Modeling optical properties of nonspherical soil-derived dust particles for remote sensing applications

The quality of satellite aerosol retrievals depends critically upon the modeling accuracy of the aerosol optical properties such as the scattering phase function, single-scattering albedo and extinction coefficient. Those for mineral dust are among the most likely to warrant improvement. Mineral aerosols present a particularly difficult case for remote sensing because their optical properties depend strongly on dust particle morphology, mineralogy, and state of mixing. This thesis is a first comprehensive study addressing this problem. We developed an approach to link the data on dust particle morphology and optics modeling by integrating SEM/TEM data, the Discrete-Dipole Approximation technique, effective medium approximations, and radiative transfer codes. This strategy allow us to identify which properties of dust particle govern their optics and to determine the extent to which these properties are important to remote sensing. We demonstrate that the presence of angular-type particles of different thicknesses and circularities and the difference in the particle iron content result in various differences in optical properties of dust mixtures compared to those of volume-equivalent spheres or ellipses. The detailed analysis of the effect of dust particle nonsphericity and composition on the retrieval of aerosol optical properties from satellite and ground-based remote sensing observations are performed. To overcome the drawbacks of dust models used in current retrieval algorithms, we propose new models for the Saharan and Asian dust analogs for remote sensing applications at the solar wavelengths. The ultimate goal of this work is to enhance optical models of dust to aid in the interpretation of satellite observations and characterization of the radiative impacts of dust in the atmosphere.