Modeling turbulent heat fluxes over Arctic sea ice using a maximum- entropy-production approach

Recently, an algorithm of surface turbulent heat fluxes over snow/sea ice has been developed based on the theory of maximum entropy production (MEP), which is fundamentally different from the bulk flux algorithm (BF) that has been used in sea ice models for a few decades. In this study, we first assess how well the MEP algorithm captures the observed variations of turbulent heat fluxes over Arctic sea ice. It is found that the calculated heat fluxes by the MEP method are in good agreement with in situ observations after considering the absorption of incoming radiation in a snow/ice surface layer with infinitesimal depth. We then investigate the effects of two different schemes (MEP vs. BF) in the sea ice model of CICE6 on simulated turbulent heat fluxes and sea ice processes in the Arctic Basin. Our results show that the two different schemes give quite different representations of seasonal variations of heat fluxes, particularly for sensible heat fluxes in summer. The heat fluxes simulated by the MEP produce weak cooling effect on the ice surface in summer, whereas the BF generates a warming effect. As a result, compared to the BF, the MEP leads to a reduced seasonal cycle of Arctic sea ice mass flux by modulating snow-to-ice conversion, basal ice growth, surface ice melt and basal ice melt.