91社区

Dr. Arnold Song joins us from the US Army Cold Regions Research and Engineering Laboratory for his seminar Should sea ice models be more discrete?: High-resolution sea ice modeling with the discrete element method. Refreshments will be served.

Abstract:

With the push toward higher and higher spatial resolution in sea ice models, we enter a scale regime where the continuum assumption breaks down and is no longer mechanically equivalent to the physical system. We present the CRREL Sea Ice model, which is based on the discrete element method (DEM), as an alternative numerical framework for high-resolution modeling of sea ice dynamics. The CRREL Sea Ice model treats the ice pack and floes as a collection of ice 鈥済rains鈥� to describe the sea ice as a system where the effects of element-to-element interaction significantly influence the mechanical behavior of the bulk material. These pieces of ice have sufficiently small length scales to capture meter scale heterogeneity of the ice pack, which may govern the bulk behavior at the scale of interest (10-1000 km scale). To capture this heterogeneity, the shape, thickness and a bond strength between adjacent pieces of ice is explicitly described for each grain. In addition, a unique feature of the CRREL Sea Ice DEM is the capability to specify bond strengths that are independent of ice grain thickness, which enables the model to mimic the presence of linear flaws in the ice. Our modeling approach has been successfully applied to model sea ice processes such as pressure ridging (Hopkins 1998), aggregation due to wave-ice interaction (Hopkins & Shen 2001), and the mesoscale evolution of the floe size distribution (Hopkins & Thorndike 2006).