Summertime barotropic high pressure in the Arctic and its induced warmer and wetter atmosphere over sea ice are suggested to be important contributors to September sea ice loss on interannual and interdecadal time scales in the past decades. Using ERA5 and other reanalysis data, we find that atmospheric warming and moistening in the Arctic, synchronized by high latitude atmospheric circulation variability, work in tandem to melt sea ice through increasing downwelling longwave radiation at the surface. To what extent this atmosphere-longwave radiation-sea ice relationship can be captured in CMIP5 and 6 remains unknown and thus addressing this question is the objective of this study. To achieve this goal, we construct a process-oriented metric emphasizing the statistical relationship between atmospheric temperature and humidity with sea ice, which can effectively rank and differentiate the performance of 30 CMIP5 climate models in reproducing the observed connection. Based on our evaluation, we suggest that most available models in CMIP5 and 6 have a limitation in reproducing the full strength of the observed atmosphere–sea ice connection. This limitation likely stems from a weak impact of downwelling longwave radiation in linking sea ice with circulation associated with the weak sensitivity of the temperature and humidity fields to circulation variability in the Arctic. Thus, further efforts should be devoted to understanding the sources of these models’ limitations and improve skill in simulating the effects of atmospheric circulation in coupling temperature, humidity, surface radiation and sea ice together during Arctic summer.
ASJC Scopus subject areas
- Atmospheric Science