LAUSR

Abstract To meet the requirements of the Clean Air Act's “good neighbor” provision, it is necessary to quantify the interstate transport of air pollution and determine if upwind states significantly contribute to noncompliance with the National Ambient Air Quality Standards (NAAQS) at downwind states. This study aims to estimate the contributions of ten southeastern states (Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, Tennessee, Virginia and West Virginia) to ozone (O3), and identify significant associations with nonattainment or maintenance at monitors across the eastern United States (US) in 2017. Two widely used air quality models, the Comprehensive Air quality Model with extensions (CAMx) and the Community Multiscale Air Quality model (CMAQ), using different versions of the Carbon Bond (CB) chemical mechanism, were applied to predict 2017 O3 design values (DVFs) from 2011 O3 design values (DVCs) at the US Environmental Protection Agency's (EPA) Air Quality System (AQS) sites throughout the eastern US. Since 2017 design values are already published, the accuracy of the EPA-recommended modeling methodology for projecting current design values into the future only by changing the emissions was evaluated. The assumption that good model performance in the base year would result in accurate future year ozone projections was tested. In general, CAMx/CB6r2 simulated higher DVFs than CMAQ/CB05, with an average difference of 0.5 ppb. As a result, this study identified 24 sites in the eastern US as nonattainment (DVF ≥ 76 ppb) with CAMx/CB6r2 compared to only 16 sites with CMAQ/CB05. Using monitor cell relative response factors (RRFs) instead of 3×3 cell matrix maximum RRFs (the approach used in EPA's proposed Cross-State Air Pollution Rule (CSAPR) modeling) (EPA, 2015a), led to higher DVFs at some monitors and lower DVFs at others. Source apportionment analysis with the CAMx Anthropogenic Precursor Culpability Assessment (APCA) probing tool revealed that NOx emissions dominated over VOC emissions in contributing to O3 in downwind states. Contributions of total anthropogenic NOx emissions and contributions of Electric Generating Unit (EGU) NOx emissions from the southeastern states to O3 at nonattainment and maintenance sites were quantified separately using CAMx/APCA as well as CMAQ/zero-out simulations. Contributions obtained from CAMx/APCA were typically larger than the contributions obtained from CMAQ/zero-out. Therefore, a “significant contribution” determined by CAMx/APCA may be insignificant according to CMAQ/zero-out. “Significant contribution” was defined as an impact larger than 1% of the 2008 O3 NAAQS (i.e., ≥ 0.76 ppb). Alabama, Kentucky, Mississippi, Tennessee, Virginia and West Virginia were linked to 2, 3, 1, 1, 6 and 5 nonattainment sites with CAMx/APCA, while only Kentucky, Virginia and West Virginia were linked to 2, 3 and 2 nonattainment sites with CMAQ/zero-out. Additionally, for significant linkages, EGU NOx contributions were 10–55% of the total anthropogenic NOx contributions. The information produced in this study can be used by southeastern states to help develop “good neighbor” State Implementation Plans (SIPs).