| A coupled ice-ocean model of the Arctic Ocean and adjacent seas is developed in order to study the effects of anomalies of precipitation and river runoff on sea ice. A dynamic-thermodynamic sea ice model is coupled to an ocean general circulation model which includes a turbulence closure scheme for the treatment of stability-dependent vertical mixing of temperature, salinity, and momentum. The ice and ocean models are coupled by the fluxes of heat, momentum, and salinity computed from the predicted ice and ocean temperatures, velocities, and net ice growth. The model is forced by interannually-varying atmospheric temperature and pressure data from 1980-1989, parameterized incoming radiation, and hydrologic forcing from precipitation and river runoffs. The modeled ice thickness, ice extent, and ice drift are generally similar to observed patterns, with a bias toward thinner ice and some regional differences in extent, which reduces the export of ice out of the Arctic. The ice extent is shown to be sensitive to the ocean heat flux computed in the model. Ice thickness and sea surface salinity are shown to be very sensitive to how the vertical mixing scheme in the ocean maintains the strong stratification. Sensitivity experiments show that without any precipitation input, sea ice thickness decreases rapidly because of the destabilization of the upper ocean. Eliminating the river runoffs causes a much slower decreasing trend in ice thickness, due to the decadal timescale of the dispersal of runoff in the Arctic Ocean. Interannual precipitation anomalies are shown to create anomalies of ice thickness that last 1-2 years, while runoff anomalies do not produce significant ice anomalies. The ice-ocean model results suggest that changes in arctic precipitation can affect sea ice more rapidly and more significantly than river runoffs, although rivers contribute the greater volume of fresh water to the Arctic, and that river runoffs can affect the sea ice on the timescale of a decade. |
