The Arctic climate is changing rapidly, warming at about twice the rate of the planet. Global climate models (GCMs) are invaluable tools both for understanding the drivers of these changes and predicting future Arctic climate evolution. While GCMs are continually improving, there remain difficulties in representing cloud processes which occur on scales smaller than GCM resolution. Since clouds influence the Arctic energy and water cycles, their accurate representation in models is critical for robust future projections. In this work, we examine the representation of Arctic clouds and precipitation in the Community Earth System Model (CESM) with the Community Atmosphere Model (CAM), comparing the newly released version (CESM2 with CAM6) with its predecessor (CESM1 with CAM5). To isolate changes in the Arctic mean state, we compare preindustrial control runs. Arctic cloud ice has decreased slightly, while cloud water has increased dramatically in CESM2. Annual mean liquid-containing cloud (LCC) frequency has increased from 19% in CESM1 to 51% in CESM2. Since LCCs strongly modulate downwelling radiation at the surface, their increase has led to an increase in mean downwelling longwave (+22 W m-2) and corresponding decrease in downwelling shortwave (-23 W m-2) radiation. The mean Arctic surface temperature increased from 257 K in CESM1 to 260 K in CESM2, with the largest seasonal difference in winter (+6 K). Annual average snowfall has decreased slightly (-1 mm month-1), while rainfall has increased (+5 mm month-1).
Keywords: Arctic; CAM6; CESM2; clouds; precipitation.
©2020. The Authors.