Beesley, J. A. and R. E. Moritz, 1997: Toward an explanation of the annual cycle of cloudiness over the Arctic ocean, Journal of Climate (submitted).

The annual cycle of low cloud amount over the Arctic Ocean is studied using climatological data and a time-dependent atmospheric column model. Three hypotheses for the annual cycle are formulated, compared with climatological data for consistency, and tested using the model. The hypotheses distinguish three factors that might account for the annual cycle of cloud amount over the Arctic Ocean: (1) the annual cycle of water vapor advection into the Arctic, (2) the annual cycle of evaporation at the surface of the ice pack, and (3) the effect of the annual cycle of temperature on the formation and precipitation of atmospheric ice. None of the hypotheses can be ruled out by inspection of available climatological data, although the transition between the winter and summer cloudiness regimes appears to occur one month before the influx of atmospheric moisture increases from its wintertime level, which weakens the first hypothesis.

The atmospheric column model represents physical processes necessary to simulate low stratiform clouds observed at lower latitudes, plus ice phase physical processes that are likely to be important in the Arctic. The prescribed boundary conditions consist of incoming solar radiation at the top of the atmosphere; vertical profiles of divergence, temperature advection, and moisture advection; and open water fraction and ice thickness at the surface. When forced by standard boundary conditions corresponding to winter and summer, the model simulates the qualitative regimes of the respective seasons: large mean cloud amount in summer and small mean cloud amount in winter. Two sensitivity experiments are performed in which the standard winter experiment is re-run with summertime values of moisture advection, and intensity of ice-phase physical processes, respectively. Only the suppression of ice-phase physical processes produces a significant increase in the simulated amount of winter low cloudiness. Three sensitivity experiments are performed in which the standard summer experiment is re-run with wintertime values of moisture advection, evaporation and ice-phase physical processes. Only the enhancement of ice-phase processes produces a significant reduction in the simulated summer cloudiness. These results suggest that the temperature-dependence of ice phase precipitation processes is an essential element of the explanation for the annual cycle of low cloudiness. These processes limit the duration of clouds in winter by converting atmospheric moisture from liquid cloud particles to larger, rapidly-precipitating ice particles, and by preventing the mixing ratio in clear air from reaching saturation with respect to liquid.

It is shown that, among general circulation models participating in the Atmospheric Model Intercomparison Project, those that represent ice-phase precipitation processes produce qualitatively more realistic simulations of the mean annual cycle of cloud amount over the Arctic Ocean than models that omit these processes. It is predicted that under a general warming of Arctic climate, the cloudy summer season would lengthen and vice versa under a general cooling.