Measurements of aerosol size distributions made with a differential mobility particle sizer (DMPS) during three expeditions to the central Arctic Ocean in summer showed a consistent source of both Aitken-mode and accumulation-mode particles within the Arctic pack ice. Above the shallow atmospheric mixed layer in the Artic the stability of the atmosphere was such that transport from the free troposphere was limited to rare occasions, indicating a surface source. Profiles of aerosol concentration made with a helicopter suggested that there were increases within regions where low clouds had evaporated. A similar result was observed close to the surface, both Aitken-mode and accumulation-mode concentrations increasing during and immediately after fog events.

Transmission electron microscope (TEM) studies of particles sampled in the surface film of the open water between ice floes and in the air showed many common features, suggesting that airborne particles resulted from the bursting of bubbles that had scavenged particulates from the water. Accompanying the particles in both air and water was a gel-like substance, the properties of which identified it as having formed from the exopolymer secretions (EPS) of resident ice algae and bacteria. The most numerous particle type common to water and air was an aggregate of many particles, each predominantly smaller than 50nm diameter and apparently bonded with EPS. Ultraviolet light and acidification chemically alter EPS, and comparison of the size distribution of airborne and water aggregates suggest that this process allowed the atmospheric aggregates to fragment into progressively smaller sizes. The strongly surface-active, strongly hydrated EPS coatings on aggregates should make them active in cloud drop formation. Acidification through aqueous oxidation of sulfur gases in cloud drops may speed the fragmentation of aggregates, causing an increase in particle numbers when the cloud evaporates.

Individual components of the aggregates were too small to be captured efficiently with TEM collectors, but the fact that they were the only particle types observed that were smaller than 50nm diameter suggests that they formed the bulk of such size of particles seen by the DMPS. The close agreement between the size distributions obtained from TEM photographs of aggregates and their individual components and that of the DMPS between 15 and 70nm (Aitken mode) supported this suggestion. Such particles would be preferred sites for deposition of oxidation products of DMS, leading to formation of sulfur-containing CCN, or, if enough of the highly surface-active highly hydrated EPS gel remained attached, would themselves act as CCN. There is therefore a stronger possible link between marine biology, cloud properties and climate than is provided by DMS alone.