A nonlinear regression model estimating single source concentrations of primary and secondarily formed PM2.5

Abstract

Various approaches and tools exist to estimate local and regional PM2.5 impacts from a single emissions source, ranging from simple screening techniques to Gaussian based dispersion models and complex grid-based Eulerian photochemical transport models. These approaches either lack a realistic chemical and physical representation of the atmosphere for secondary PM2.5 formation or in the case of photochemical models may be too resource intensive for single source assessments. A simple non-linear regression model has been developed to estimate annual average downwind primary and secondarily formed PM2.5 nitrate and sulfate from single emissions sources. The statistical model is based on single emissions sources tracked with particulate source apportionment technology in a photochemical transport model. This non-linear regression model is advantageous in that the underlying data is based on single emissions sources modeled in a realistic chemical and physical environment of a photochemical model and provides downwind PM2.5 impact information with minimal resource burden. Separate regression models are developed for primary PM2.5, PM2.5 sulfate ion, and PM2.5 nitrate ion. Regression model inputs include facility emissions rates in tons per year and the distance between the source and receptor. An additional regression model input of receptor ammonia emissions is used to account for the variability in regional ammonia availability that is important for PM2.5 nitrate ion estimates.