Authors:王松*，苏洁Abstract: Based on the trend analysis and correlation analysis of sea ice concentration provided by the National Snow and Ice Data Center and AD index extracted from the ERA-Interim sea-level pressure and monthly mean sea level pressure data, it is concluded that the average AD index from December to February is positively correlated with the range of Arctic sea ice in February, and the average AD index from May to August is negatively correlated with the range of Arctic sea ice in September. In order to confirm the influence of thermal and dynamic forcing on sea ice concentration under the influence of AD atmospheric circulation in winter and summer, series of simulation experiment were carried out using CICE sea ice model. The study shows that the model can reproduce the main characteristics of sea ice distribution under positive and negative AD phases, and the simulation ability of arctic sea ice meets the research needs. In winter, the Arctic sea ice concentration is not sensitive to the change of AD type wind field. Under the forcing of negative phase AD, the sea ice concentration in the marginal sea of the Atlantic sector decreases through thermal effector forcing beyond wind drive. In summer, under the influence of positive phase AD, sea ice concentration in the sea area outside east Greenland decreases, which is mainly caused by the dynamic action of wind field, thermodynamic effects also contributed to the smaller of sea ice concentration in the marginal and Barents Seas of the Atlantic sector.