Effects of Wintertime Snow Depth Assimilation on Summertime Sea Ice Prediction Experiments Utilizing a Fully Coupled Regional Arctic Model With Perfect Boundary Conditions

Abstract

During the past decades, the rapid decline of sea ice in the Arctic Ocean has substantially favored commercial navigation activities in the summertime. Numerical sea ice prediction on a seasonal scale plays a crucial role in guiding the programme of such activities, yet the role of snow depth data assimilation in numerical sea ice prediction has not been clearly addressed in previous studies. With the aid of a coupled Arctic sea ice‒ocean‒atmosphere modeling system, two sets of runs have been conducted: one assimilates sea ice concentration, sea ice thickness, and sea surface temperature in the ice‐free region; the other one also assimilates snow depth overlying the ice synchronously. We have found that driven by perfect boundary conditions, wintertime snow depth assimilation generally has a positive effect on the September sea ice prediction initialized in early June, and the positive effect gradually weakens along with the delay of prediction onset from early June to early August. The effect has an intimate relationship with the changes in sea ice thickness and snow depth in the model state at the prediction onset, which eventually affects September sea ice prediction in the coupled system via model physics.