A Numerical Simulation of Sea Ice Cover in Hudson Bay

View More View Less
  • 1 Department of atmospheric and Oceanic Sciences and Centre for Climate and Global Change Research, McGill University, Montreal, Quebec, Canada
© Get Permissions
Full access

Abstract

Hibler's dynamic-thermodynamic sea ice model with viscous-plastic rheology is used to simulate the seasonal cycle of sea ice motion, thickness, compactness, and growth rate in Hudson Bay under monthly climatological atmospheric forcing and a prescribed ocean surface current field. The sea ice motion over most of the domain is driven mainly by the wind stress. Wintertime sea ice velocities are only of the order of 1–5 (× 10−4 m s−1) due to the nearly solid ice cover and the closed boundary constraint of Hudson Bay. However, the velocities rise to 0.10–0.20 m s−1 during the melting and freezing seasons when there is partial ice cover. The simulated thickness distribution in mid–April, the time of heaviest ice cover, ranges from 1.3 m in James Bay to 1.7 m in the northern part of Hudson Bay, which compares favorably with observations. The area-averaged growth rate, computed from the model is 1.5–0.5 cm day−1 from December to March, is negative in May (indicative of melting) and reaches its minimum value of −4.2 cm day−1 (maximum melting rate) in July. During autumn, the main freezing season, the growth rate ranges from 1 to 2 cm day−1. In the model, sea ice remains along the south shore of Hudson Bay in summer, as observed, even though the surface air temperatures are higher there than in central and northern Hudson Bay. A sensitivity experiment shows that this is mainly due to the pile-up of ice driven southward by the northwesterly winds. The simulated results for ice cover in other seasons also compare favorably with the observed climatology and with measurements from satellites. In particular, the model gives complete sea ice cover in winter and ice-free conditions in late summer. A series of sensitivity experiments in which the model parameters and external forcing are varied is also carried out.

Abstract

Hibler's dynamic-thermodynamic sea ice model with viscous-plastic rheology is used to simulate the seasonal cycle of sea ice motion, thickness, compactness, and growth rate in Hudson Bay under monthly climatological atmospheric forcing and a prescribed ocean surface current field. The sea ice motion over most of the domain is driven mainly by the wind stress. Wintertime sea ice velocities are only of the order of 1–5 (× 10−4 m s−1) due to the nearly solid ice cover and the closed boundary constraint of Hudson Bay. However, the velocities rise to 0.10–0.20 m s−1 during the melting and freezing seasons when there is partial ice cover. The simulated thickness distribution in mid–April, the time of heaviest ice cover, ranges from 1.3 m in James Bay to 1.7 m in the northern part of Hudson Bay, which compares favorably with observations. The area-averaged growth rate, computed from the model is 1.5–0.5 cm day−1 from December to March, is negative in May (indicative of melting) and reaches its minimum value of −4.2 cm day−1 (maximum melting rate) in July. During autumn, the main freezing season, the growth rate ranges from 1 to 2 cm day−1. In the model, sea ice remains along the south shore of Hudson Bay in summer, as observed, even though the surface air temperatures are higher there than in central and northern Hudson Bay. A sensitivity experiment shows that this is mainly due to the pile-up of ice driven southward by the northwesterly winds. The simulated results for ice cover in other seasons also compare favorably with the observed climatology and with measurements from satellites. In particular, the model gives complete sea ice cover in winter and ice-free conditions in late summer. A series of sensitivity experiments in which the model parameters and external forcing are varied is also carried out.

Save