Leveraging High-Resolution Satellite-Derived NO2 Estimates to Evaluate NO2 Exposure Representativeness and Socioeconomic Disparities

Abstract

Research has typically estimated NO2 concentrations over several kilometers; thus, NO2 data at finer spatial resolution remain limited. This study used tropospheric NO2 data from the TROPOspheric Monitoring Instrument (TROPOMI) and traffic-related land use parameters to estimate long-term average NO2 concentrations at a spatial resolution of 500 m in South Korea from 2018 to 2022. Our satellite-land use hybrid regression model showed reasonably high predictability with a cross-validation R2 of 0.81, mean absolute error of 2.28 ppb and root mean squared error of 2.85 ppb. Leveraging these high-resolution data, we assessed the representativeness of ground monitors for population exposure by comparing population-weighted NO2 concentrations from estimated and measured data. Across 17 metropolitan cities and provinces, the ratios of population-weighted estimated to measured NO2 ranged from 0.62 to 1.12, with the ratio of 1 exhibiting the most representative monitoring networks. We further investigated disproportionate NO2 exposures based on socioeconomic status, revealing that NO2 exposures were consistently higher in local districts with higher socioeconomic status because of the unique historical backgrounds of rapid economic development and urban infrastructure design in South Korea. Using high-resolution NO2 data can lead to more comprehensive and precise exposure assessments, enhancing public health and regulatory applications.