Leveraging crowd-sourced environmental data to assess air pollution exposure disparity: A case of Los Angeles County

The COVID-19 pandemic has underscored the significance of air pollution exposure; however, its impacts on the exposure disparity between disadvantaged communities (DACs) and non-DACs remain understudied. We utilized crowd-sourced open data from the PurpleAir website, a widely used low-cost sensor network, and data from CalEnviroScreen 4.0, a tool for identifying disproportionate pollution burdens in California, US, to investigate air pollution exposure in Los Angeles County, CA. We compared particles with diameters smaller than 2.5 µm (PM_2.5) concentrations before and during the pandemic (March 2019 - March 2020 vs. March 2020 - March 2021) across eight regions, with a focus on DACs that often have high proportions of low-income and people of color residents. Some DACs experienced higher-than-average concentrations when lockdown measures were lifted, with a higher percentage of days exceeding the 35 µg/m³ threshold compared to non-DACs. We confirmed persistent air pollution disparities between DACs and non-DACs, as indicated by both PurpleAir data and EPA regulatory monitors. We also found that the impact of traffic and land use factors on PM_2.5 concentrations became more consistent across locations during COVID. Our research underscores the viability of leveraging crowd-sourced data to identify air pollution exposure disparities and highlights the urgency of targeted interventions (e.g., telecommuting, industrial policies) to address the disproportionate burden of air pollution on vulnerable communities, particularly during and after crises. Further research is warranted to expand this approach to advance environmental justice efforts in diverse contexts.