Ice crystals in clouds frequently form upon a subset of aerosol particles called ice nuclei (IN). IN influence cloud ice crystal concentrations, consequently affecting cloud lifetime and reflectivity. The present understanding of these effects on climate is hindered by limited data on the global distribution of IN. This thesis presents measurements of deposition-mode IN concentrations under conditions relevant to mid-level clouds, 238 K and 138% RHi. at two Canadian sites: Toronto, a major city, and Whistler, a pristine coniferous rainforest. In Toronto, chemically-resolved surface areas were estimated by single-particle mass spectrometry and regressed against IN concentrations to identify a significant relationship between IN concentrations and both carbonaceous aerosols (EC and/or OC) and dust. In Whistler, IN concentrations during a biogenic secondary organic aerosol (SOA) event did not increase from background levels (0.1 /L), suggesting that biogenic SOA particles do not nucleate ice under these conditions.

Ice crystals in clouds frequently form upon a subset of aerosol particles called ice nuclei (IN). IN influence cloud ice crystal concentrations, consequently affecting cloud lifetime and reflectivity. The present understanding of these effects on climate is hindered by limited data on the global distribution of IN. This thesis presents measurements of deposition-mode IN concentrations under conditions relevant to mid-level clouds, 238 K and 138% RHi. at two Canadian sites: Toronto, a major city, and Whistler, a pristine coniferous rainforest. In Toronto, chemically-resolved surface areas were estimated by single-particle mass spectrometry and regressed against IN concentrations to identify a significant relationship between IN concentrations and both carbonaceous aerosols (EC and/or OC) and dust. In Whistler, IN concentrations during a biogenic secondary organic aerosol (SOA) event did not increase from background levels (0.1 /L), suggesting that biogenic SOA particles do not nucleate ice under these conditions.

Microorganisms are widespread in all ecosystems and play critical roles in nature. They are major players in global biogeochemical cycles that are fundamental in nutrient cycling. Molecular ecology surveys that investigate the microbial diversity of many different environments have revealed an impressive diversity of microbes in nature and have highlighted our inability to cultivate the vast majority of them in the laboratory. The improvement of our ability to grow uncultivable microbes in laboratory conditions will help us in this challenging task. Standard cultivation methods that have helped to bring to culture many relevant microorganisms in the past century are, however, characterized by limitations which hamper the isolation of novel microbes. For this reason, alternative cultivation strategies have been developed in recent decades which have allowed to expand the collection of environmentally relevant but poorly represented microbial strains. The use of such novel approaches for investigating the microbial diversity of underexplored natural ecosystems, such as sub-tropical mangrove forests, can result in the isolation, in laboratory conditions, of bacterial strains belonging to previously undescribed taxa. Mangroves are unique environments exposed to strong selection forces with respect to other marine environments, including high temperatures, salinity and oligotrophy. I hypothesize that these unique combinations of environmental features have selected microbiomes with unique characteristics. The aim of this PhD research is to explore the bacterial diversity of the Red Sea mangrove ecosystem, by applying an alternative cultivation strategy that uses oligotrophic conditions and long incubation time. I also exploited the diffusion chamber to cultivate bacterial taxa belonging to rarely isolated or even novel genera. This approach allowed me to isolate four novel bacterial taxa. Using 16S rRNA gene sequencing, the isolated bacteria were identified as one novel species and three novel genera belonging to Alpha-proteobacteria, Bacteroidetes, and Firmicutes, respectively. These isolates were further characterized and described through genomic, phylogenetic, chemotaxonomic, and phenotypic analysis to describe their ecological significance in the ecosystem of origin (i.e., mangrove sediments). This study reveals that the extreme conditions of the Red Sea mangroves have selected a unique and yet mostly untapped culturable microbiome with great potential for environmental applications.

Small water ice crystals are the main component of cold tropospheric clouds such as cirrus. Because these clouds cover large areas of our planet, their role in the radiation budget of incoming and outgoing radiation to the planet's surface is important. At present, the representation of these clouds in climate and weather models is subject to improvements: a large part of the uncertainty error stems from the lack of precise micro-physical and radiation model schemes for ice crystal clouds. To improve the cloud representations, a better understanding of the life time dynamics of the clouds and their composition is necessary, comprising a detailed understanding of the ice particle genesis, and development over their lifetime. It is especially important to understand how the development of ice crystals over time is linked to the changes in observable variables such as water vapour content and temperature and how they change the light scattering properties of the crystals. Recent remote and aircraft based in-situ measurements have shown that many ice particles show a light scattering behaviour typical for crystals having rough surfaces or being of complex geometrical shapes. The aim of this thesis was to develop the experimental setup and experiments to investigate this further by studying the surface morphology of small water ice crystals using scanning electron microscopy (SEM). The experiments I developed study the growth of water ice crystals inside an SEM chamber under controlled environmental conditions. The influence of water vapour supersaturation, pressure and temperature is investigated. I demonstrate how to retrieve the surface topology from observed crystals for use as input to computational light scattering codes to derive light scattering phase functions and asymmetry parameters, which can be used as input into atmospheric models. Difficulties with the method for studying the growth of water ice crystals, such as the effect of the electron beam-gas ionization and charging effects, the problem of facilitating repeated and localized ice growth, and the effect of radiative influences on the crystal growth are discussed. A broad set of nucleation target materials is studied. In a conclusion, I demonstrate that the method is suitable to study the surface morphologies, but is experimentally very challenging and many precautions must be taken, such as imaging only once and preventing radiative heat exchange between the chamber walls and the crystals to avoid unwanted effects on the crystal morphology. It is also left as a question if a laboratory experiment, where crystals will need to be grown in connection to a substrate, can represent the real world well enough. Deriving the required light scattering data in-situ might be an alternative, easier way to collect data for modelling use.