Accounting for Aerosol Particle Non-sphericity in Passive and Active Remote Sensing of Desert Dust Using GRASP Algorithm

This study investigates the applicability of several non-spherical aerosol particles models including spheroids (Dubovik et al., 2006), superspheroids (Lin et al., 2018) and irregularly shaped hexaedrons (Saito et al., 2021) to the remote sensing of desert dust from real observations of ground-based advanced lidars, and a combination of ground-based sun-photometer with space-borne multi-angular polarimeter (GROSAT).

A highly versatile GRASP algorithm is used to retrieve detailed vertical distribution profiles of several aerosol components from the multi-spectral Raman depolarization or HSRL lidar (Lopatin et al., 2021), providing observations of volume depolarization, particle extinction and backscatter simultaneously at three wavelengths (Haarig et al., 2022) as well as concentration of several aerosol components combined with detailed BRDF and BPDF properties of the underlying surface from the multi-spectral multi-angular polarimetric observations performed over selected AERONET sites, whose observations are usually significantly influenced by the presence of desert dust particles. In both lidar and combined ground-based/satellite polarimetric retrievals GRASP is configured to treat aerosol as an external mixture of several aerosol components with their detailed microphysical properties including size and shape distributions and complex refractive index predefined, with one of the components representing coarse non-spherical "dust-like" particles (Lopatin et al., 2021). The latter could be represented by spheroidal, superspheroidal and irregular hexaedron particles all having same microphysical properties such as size distribution and complex refractive indices, as well as their external mixture with fractions of each particle type retrieved.

The results of retrievals of aerosol properties with different non-spherical components as well as their mixtures are inter-compared, validated and discussed.