model concentrations at the lowest level accurately represent the spatial variability, temporal episodicity, and absolute magnitudes of surface concentrations at available monitoring stations; over 50% of the model results are within a factor of 3 of the observations. Over 50% of weekly model wet deposition amounts are within a factor of 3 of observations at available monitoring stations. Over the period modeled, contributions from anthropogenic sources to sulfate over the mid North Atlantic Ocean ranged from 44 to 66%, contributions from biogenic sources ranged from 6 to 12%. Calculated average yields for sulfate (47 to 72%) and MSA (13%), and turnover times for SO_2 (2 to 3 days) and MSA (4 to 8 days) are comparable to previous estimates; however, these quantities depend on meteorological conditions and on the geographic and vertical distributions of the material.