Aerosol Retrievals using GRASP from HARP CubeSat Data in Near Cloud Regions: the Impact of Different Cloud Masking Techniques.

Understanding aerosol interactions within the Earth system is vital to determining our planet’s energy balance. Aerosol-cloud interactions play a complex but critically important role in predictions of radiative flux in cloudy regions. Most global aerosol information is currently gathered through satellite-based measurements of radiometric intensities. The Hyper Angular Rainbow Polarimeter (HARP) CubeSat, an instrument developed at the University of Maryland Baltimore county, was launched into space in 2019 and began collecting data in April 2020 from a 52° inclination orbit. HARP measured the intensity and linear polarization state of light at the top of the atmosphere at four different wavelengths with 10 or more viewing angles per wavelength. During the two-year mission, HARP has captured several aerosol events, including the 2020 major “Godzilla” trans-Atlantic dust transport from the Sahara. HARP-2, the successor to HARP CubeSat, will fly aboard NASA’s upcoming Plankton Aerosol Cloud Ocean Ecosystem (PACE) satellite. HARP-2 will have improved radiometric and polarimetric accuracy compared to the original HARP. Our study presents aerosol retrievals in near cloud regions or Twilight zones using the Generalized Retrieval of Aerosols and Surface Properties (GRASP) algorithm. We observe the aerosol optical depth (AOD) variation near cloud regions and try to disentangle physical near cloud enhancements and retrieval artifacts that are produced by 3D radiative transfer effects. First, we compare GRASP estimates of AOD retrieved from HARP observation with other available data, namely those obtained from AERONET ground-based measurements as well as retrievals from VIIRS (on Suomi-NPP) and ABI (on GEOS-R). In this process we observed that distinguishing ‘cloud’ pixels from ‘clear’ in all available view angles referenced to a single ground position is very important before retrieving aerosol properties. Hence, cloud masking should be performed for all the view angles separately prior to retrievals. In response, we propose a new cloud masking technique for HARP CubeSat to help improve aerosol near cloud retrieval accuracy without impacting spatial coverage.