Simulating fractal properties of sea ice deformation in a high-resolution Arctic System Model

Recently, strong spatial and temporal scaling of sea ice deformation has been observed in the Arctic using the Radarsat Geophysical Processing System (RGPS) and Global Positioning System (GPS) equipped buoys. However, no such scaling properties have yet been demonstrated in sea ice models that use Elastic Viscous Plastic (EVP) or Viscous Plastic (VP) mechanics, despite several groups searching for a fractal signal in their model output. This outcome has led some to conclude that new rheologies are required to simulate high spatial and temporal resolution sea ice motion in order to correctly capture observed “Linear Kinematic Features” as well as shear and divergence scaling observed at spatial scales of ~5-500 km and temporal scale of ~1 hour – 1 month. However, no scaling tests have yet been reported that use high resolution, fully coupled ice-ocean-atmosphere-land models, where the presence of brown noise from mesoscale atmospheric systems and eddy-permitting ocean circulation more realistically emulate the ice-ocean boundary layer than do stand-alone ice-ocean hindcasts or typical Global Earth System Models. We demonstrate that the observed sea ice scaling behavior does, in fact, exist in a current-generation EVP sea ice model. Using the Regional Arctic System Model, composed of the Parallel Ocean Program (POP) and Los Alamos Sea ice Model (CICE) at 9km resolution, coupled to the Weather Research and Forecasting Model (WRF) and Variable Infiltration Capacity (VIC) model at 50km resolution, we detect fractal dimensions between -0.23 and -0.32, compared to the observed range of -0.18 to -0.20 in spatially scaled sea ice deformation. The range in our modeled fractal dimension depends on the chosen sampling period, a factor overlooked in most observational studies. We demonstrate that the presence of inertial waves in the ice-ocean boundary layer, induced by frequent (20 minute) atmospheric coupling, plays a significant role in fractal-like sea ice model behavior, as compared with poor scaling properties of our stand-alone ice-ocean model simulations using 6-hourly reanalysis forcing.