A High Resolution Sea-Ice Model with Variable Snow Cover and Resolved Melt Ponds

Tuesday, 16 December 2014: 8:45 AM
Eric D Skyllingstad, Oregon State Univ, Corvallis, OR, United States, Karen M Shell, Oregon State University, Corvallis, OR, United States and Chris Polashenski, Organization Not Listed, Washington, DC, United States
A three-dimensional sea ice model is presented capable of resolving spatial snow depth variations and melt ponds over an O(100 m) area. The model calculates a surface heat budget and internal heating from solar radiative transfer, and simulates the formation and movement of brine/melt water through the ice/snow system. Initialization for the model is based on observations of snow topography made during the summer melt seasons of 2009, 2010, and 2012 from a location off the coast of Barrow, AK. Experiments are conducted that examine the melting period from the middle of May through June. Key processes simulated by the model are the flooding and drainage of melt water from snow that is commonly observed at the beginning of the melt season. Results indicate that resulting long-term melt pond coverage is sensitive to both the spatial variability of snow cover and the minimum snow depth. For thin snow cover, initial melting results in extensive flooding that limits pond formation after drainage of the melt water. Deeper snow tends to delay flooding of the ice surface, and can lead to more ponds after the surface drains. Simulated pond albedo tends to be too high in comparison with measurements, suggesting that the radiative transfer scheme is not properly calibrated for first year ice or that the role of sediments in the ice is underestimated. Nevertheless, predicted pond fraction time series are in good agreement with observations for each of the study years.