Thermodynamic processes in arctic ice are mainly considered in numerical models of Maykut et al (1971), Semtner (l976), Parkinson et al (1979), etc, to describe important thermodynamic factors. Meanwhile, great progress of arctic ice models has been made in improving the dynamic description of arctic sea ice behavior by Campbell (1965), Coon (1980), Hibler (1979), etc. Based on the studies mentioned above, Hibler (1980), Walsh (1985), Semtner (1987) and Oberhuber (1993) developed dynamic - thermodynamic models and coupled 0cean-ice models for realistic simulation considering dynamic and thermodynamic processes of arctic sea ice. Up to now, the dynamic - thermodynamic ice model can reasonably simulate the drifting pattern, ice cover extent and thickness distribution in arctic ocean. But it is still beyond the ultimate goal to couple these models with atmospheric and oceanic models.

The further improvement on arctic ice model can be expected in the following research directions. (1) The ice model’s parameterization schemes for thermodynamic processes are not enough to model these processes in marginal ice zone as well as between water and sea leads, but we can expect to get further improvement. (2) Developing variable thickness ice models to study the dynamic and thermodynamic effects of varying thickness of the ice cover. (3) Developing more reasonable rheology to improve the modeling of the internal ice stresses. (4)The coupled ocean-ice model can improve the results of calculation on the location of ice edge and other ice cover properties such as mean ice thickness and concentration. Development of air-ice-ocean coupling model will be an important future direction in studying the role of the Arctic in the global climate.