Modelling the effects of organic aerosol phase partitioning processes on cloud formation

Atmospheric aerosols particles may act as cloud condensation nuclei (CCN) that provide sites for condensation of water vapour for the formation of cloud droplets, called cloud droplet activation. Whether aerosol particles are CCN is determined by their size, composition and the ambient humidity. Cloud macrophysical properties together with the size and number concentration of droplets determine the optical properties of liquid phase clouds. Clouds are an important component in the Earth's radiation balance and aerosol-cloud interactions (ACI) are associated with the largest uncertainty in estimates made of anthropogenic radiative forcing in earth system models. To constrain ACI and reduce uncertainties, an improvement in our understanding of CCN activation is required. Owing to its complex phase structure and chemical heterogeneity, the organic fraction of atmospheric aerosol introduces significant challenges in developing an exact description of cloud formation. In this thesis, a cloud parcel model is employed to systematically address parametric and process uncertainties in estimates of cloud droplet sizes and number concentrations (CDNC). To do so, the unified framework for organic aerosol (UFO) scheme was developed and embedded into the cloud parcel model, ICPM-UFO. The ICPM-UFO simulates partitioning of organic mass between the gas and aqueous bulk and surface phases, thereby providing means to theoretically diagnose changes in droplet nucleating potential of aerosol particles due to organic aerosol mass transfer processes. Partitioning of surface active organic aerosol mass from the bulk particle phase to the surface phase results in a lowered, size-dependent surface tension that enhances activation potential of CCN and therefore simulated CDNC. A large fraction of organic aerosol constituents exist partitioned across particle and gas phases and simulation of cloud formation events show this semi-volatile organic mass to condense to the particle phase as humidity increases through the cloud base. This additional particle phase mass may be partially soluble. The more soluble component increases the activation potential by lowering the water activity, while the less soluble but more surface active component also increases the activation potential by further lowering of the surface tension. The compounding effects of the gas-particle and bulk-surface partitioning processes result in significant changes in CCN concentrations and CDNC for simulation on boreal aerosol. These results exhibit a significant over prediction of typical boreal CCN concentrations relative to in-situ measurements, though further sensitivity analysis with respect to the soluble fraction and surface phase description may be advantageous. Based on multivariate statistical approaches applied, resolution of the surface phase in cloud formation parameterisations within climate models is however not currently recommended. Theoretical description of both partitioning processes require prescription of input parameters that are challenging to measure in-situ. These parameters include: SVOC volatility and enthalpy of vaporisation and organic component surface tension and film thickness. Further work using the inverse modelling framework established herein is recommended to provide estimation of these parameters while simultaneously matching simulated CDNC and/or CCN concentrations with observational data. It is envisaged that such an investigation will also yield insights into structural uncertainties associated with the choice of surface phase model - a point of contention both within this thesis and the wider literature. At the time of the doctoral defense, the following paper was unpublished and had a status as follows: Paper 4: Manuscript.