MICROSTRUCTURES OF LOW AND MIDDLE-LEVEL CLOUDS OVER THE BEAUFORT-SEA

Airborne measurements in low and middle-level clouds over the Beaufort Sea in April 1992 and June 1995 show that these clouds often have low droplet concentrations (<100 cm(-3)) and relatively large effective d roplet radii. The highest average droplet concentrations overall were measured in altocumulus clouds that formed in airflows from the south that passed either over the North American continent or were from Asia . Droplet concentrations in low clouds tended to be higher in April th an in June. The low clouds in June occasionally contained drops as lar ge as 35 mu m diameter; in these clouds the collision-coalescence proc ess was active and produced regions of extensive drizzle. Cloud-top dr oplet concentrations were significantly correlated with aerosols benea th their bases, but appeared to be relatively unaffected by aerosols a bove their tops. Anthropogenic sources around Deadhorse, Alaska, incre ased local cloud droplet concentrations. Ice particle concentrations w ere generally low in April, but high ice particle concentrations were encountered in June when cloud-top temperatures were considerably high er. On two days in June, tens per litre of columnar and needle ice cry stals were measured in stratocumulus with top temperatures between -4 and -9 degrees C. Ice particle concentrations were poorly correlated w ith temperature (r = 0.39) but, for the two data sets as a whole, the concentrations of ice particles tended to increase with increasing tem perature from -30 to -4.5 degrees C. Ice particle concentrations corre lated better with the size of the largest droplets (r = 0.61). The mos t common mixed-phased cloud structure encountered was a cloud topped b y liquid water that precipitated ice. Liquid-water topped clouds were observed down to temperatures of -31 degrees C. They are likely common in the Arctic, and may play an important role in the radiation balanc e of the region. Temperature lapse rates in the clouds were generally complex, reflecting either layering, due to differential advection, or radiational effects. In these cases, vertical profiles of liquid-wate r content and effective cloud droplet radius did not vary systematical ly with height above cloud base, as they do in well-mixed clouds. Howe ver, when the temperature in a cloud decreased with height at or very near the pseudo-adiabatic lapse rate, the cloud liquid-water content ( and various measures of the broadness of the cloud droplet size distri bution) generally increased monotonically with height until very close to cloud top. For clouds consisting entirely of droplets, or droplets and ice crystals, cloud coverage increased by about 10% when the defi nition of a cloud was changed from 10 to 5 droplets per cubic centimet re. For clouds containing ice particles, cloud coverage and/or cloud d epth increased by about 40% when the definition of a cloud was changed from 1 to 0.1 ice particles per litre. The significance of these obse rvations with respect to the effects of clouds on the radiation budget of the Arctic, and the potential for modification of arctic clouds by pollution, are discussed.