Publication Date:
2018-02-18
Description:
An optimization algorithm is developed to retrieve liquid water cloud properties including cloud optical depth (COD), droplet size distribution and cloud top height (CTH), and aerosol above cloud properties including aerosol optical depth (AOD), single scattering albedo and microphysical properties from sweep-mode observations by JPL's AirMSPI instrument. The retrieval is composed of three major steps: (1) retrieval of an initial estimate of the mean droplet size distribution across the entire image of 80-100 km along-track by 10-25 km across-track from polarimetric cloudbow observations; (2) coupled retrieval of image-scale cloud and above-cloud aerosol properties by fitting the polarimetric data at all observation angles; and (3) iterative retrieval of 1D-RT based COD and droplet size distribution at pixel-scale (25 m) by establishing relationships between COD and droplet size and fitting the total radiance measurements. Our retrieval is tested using 134 AirMSPI datasets acquired during the NASA ORACLES field campaign. The retrieved above-cloud AOD and CTH are compared to coincident HSRL-2 (NASA LaRC) data, and COD and droplet size distribution parameters (effective radius r eff and effective variance v eff ) are compared to coincident RSP (NASA GISS) data. Mean absolute differences (MADs) between AirMSPI and HSRL-2 retrievals of above-cloud AOD at 532 nm and CTH are 0.03 and 〈 0.5 km, respectively. At RSP's footprint scale (323 m), MADs between RSP and AirMSPI retrievals of COD, r eff and v eff in the cloudbow area are 2.33, 0.69 μm and 0.020, respectively. Neglect of smoke aerosols above cloud leads to an underestimate of image-averaged COD by ~15%.
Print ISSN:
0148-0227
Topics:
Geosciences
,
Physics
Permalink