O02: Implementing Scale-Aware Sea Ice Ridging in E3SM

Implementing Scale-Aware Sea Ice Ridging in E3SM

• @Andrew Roberts

  • afroberts@lanl.gov

@Elizabeth Hunke

Ocean - Ice

E3SM

For some time, is has been apparent that a core state-space descriptor of sea ice models, the thickness distribution, g(h), insufficiently describes pack morphology necessary for representing features integral to coupled modeling:  snow-on-sea-ice, melt ponds, oceanic and atmospheric drag, land fast ice, and sea ice strength.  A significant problem with g(h) is that it offers only a glimpse of the much richer morphological sea ice state describing sub-grid scale ice roughness and porosity seen in the real world. g(h) itself relies on a constant dynamic density of ice through the thickness, h, when in reality, macro-porosity, 𝜙, is non-uniform in ice rubble build-up during ridging, and can result in as much as 30% reduction in local pack density. We are implementing a new variational approach to ridging in E3SM that tackles this problem, and extends the thickness state space to include macro-porosity with a distribution g(h,𝜙) to replace g(h).  This work builds upon three publications that established the theory of variational ridging, verified idealized results against a discrete element model of ridges, and evaluated a column implementation of the new ridging model against a large dataset of deformed Arctic sea ice. The method has also been favorably assessed against sea ice topography detected by ICESat-2 in an independent study. Based on these developments, we outline here the implementation and evaluation plan using orbital emulators to be completed during E3SM Phase III.

We are at the meeting, but the poster is virtual

https://acme-climate.atlassian.net/wiki/spaces/ICE/pages/3833365075