Synergistic geostationary retrievals of aerosol properties using a combined implementation of the Dark Target and GRASP retrievals

While there has been significant progress in aerosol characterization from satellite remote sensing, limited temporal sampling has limited our understanding of the diurnal cycle. The new generation of sensors in geostationary orbit (GEO), such as the Advanced Baseline Imagers (ABI) on GOES-16 and GOES-17, have more channels and better radiometric accuracy than prior geostationary instrumentation, enabling high quality aerosol retrievals at previously unobtainable temporal resolutions. We combine novel with well-established retrieval techniques to obtain improved estimates of aerosol properties from these new geostationary satellites. For data preparations (e.g., cloud masking, pixel selection and gas absorption corrections), we use the methods employed by the well-established Dark Target algorithm. These preprocessed reflectances are then be fed into the advanced and highly flexible Generalized Retrieval of Aerosol and Surface Properties (GRASP) algorithm, which allows for highly customizable aerosol and surface priors. To derive the assumptions that ultimately best use the information in the GEO observations, we begin with synergistic retrievals that input both satellite observations and collocated ground-based measurements made by the Aerosol Robotic Network (AERONET). The resulting information on aerosol size, refractive index and shape then drives decisions pertaining to the application of the retrieval to exclusively geostationary-based observations. This procedure helps ensure that the retrieval assumptions map the observed reflectances to the relevant aerosol properties in the most consistent manner possible. Additionally, we make substantial use of GRASP's ability to impose multi-pixel temporal and spatial smoothness constraints, which enables more complete aerosol property retrievals by exploiting the high temporal frequency of GEO observations.