资源环境科技发展态势分析平台

Clouds are an important component of the Arctic climate system through their regulation of the surface energy budget; however, Arctic clouds are poorly simulated in global climate models (GCMs). In this study, we evaluate the Arctic clouds simulated by a multiscale modeling framework (MMF). The results are compared against a merged CloudSat-CALIPSO radar-lidar cloud product and contrasted with an atmospheric reanalysis and conventional GCMs. The comparisons focus on the annual cycle of cloud covers, vertical structures of cloud fraction, and condensate mixing ratio, as well as the relationships between low-cloud cover and atmospheric static stability. The MMF is found to represent Arctic boundary layer clouds slightly more realistically than the reanalysis and GCMs in both the annual cycle and vertical distribution except that middle- and high-cloud covers are underestimated and the amplitude of annual cycle of total cloud cover is larger. The relationship between low-cloud cover and near-surface atmospheric stability produced by MMF is remarkably similar to the satellite observation during autumn, winter, and early spring, as low-cloud cover decreases with colder surface and stronger stability. Such relationships over the annual cycle are not reproduced by other modeling approaches. Lastly, MMF yields a positive correlation between low-cloud cover and atmospheric stability over the Arctic ocean from May to August, opposite to the satellite observation, implying stronger control of horizontal advection on low-cloud formation. This modeled relationship is contributed by cloud fraction near the surface, which is known to be underestimated due to radar's surface clutter.