TROPOMI CO and NO₂ as Observational Constraints on the Sources of Air Pollution Inequalities in US Cities

Carbon monoxide (CO) and nitrogen dioxide (NO₂) are combustion pollutants that disproportionately harm communities of color in US cities. While distributive inequalities are established, we have lacked observational constraints on variability in air pollution mixtures and the sources that drive exposure disparities. We present the first application of the TROPOspheric Monitoring Instrument (TROPOMI) to describe census tract-scale CO inequalities and population-weighted NO₂/CO enhancement ratios (ΔXNO₂/ΔXCO) in six major US urbanized areas (UAs). Because CO has a long atmospheric lifetime, we develop an approach to describing CO inequalities using column-averaged dry-air mole fractions normalized by the daily urban variation (XCO_norm). Annual daily mean inequalities in XCO_norm are significant in some UAs, e.g., Los Angeles, Phoenix, New York City, and Chicago, where they are 7-11% for Black, 7-15% for Latino, and 1-11% for Asian residents compared to non-Hispanic/Latino white residents; however, in Houston and Atlanta, inequalities in XCO_norm are near or indistinguishable from zero. We report direct satellite evidence of cumulative exposure burdens as CO and NO₂ disparities are correlated across cities (r = 0.89). We compare inequalities on weekdays and weekends and with conditions that differently affect short- and long-lived, primary and secondary species, with evidence suggesting that CO inequalities are influenced not only by gasoline vehicles but also chemical production. TROPOMI ΔXNO₂/ΔXCO indicate that combustion mixtures in primarily Black, Latino, and Asian neighborhoods are disproportionately influenced by heavy-duty diesel vehicles (HDDVs), even on weekends. To address exposure disparities, efforts should focus on differences in source composition not just magnitude. Spatial discrepancies between ΔXNO₂/ΔXCO and the 2020 National Emissions Inventory (NEI20) suggest the NEI20, at least without downscaling, underpredicts the extent to which Black, Latino, and Asian neighborhoods are affected by high NO₂/CO sources, e.g., HDDVs. Such errors in the NEI20 potentially direct attention and resources away from HDDV controls with environmental justice benefits.