INFERRING ATMOSPHERIC AEROSOL PROPERTIES FROM SATELLITE OBSERVATIONS AND A GLOBAL CHEMICAL TRANSPORT MODEL
Atmospheric aerosols have significant impacts on climate and human health. However, the exact magnitude of the climate and health effects of aerosols remains highly uncertain, due to the large variability in aerosol physical and chemical properties. The use of satellite observations of aerosol properties in conjunction with global chemical transport models can improve our understanding of the interactions of aerosols with radiation and their impacts on health. Using GEOS-Chem coupled with a radiative transfer model, we develop a global simulation of the Ultraviolet Aerosol Index (UVAI) to interpret satellite UVAI observations. This simulation allows us to constrain the absorption by brown carbon (BrC) aerosol produced from biomass burning. Inclusion of absorption by BrC in GEOS-Chem reduces tropospheric hydroxyl radical by reducing the frequency of the photolysis of ozone. We calculate the direct radiative effect (DRE) of BrC, and find that absorbing BrC changes the global annual mean all-sky top of atmosphere DRE by +0.03 W m−2 and all-sky surface DRE by −0.08 W m−2. We interpret trends in satellite observed UVAI values using our UVAI simulation for 2005-2015 to improve our understanding of trends in global aerosol composition. Trends in absorption by dust dominate the simulated UVAI trends over desert regions. The UVAI simulation underestimates positive UVAI trends over Central Asia, possibly due to an increasing dust source from the desiccating Aral Sea that may not yet be represented by models. Trends in absorption by BrC dominate UVAI trends over biomass burning regions. Trends in scattering by secondary inorganic aerosol dominate UVAI trends over the eastern United States and eastern India. We estimate surface PM2.5 concentrations using information from satellites, simulation, and ground monitors for the years 2000-2017. These combined PM2.5 estimates benefit from recent updates to satellite AOD sources, developments in chemical transport models, and expanded ground monitor measurements. We find improved agreement between our PM2.5 estimates and ground monitors versus prior work. We use our improved estimates to calculate trends in PM2.5 both globally and regionally in order to understand the exposures of the global population.