Improved estimation of PM2.5 using Lagrangian satellite-measured aerosol optical depth

Abstract

Accurate estimates of fine suspended particulate matter (PM2.5) concentrations are important in air quality and epidemiological studies. Aerosol optical depth (AOD) retrieved by the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite instrument has been used as an empirical predictor to estimate ground-level concentrations of PM2.5. However, these estimates usually have large uncertainties. The main objective of this work is to assess the value of upwind (Lagrangian) MODIS-AOD as predictors in empirical models of ground-level PM2.5. We also explored the reconstruction of missing MODIS data and developed a daily average uniformly-gridded AOD product. The empirical models developed in this work were tested in ten different sites across the continental United States. Multiple linear regression models that included Lagrangian AOD along in situ AOD as predictors showed statistically significant improvement over the simple linear regression models (PM2.5 and in situ AOD). A set of seasonal categorical variables was included in the regressions to account for the variability of regression performance with respect to seasons. The extended multiple linear regression models exhibited statistically significant improvement over the simple and multiple linear regression models that only contained AOD as predictors.