Leveraging a Full Day of Observations from ABI on GOES-16/17 to Retrieve Aerosol Amount, Size, and Absorption

Individual observations of shortwave radiance with near-daily global coverage are currently only available from instruments that are limited to a single viewing angle. While spectral dependencies in these observations allow Aerosol Optical Depth (AOD) and some other basic parameters to be reliably extracted, an instantaneous single-angle measurement lacks sufficient information to retrieve many other desired aerosol quantities. Geostationary (GEO) satellites image a given ground location many times over the course of a day, therefore acquiring data over a diverse set of solar geometries and, in some locations, overlapping fields-of-view between two sensors provides observations from multiple viewing geometries near-simultaneously. Here, we apply new multipixel retrieval techniques with temporal smoothness constraints to asynchronous GEO data in order to exploit the additional information contained in the solar and viewing angle dependence of the measurements. Specifically, observations taken over the course of a day from both Advanced Baseline Imagers (ABI) onboard GOES-16 and GOES-17, are concurrently fed into the Generalized Retrieval of Aerosol and Surface Properties (GRASP) algorithm. GRASP is then configured with a prior assumption that some aerosol properties and the land surface parameterization vary slowly in time. The reduced degrees of freedom in the surface and aerosol state then permits the retrieval of parameters like particle size and absorption that are difficult to obtain with single pixel, single view angle approaches. Additionally, in order to determine a set of a priori retrieval assumptions that best uses the information content of all GEO observations, we begin with synergistic retrievals that include both measurements made by collocated ground-based Aerosol Robotic Network (AERONET) sun/sky photometers and both ABI instruments. The resulting GEO and AERONET jointly retrieved surface properties, aerosol size information, refractive index, and nonspherical fraction then drive decisions pertaining to retrievals exclusively based on geostationary observations. This procedure helps ensure that the retrieval assumptions map the reflectances observed from space to the retrieved aerosol properties in the most consistent manner possible. The geostationary-only retrieval results are then validated with standard AERONET inversion products to assess the accuracy of the new approach.