Trans-Pacific dust is thought to increase precipitation from atmospheric rivers (ARs) by enhancing ice formation in mixed-phase clouds. However, constraints on measurements of dust have limited our understanding of the magnitude of this effect.

In this dissertation, we utilize satellite-retrievals, ground-based observations, and reanalysis to: 1. Develop global, daily, observation-based datasets of dust aerosol optical depth extending from 2001 through 2018, 2. Develop a dust score to characterize the dust-content of the environment surrounding ARs that made landfall along the U.S. west coast between 2001 and 2018 and characterize the year-to-year variability of the dust content of the surroundings of ARs, and 3. Analyze the meteorological conditions associated with the presence of dust in the vicinity of ARs. We show that dusty ARs occur primarily in March, at the confluence of the end of the AR season and the beginning of the trans-Pacific dust season. We show that dust is preferentially found within the warm sector of the extra-tropical cyclones associated with ARs making landfall along the U.S. but dust is also enhanced in the cold sector. Our results suggest that the year-to-year variability in the dust score is primarily the result of variability of the dust content of the eastern North Pacific and correlated variability of the frequency of ARs and the dust content of surrounding area. Finally, we investigate the meteorological conditions that lead to the presence of dust in the vicinity of an AR by analyzing the conditions over trans-Pacific dust source regions and over the North Pacific in the days prior to an AR. Dusty ARs are associated with enhanced upper- and mid-level westerly winds over Asia, an extended North Pacific Jet, and eastward migrating extra-tropical cyclones. In contrast, pristine ARs are associated with a persistent ridge over the North Pacific.

The research presented here addresses gaps in our understanding of how often and why dust is present in the vicinity of an AR, a critical step in constraining the influence of dust on precipitation from ARs.

Dust over the Atlantic undergoes decadal variability. Over the past three decades, satellite and in-situ measurements show a downward trend in dust concentration over the Atlantic, indicating there is less dust emission in the Sahara. Meanwhile, observations show an upward trend in the Sahel precipitation. However, it is unknown what affects dust emission in the Sahara and precipitation in the Sahel simultaneously. Dust emission in North Africa also undergoes synoptic variability and high concentration of dust is often observed in the western Sahara, but it is unclear whether dust over the western Sahara is due to local emission or dust transport from the east.

The aim of this thesis is twofold, first, to examine the role the Saharan heat low (SHL) plays in dust emission and precipitation from synoptic to decadal timescales. and second, to identify bias of dust in the model. To do so I use reanalysis and satellite products to construct the time series of dust, precipitation and the SHL and perform regression analysis to examine the effect of the SHL on precipitation and winds, which are associated with dust emission. Then I study the lead and lag relationship between the SHL and dust concentration in North African on the synoptic scale to identify source regions. Lastly, I identify the biases in a dust source function using dust output from a climate model.

The results from this thesis suggest that the warming of the SHL results in the increase of precipitation in the Sahel and the reduction in dust emission by changing low-level winds on the decadal timescale. On the synoptic scale, cool phases of the SHL is associated with more dust emission and 10% of the dust over western Africa is from source regions to the east. Lastly, the dust source function may overestimate dust emission in the Sahel and western Africa.

Mineral aerosols (i.e. dust) perturb the Earth’s energy budget by scattering and absorbing shortwave (SW) and longwave (LW) radiation. Generally dust has a net (SW + LW) negative direct radiative effect (DRE) at the surface and top of the atmosphere (i.e. cooling of the Earth’s climate system). A few studies found that the net top of the atmosphere (TOA) DRE of dust, the difference between the net radiative flux in clear-sky (cloud-free) and pristine-sky (aerosol and cloud free) conditions, can be near zero or positive over desert regions. It is important to constrain the net DRE of dust over source regions since biases in this parameter can lead to errors in estimates of the global average. In Chapter 2 an observation-based method to estimate the dust SW forcing efficiency (FE), the DRE normalized by the aerosol optical depth, is developed and validated. Associated uncertainties in this method are further investigated. By accounting for the relationship between dust and water vapor, biases in the TOA SW FE of dust are reduced and estimates of the TOA SW F E of dust over the Sahara Desert range from −26.4±17.6 to 2.35±139 W m^-2 per optical depth. In Chapter 3 this method is applied to observations and retrievals from a small dust source region in southeastern California to estimate the surface and TOA dust SW FE. These results are compared with the dust SW FE estimated with the output from a radiative transfer model. After validating the model, modeled and observed estimates of the instantaneous SW FE are statistically similar and therefore the model is used to estimate the instantaneous dust LW FE and the diurnally averaged net dust FE. Values of the net FE of dust are −29±10, −10±11, and 19±15 W m^-2 per optical depth at the surface, TOA, and atmosphere, respectively, over the small dust source region. The findings presented in this dissertation contribute towards the breadth of observational estimates of the SW dust FE over dust source regions which is critical to constrain model-based estimates of regional and global values of the dust SW FE.