Racial and ethnic disparities in exposure to short-term NO2 air pollution in California during 1980-2022

Abstract

Historical racial and ethnic disparities in short-term exposure to ambient nitrogen dioxide (NO₂) have rarely been investigated, primarily due to the lack of spatiotemporally resolved NO₂ data covering the historical period. In this study, we used publicly available geospatial data within a deep learning framework to estimate high resolution (approximately 1 km × 1 km grid; daily) NO₂ concentrations in California, United States, from 1980 to 2022. Chemical transport model outputs informed a priori geophysical information to help capture the spatiotemporal variations in NO₂ concentrations in the absence of satellite observations. The NO₂ models exhibited robust performance throughout the study period, with coefficients of determination ranging from 0.72 to 0.83 for grid-based 10-fold cross-validation. Despite significant declines in NO₂ exposures for all racial and ethnic groups over the study period, we observed increasing relative disparities in short-term NO₂ exposures among Hispanic or Latino, Not Hispanic or Latino African American or Black, and Not Hispanic or Latino American Indian, Alaska Native, Asian, and Pacific Islander groups relative to Not Hispanic or Latino White. In 2020, Hispanic or Latino individuals experienced nearly three times as many days with NO₂ concentration exceeding 50 µg/m³ compared to Not Hispanic White individuals, up from 32 % more days in 1980. This study quantifies spatiotemporal racial and ethnic disparities in short-term NO₂ exposures, a novel contribution to the environmental justice literature and critical information for targeted interventions.

Keywords: California; Deep Forest; Exposure disparities; NO(2) pollution; Temporal trends.