We harness satellite observations from a variety of instruments to characterize highly spatially resolved exposures: VIIRS observations in the day-night band combined with sky glow measurements provide estimates of artificial light at night, MISR observations of aerosol optical depth coupled with ground-level particulate matter concentrations provide estimates of source-specific air pollution, and MODIS observations of enhanced vegetation index (EVI) provide estimates of greenspace.
By linking these satellite-derived exposures with the residential addresses of 2,400 Southern California children aged 13 to 16 years, we assessed their single and combined associations on self-reported stress using mixed effects models with adjustments for subject-specific characteristics.
In single pollutant models with adjustment for sex, race, ethnicity, height, BMI, exposure to smoke, and community we find that ALAN and air pollution are significantly associated with increased stress while EVI is associated with decreased stress. Modeled jointly these patterns hold, where an IQR increase in ALAN is associated with a 0.57 (95% CI 0.05, 1.09) unit increase in stress, together with a -0.27 (95% CI -0.47, -0.06) unit decrease in stress per IQR increase in EVI.
In conclusion, satellite observations provide invaluable information to assess multiple environmental exposures to examine health effects. Our results indicate that in Southern California, a sprawling urban landscape, ALAN is the environmental factor most strongly related to children’s stress, while residential greenspace partially mitigates its detrimental impact. These findings have important policy implications related to urban planning and air pollution mitigation.