Modeling regional sulfate during the BRAVO study: Part 1. Base emissions simulation and performance evaluation

Abstract

The Regional Modeling System for Aerosols and Deposition (REMSAD) was employed to predict sulfate formation and transport in the Big Bend Regional Aerosol and Visibility Observational Study (BRAVO). Fine particulate sulfate was the main component of haze measured at Big Bend National Park (BBNP) during the BRAVO field measurement campaign, which was conducted July–October 1999. Predicted sulfate was evaluated against measurements collected from BRAVO and CASTNet (Clean Air Speciation and Trends Network) monitors. During the four month study period, average observed and predicted sulfate concentrations at the 37 BRAVO sites were 3.1 and 3.3 μg m −3, respectively, and 4.5 and 5.0 μg m −3 at the 62 CASTNet sites, respectively. Spatial and temporal biases were clearly apparent. Underestimations of sulfate during July and August were evident throughout the BRAVO monitoring network (fractional biases of −0.60 and −0.44, respectively), corresponding to a period of increased atmospheric transport from Mexico to Texas. Biases in estimated sulfate in the vicinity of BBNP may be attributed to an overestimation of predicted precipitation rates during July and August. Positive biases within the BRAVO network were apparent in October (fractional bias=0.39), when atmospheric transport from the eastern US was more prevalent. Overall performance statistics for sulfate predictions relative to the BRAVO and CASTNet sites were: correlation coefficient=0.61 and 0.90, respectively; fractional error=0.56 and 0.35, respectively; fractional bias=−0.10 and 0.04, respectively. This simulation provided the basis for a series of emission sensitivity simulations that were used to estimate a sulfate source apportionment for BBNP.