The modeling component comprises three subcomponents involving a hierarchy ranging from one-dimensional column modeling (1D), large eddy simulation modeling (LES), to regional scale modeling (RS) efforts. The modeling activities are intricately linked to the field program and to one another, with three basic objectives: to provide real-time service to the field program, to provide an interpretive and comparative basis for the analysis and synthesis of field observations, and to develop a physically-based parameterization of NES effects suitable for use in regional and global models.

There exists a variety of 1D, LES, and RS models in general usage, and we will use particular examples of such models with which we have greatest familiarity. For the 1D modeling we will utilize the Local Turbulence Closure (LTC) model of McPhee (1999) as well as bulk models currently under development parameterizing LES results on internal layer convection. The LTC model describes the variation of eddy diffusivity and viscosity in the vertical as dependent upon local flow field and hydrographic structure and similarity based dominant eddy scales. The LES model (Harcourt et al. 2002) is described above. The RS model is the fully coupled atmosphere-sea ice-ocean model of Holland (2003).

During the field program, modeling under way will assimilate concurrent observations and help guide measurement decisions using 1D models run on board and the RS model run remotely. Following the field season, LES modeling will contribute to the evaluation of observations by simulating ensembles of virtual observations in turbulence-resolving simulations of NES instabilities based on observations. We will use data from the winter process study along with the LES results to develop more accurate NES parameterizations in the 1D models. Once validated against the observations and the fine-scale modeling output, we will incorporate the 1D parameterizations into RS modeling (Holland 2001c) to simulate the regional scale behavior and, by implication, these parameterizations will be suitable for direct employment in a global ocean general circulation model. Following this strategy, we ultimately hope to make a meaningful contribution to the robustness and predictive capabilities of global climate models with regards to NES processes