On the sources, composition, and climatic effects of mineral dust in the atmosphere

This dissertation is an exploration of several aspects of the global dust cycle. The first part is a comparison of climatological dust optical thickness over the Saharan region with analysis of the land surface topography. We find that high dust producing regions appear associated with extremely low-slope environments, and that they also appear to have roughness-to-slope ratios that are higher than typical. This indicates the importance of aeolian landforms such as sand dunes in contributing to dust emission, in addition to the more traditional view that dust arises mainly from closed topographic depressions. We extrapolate globally from these relationships, and suggest landscapes that could become dust sources in different climate or land-use regimes. The second part examines the optical properties of dust and shows an inversion calculation in which dust mineralogy is calculated based on wavelength dependent measurements of dust absorption and scattering. The results show significant geographic variation in the mineralogy of dust, and suggest that an external mixing model may be more accurate than internal mixing models in predicting dust optical properties as a function of mineralogy. The third part is a model experiment of the effects of dust on enhancing droughts in the US Great Plains that are initiated by sea surface temperature anomalies. The results indicate that dust does act to decrease precipitation and reduce soil moisture, thereby enhancing the drought. The experiment includes dust with three different single scattering albedo values, and the results show that the precipitation reduction is independent of the dust optical properties. Together, these chapters underscore the importance of the land surface in the global dust cycle, in terms of its effect on dust source areas, optical properties, and the relationship between the dust and hydrologic cycles.