Sensitivity of the Modeled Arctic Sea-Ice Extent and Thickness to Changes in Various Parametrizations

Abstract

Seven experiments are conducted on a coupled ice-ocean model in order to gain insight in the sensitivity of the modeled Arctic sea-ice extent and thickness to crucial changes in the parametrization of the models albedo scheme and thermodynamics. In addition to a control experiment there are three sensitivity experiments which change how the snow/ice-albedo is parametrized. The first replaced the model's albedo scheme with a more basic one, while the remaing two made changes to the existing albedo scheme. Two experiments are conducted which altered the lead closing parameter, ho, affecting the thickness of ice formed in open water. The final sensitivity experiment added an increament of maximum one degree Celsius to the 2 m atmospheric temperature. The ice thickness and extent is most sensitive to the experiment which replaced the model's albedo scheme with a basic one. This led to an increase from the control run in the sea ice extent and thickness of about 22%. Increasing the 2 m atmospheric temperature led to a decrease of 60% in the mean snow thickness. Thus the snow thickness is sensitive to the parametrizations of sensible heat and longwave radiation. The seasonal cycle of the sea ice extent was sensitive to the changes made to the lead closing parameter ho. This parameter can be used to tune the modeled cycle of the sea ice extent to coincide better with the observed sea ice extent cycle.