Abstract
We present observations from the 2nd Aerosol Characterisation Experiment where over a 29-h period between 16−18 July 1997 a tagged column of air was followed by a fully instrumented aircraft. The Lagrangian framework this offered made it possible to measure the evolution of the aerosol size distribution, the cloud structure and microphysics, and the dynamic and thermodynamic structure of the marine boundary layer within a polluted airmass advecting off northwest Europe over the sub-tropical North Atlantic Ocean. The salient observations are presented and analysed. Processes responsible for the evolution are suggested, but quantification of their respective rates must be taken up by future modelling studies. Stratocumulus capped the boundary layer throughout the period that produced negligible washout of aerosol. This implies that the conversion of a continental to a maritime airmass within the cloud-capped sub-tropical marine boundary layer is not controlled by the drizzle process but by entrainment from the free troposphere. We find evidence of processing of aerosol particles by stratocumulus cloud, in particular by aqueous-phase reactions. The processing of the aerosol, realised by modification of the aerosol size distribution in the particle diameter range 0.1−0.5 μm, was complicated by rapid changes in boundary layer height and structure, and also by entrainment of both polluted and relatively clean aerosol from the free troposphere. The cloud microphysics was affected by these changes in the boundary layer aerosol through changes in the cloud condensation nuclei activation spectra. The cloud microphysics was also strongly affected by changes in the dynamics of the boundary layer which included variations (e.g., diurnal) in cloud thickness and an increase in vertical wind speed. Thermodynamic changes within the boundary layer included decoupling due to an increasing sea-surface temperature and a change in the subsidence rate in the free troposphere superimposed on diurnal decoupling. Hypotheses have been devised so that future modellers can focus their efforts to either validate or invalidate potentially important processes.