This work tests the validity of using a commercial cloud condensation nuclei (CCN) counter (CCNc) on the Facility for Airborne Atmospheric Measurements (FAAM) research aircraft. The CCNc was suitable for aircraft work with sta- ble and repeatable supersaturation, temperature and pressure relationships. The sample architecture of the aircraft fitted CCNc was found to transmit particles with acceptable losses in the diameter range of interest as was a pressure control device designed for airborne work. Rosemount inlets, used to sample aerosol, were found to be sensitive to particle density resulting in disparate aerosol being sam- pled with different efficiencies. In dust dominated aerosol inlet efficiency peaks at 10.24 at an optical diameter of 2.91 μm, with a minimum inlet efficiency between 1.78 and 1.51 at 0.28μm. In less dense aerosol inlets sample representatively below 0.6 μm and comparably below 1.0 μm. The thorough testing of the CCNc, associated sampling architecture and mea- surement strategies, enabled vertical and horizontal CCN to be investigated along with other aerosol and cloud microphysical properties in the Southern Equato- rial Pacific (SEP). The primary source of particulates was the South American continent, with sulphate dominating composition. There were strong gradients in aerosol and gas phase chemistry concentration with distance from the coast and in the cloud microphysics measurements where highest droplet numbers and smallest diameters were close to the coast. These data represent an important validatory and parameterisation data set for models of all scales. CCN data were used to calculate the aerosol hygroscopicity parameter, the mean project value, κ, was 0.21 ± 0.18 . There was no evident variation in hygroscopicity with distance from the Chilean coastline suggesting a single dominant source and a well mixed boundary layer up to 907km to the west. CCN measurements were also com- pared to predictions from multiple models of different composition and mixing state assumptions. The best CCN closure used an external mixture of inorganic and organic aerosol components, with a modelled to observed ratio of 1.37 ± 0.32. It was hypothesised that this large ratio and the relatively low bulk hy- groscopicity was influenced by an external mixture. Incorporating this external mixture is imperative if CCN are to be accurately modelled and any subsequent cloud processes accurately captured.

Aerosols have significant impacts on earth's climate and hydrological cycle. They can directly reflect the amount of incoming solar radiation into space; by acting as cloud condensation nuclei (CCN), they can indirectly impact climate by affecting cloud albedo. Our current assessment of the interactions of aerosols and clouds is uncertain and parameters used to estimate cloud droplet formation in global climate models are not well constrained. Organic aerosols attribute much of the uncertainty in these estimates and are known to affect the ability of aerosol to form cloud droplets (CCN Activity) by i) providing solute, thus reducing the equilibrium water vapor pressure of the droplet and ii) acting as surfactants capable of depressing surface tension, and potentially, growth kinetics. My thesis dissertation investigates various organic aerosol species (e.g., marine, urban, biomass burning, Humic-like Substances). An emphasis is placed on the water soluble components and secondary organic aerosols (SOA). In addition the sampled organic aerosols are acquired via different media; directly from in-situ ambient studies (TEXAQS 2006) environmental chamber experiments, regenerated from filters, and cloud water samples. Novel experimental methods and analyses to determine surface tension, molar volumes, and droplet growth rates are presented from nominal volumes of sample. These key parameters for cloud droplet formation incorporated into climate models will constrain aerosol-cloud interactions and provide a more accurate assessment for climate prediction.

It is well known that atmospheric aerosols provide the sites for forming cloud droplets, and can affect the Earth's radiation budget through their interactions with clouds. The ability of aerosols to act as cloud condensation nuclei is a strong function of their chemical composition and size. The compositional complexity of aerosol prohibits their explicit treatment in atmospheric models of aerosol-cloud interactions. Nevertheless, the cumulative impact of organics on CCN activity is still required, as carbonaceous material can constitute up to 90% of the total aerosol, 10-70% of which is water soluble. Therefore it is necessary to characterize the water soluble organic carbon fraction by CCN activation, droplet growth kinetics, and surface tension measurements. In this thesis, we investigate the water soluble properties, such as surface tension, solubility, and molecular weight, of laboratory and ambient aerosols and their effect on CCN formation. A mechanism called Curvature Enhanced Solubility is proposed and shown to explain the apparent increased solubility of organics. A new method, called Köhler Theory Analysis, which is completely new, fast, and uses minimal amount of sample was developed to infer the molar volume (or molar mass) of organics. Due to the success of the technique in predicting the molar volume of laboratory samples, it was applied to aerosols collected in Mexico City. Additionally the surface tension, CCN activity, and droplet growth kinetics of these urban polluted aerosols were investigated. Studies performed for the water soluble components showed that the aerosols in Mexico City have surfactants present, can readily become CCN, and have growth similar to ammonium sulfate. Finally, aerosols from three different polluted sources, urban, bovine, and ship emissions, were collected and characterized. The data assembled was used to predict CCN concentrations and access our understanding of the system. From these analyses, it was evident that knowledge of the chemical composition and mixing state of the aerosol is necessary to achieve agreement between observations and predictions. The data obtained in this thesis can be introduced and used as constraints in aerosol-cloud interaction parameterizations developed for global climate models, which could lead to improvements in the indirect effect of aerosols.

It has been historically assumed that most of the uncertainty associated with the aerosol indirect effect on climate can be attributed to the unpredictability of updrafts. We assess the sensitivity of cloud droplet number density to realistic variations in aerosol chemical properties and to variable updraft velocities using a 1-dimensional cloud parcel model. The results suggest that aerosol chemical variability may be as important to the aerosol indirect effect as the effect of unresolved cloud dynamics, especially in polluted environments. We next used a continuous flow streamwise thermal gradient Cloud Condesnation Nuclei counter (CCNc) to study the water-uptake properties of the ambient aerosol, by exposing an aerosol sample to a controlled water vapor supersaturation and counting the resulting number of droplets. The heat transfer properties and droplet growth within the CCNc were first modeled and experimentally characterized. We describe results from the MIRAGE field campaign at a ground-based site during March, 2006. Size-resolved CCN activation spectra and hygroscopic growth factor distributions of the ambient aerosol in Mexico City were obtained, and an analytical technique was developed to quantify a probability distribution of solute volume fractions for the CCN, as well as the aerosol mixing-state. The CCN were shown to be much less CCN active than ammonium sulfate, with water uptake properties more consistent with low molecular weight organic compounds. We also describe results from the GoMACCS field study, an airborne field campaign in Houston, Texas during August-September, 2006. GoMACCS tested our ability to predict CCN for highly polluted conditions with limited chemical information. Assuming the particles were composed purely of ammonium sulfate, CCN closure was obtained with a 10% overprediction bias on average for CCN concentrations ranging from less than 100 cm-3 to over 10,000 cm-3, but with on average 50% variability. Assuming measured concentrations of organics to be internally mixed and insoluble tended to reduce the overprediction bias for less polluted conditions, but led to underprediction bias in the most polluted conditions. Comparing the two campaigns, it is clear that the chemistry of the particles plays an important role in our ability to predict CCN concentrations.