Sculpting Polar Seascapes
 
Andrew Roberts
Fluid Dynamics and Solid Mechanics, Theoretical Division, LANL
 
Sea ice covers roughly 7-10% of Earth’s ocean surface, and its areal coverage significantly affects the total solar reflectivity of our planet, partly due to the ability of frozen sea water to carry snow.  The thickness and equatorward extent of sea ice has undergone rapid change during the 21st century, both in the Arctic and Southern Ocean.  Simulations suggest that earth-system feedbacks from the loss of perennial Arctic sea ice are contributing to a loss of permafrost on land, changes in weather patterns, and transitions in the oceanic biogeochemistry of the high north.  However, there are limitations in the Earth System Models (ESMs) making these predictions, not least of which in the physics of sea ice represented within them. Some sea ice phenomena have proven stubbornly difficult to represent mathematically.  Consequently, aspects of sea ice models, including in the Department of Energy’s Energy Exascale Earth System Model (E3SM), retain empiricisms of the past.  In this talk, I present a solution to one of those problems: vertical sea ice deformation, otherwise called ridging, for which my colleagues and I have derived a solution from first principles that will permit its representation in models free from empirical approaches of the 1970s still being used in advanced ESMs.   The consequence of our variational method is that we have a new theoretical framework to better understand wind and current drag on sea ice, the grounding of sea ice causing landfast ice along coastlines, and most importantly, the compressive strength of the pack at variable resolution and on unstructured computational meshes. We are poised to implement our new physics in E3SM and make direct comparisons between theoretically-derived polar seascapes and sea ice topographic measurements using satellite laser altimetry, addressing a requirement of E3SM to provide practical guidance on the current and future state of our polar regions and accuracy of our model.