There are many indications that the Arctic sea ice cover is undergoing significant climate-induced changes, affecting both its extent and thickness [e.g., Parkinson and Cavalieri, 2002; Parkinson et al., 1999; Comiso, 2002; Stroeve et al., 2005; Stroeve, pers. comm.; Tucker et al., 2001; Wadhams and Davis, 2000; Rothrock et al., 1999]. However, the details of the complex atmosphere-ice-ocean interaction are not well understood. The mass balance of the ice cover is of particular interest as a key climate change indicator, since it is an integrator of both the surface heat budget and the ocean heat flux whereby net warming over time causes thinning while net cooling leads to thicker ice.

The mass balance of sea ice can be thought of as the evolution of thickness distribution. It is controlled by thermodynamic ice growth and melt, mechanical redistribution through ridging and rafting, and transport. For simplicity, we consider a regional Lagrangian frame of reference, and track the evolution of a region of ice, eliminating the need to consider transport. Thermodynamic forcing is typically modeled as regionally uniform or varying smoothly with latitude, snow cover and cloud cover. The impact of forcing on the growth or melt rate of level ice is dominated by heterogeneity at the meter scale, associated with spatial variability of ice thickness, snow depth and surface conditions [Perovich et al., 2003]. The heterogeneity is controlled by the super-position of thermodynamic response (growth/melt) on an icescape created by mechanical redistribution (leads/ridging/rafting).

From a remote sensing perspective, analysis of RADARSAT SAR imagery [Kwok, 2001] shows that lead systems often extend 100s of kilometers across the Arctic Basin, and these "linear kinematic features" (LKFs) display strain rates an order of magnitude higher than the surrounding ice pack. Ice growth in leads results in level ice which is often ridged or rafted when these leads close thereby introducing meter-scale heterogeneity into the spatial distribution of ice thickness. These processes constantly rework the surface morphology on sub-daily and synoptic time scales. Thus, sea ice deformation serves as the initial sculptor of spatial variability of sea ice thickness and surface morphology. It is the process of ice deformation and its impact on the mass balance of the sea ice cover that is the focus of this proposal.

To determine the sign and magnitude of the feedbacks we must improve our understanding of how new ice growth, ridging and rafting will respond to such things as: (a) increasing storminess in the Arctic; (b) a seasonal ice pack of reduced thickness; and (c) large scale changes in drift modifying ice stress. As we do not know whether current models adequately represent leads, we are uncertain of their ability to correctly represent observed sea ice strain rate and its impact on ice growth and redistribution. Models show that increasing deformation rates and variability result in increased total ice mass [Heil and Hibler, 2002; Kwok et al., 2003]. Yet, we lack quantitative validation of model estimates and the effect of deformation on ice mass. This is in part because the in situ data required for such validation is sparse and incomplete. As sea ice thickness is highly sensitive to modification of rheological and surface stress parameterizations, typically showing 10-20\% sensitivity [Kreysher et al., 2000; Hutchings, 2001], it is very important to simulate the stress-strain rate relation adequately.