Neighborhood-Scale O3, NO2, and PM2.5 in Washington, D.C. from a 1 × 1 km2 WRF–SMOKE–CMAQ Framework

Abstract

Quantifying air pollution concentrations at fine spatial scales is important for resolving intra-urban exposure variability and identifying sources that disproportionately impact population groups across locations. We developed a neighborhood-scale resolution (1 × 1 km2) WRF–SMOKE–CMAQ modeling framework to characterize O3, NO2, and PM2.5 concentrations in Washington, D.C. in 2017. In evaluations against EPA Air Quality System observations and the National Climatic Data Center meteorological data, WRF reproduced the temporal variability of temperature and specific humidity with reasonable agreement (average r = 0.8 and 0.7, respectively), although biases varied across seasons. CMAQ reproduced seasonal and diurnal patterns of O3 and PM2.5 with r up to 0.7 for hourly O3, though NO2 was negatively biased (−14% to −29%), reflecting challenges in resolving near-road gradients. We identified pronounced intra-urban concentration variability of NO2, whose coefficient of variation was greater than that of O3 and more than of three times that of PM2.5. O3 concentrations were lower in all months by 4% to 11% in the urban core compared to outside the urban core due to NOx titration/inhibition. Relative to 4 × 4 km2 resolution, 1 × 1 km2 simulations increased population-weighted exposure (PWE) for O3 (+0.5% to +1.5%) but decreased PWE for NO2 (−3% to −6%) and PM2.5 (−1.5% to −3%). VOC/NOx ratios indicated predominantly NOx-saturated chemistry within the urban core. This work presents the value and the remaining challenges of high-resolution regional air quality modeling in heterogeneous urban settings.