ADVANCING AIR QUALITY MODELS FOR REGULATORY APPLICATIONS

Abstract

Air pollution is one of the top risk factors for public health burden. These risk factors are exacerbated on a local scale for sulfur dioxide (SO2). The estimates of these health risks have been focused on particulate matter and ozone which fall short of showing the full picture on how local scale air pollution is different from those on a regional/state-based exposures. Additionally, the capabilities of complex atmospheric models and observational methods does not fill the data gaps that exist with sulfur dioxide. In this study, we present a computationally feasible methodology by using the American Meteorological Society (AMS)/Environmental Protection Agency’s (EPA) Regulatory model, AERMOD, a local scale gaussian dispersion model to fill the data gaps in surface level concentrations near large point sources of SO2 emissions. We also assess the potential for leveraging advanced photochemical model output for background concentration characterization. We study sources based on the volume of annual emissions as well as the availability of regulatory in-situ SO2 monitors near those sources for model evaluation. We find that local-scale dispersion models need to be leveraged for quantifying SO2 exposures for people living near an emissions source by allowing discerning of sub-grid scale variability issues associated with coarser photochemical models. Near-source SO2 concentrations have a wide range of values around a source which are averaged out on those coarse grids. At the four study locations, the ratio maximum SO2 concentration at a point in the grid is ~2 to 7 times higher than the average of all receptors in the grid. The maximum concentrations are 350.39, 82.89, 75.95, and 259.70 µg/m3 for sites A, B, C, and D respectively while the average is 70.12, 11.89, 27.89, and 134.65 µg/m3. We also find that the maximum 1-hour SO2 concentration occurs within 4 km from the source while the average distance of an SO2 monitor from a source is 45.10 km. We find that only

32% of the large point sources of SO2 emissions have at least one SO2 monitor within a 15 km radius. This is relevant in determining a threshold for studying the environmental justice impacts of source-specific pollution exposure. When concentrations vary over such a small distance, public health exposure estimates should be derived after giving due consideration to these spatial variabilities. For background concentration characterization, we find that the network of in-situ monitors is insufficient to capture the peak concentrations occurring near the source for SO2 in gas phase. Furthermore, we assessed alternatives for determining background concentrations based on photochemical model output and found that for isolated sources, the AQS-based background overpredicts for annual and median background scenarios. In contrast to this, the Community Multiscale Air Quality (CMAQ) model-based background underestimates the ambient impacts for both the yearly average and 50th percentile basis for relatively small emission sources, whereas for a large point source, the CMAQ grid cell background concentrations overestimate the impact. This work illustrates the potential for local-scale air quality models to be leveraged for fine-scale public health exposure assessment. This can further help fill the gaps for the public health benefits associated with gaseous pollutant exposure.