UCLA

Desert dust is the dominant aerosol type by mass in the atmosphere. Dust impacts on various aspects of the Earth system depend sensitively on its size and shape. However, global aerosol models and remote sensing retrievals struggle to correctly account for dust shape and size for a few reasons. These issues include that (i) models and retrieval algorithms lack a consistent and accurate quantification of dust shape, (ii) models and retrievals substantially underestimate the abundance of coarse dust in the atmosphere relative to measurements, (iii) measurements of size distributions are also problematic because they are based on different diameter types that do not account for realistic dust shapes, and (iv) measurements of emitted dust size distributions are not available for some major soil types (e.g., active sands), making the validation process for these regions difficult. The resulting biases in dust shape and size can propagate into dust optical properties used in models and retrieval algorithms, which further contributes to inaccurate estimates of dust effects on the Earth system.This dissertation addresses the issues above. Specifically, I present the first (to my knowledge) in situ field measurements of the size distributions of dust aerosols emitted from active sands. I show that active sands emit substantially finer dust than non-sandy soils. Second, I compile dozens of in situ observations of dust shape across the globe and obtain a globally representative constraint on the probability distributions describing dust shape. I show that models and retrieval algorithms substantially underestimate dust asphericity by a factor of ~3 to 5; as aspherical dust deposits less quickly from the atmosphere, this underestimated dust asphericity causes models to underestimate dust lifetime by ~20%. Third, I use this shape constraint to correct a compilation of measurements of emitted dust size distributions that neglected this substantial dust asphericity. I find that accounting for asphericity yields a substantially coarser emitted dust size distribution and that, consequently, current parameterizations underestimate coarse dust emission by more than a factor of ~2. Finally, I account for the observational constraints on dust shape and size in obtaining the single-scattering properties of dust aerosols. This newly-developed dataset is extensively resolved by a wide range of wavelength, dust size, and dust refractive index values, which enables wide applications on models and remote sensing products. These findings have several key implications. First, active sands emit substantially finer dust, which could enhance its downwind impacts on human health, the hydrological cycle, and regional climate. Second, the findings that models and retrieval algorithms underestimate both dust asphericity and coarse dust emission help explain why models underestimate the abundance of coarse dust in the atmosphere. Third, the results highlight the importance of standardized diameter conversions. A lack of such standardization can generate substantial biases, for instance in the measurements of dust size distributions. Finally, the extensive single-scattering properties accounting for realistic dust shape and size are being implemented into several global aerosol models, including MONARCH, IMPACT, NCAR CESM, and NASA GISS ModelE. This work could ultimately help narrow the large uncertainties in dust impacts on the Earth system.