CICE5 해빙 시공간 분포 모의 평가 및 파라미터 민감도 테스트

Abstract

Using the CICE5 sea ice model with ERA5 atmospheric and ocean forcing, this study investigates the fidelity of modeling Arctic sea ice extent and volume and their sensitivity to key physical parameters. The model well simulated the overall seasonal cycle of sea ice extent and sea ice volume, but exhibits systematic biases such as an overestimation of summer sea ice extent and an underestimation of winter sea ice volume. These biases stem from the model’s overestimation of sea ice concentration in the marginal Arctic Ocean in summer and underestimation of sea ice thickness in the central Arctic Ocean in winter. A parameter sensitivity analysis is performed by changing values of physical parameters representing radiation, thermodynamical and dynamical processes. The result indicates that the radiative and thermodynamical parameters such as emissivity of snow and ice, ice density, snow density, and thermal conductivity of snow exhibits the higher sensitivity, with the first two parameters having the highest sensitivity. The dynamic parameters play little role. This information can be used to optimize parameter setting in sea ice model and further to enhance the capability of simulating polar climate in coupled model and Earth system model.