A New Parameterization of Dilation Using GODAR

Capturing all sea ice dynamical aspects in a model is notoriously challenging due to the complex interplay of granular and fracture-dominated processes. In the central Arctic, linear kinematic features (LKFs) dominate deformation patterns, while the marginal ice zone (MIZ) is characterized by fragmented floes where the collisional mode is dominant. The rheological properties of sea ice in these region differ significantly, and a rheological model that could be used in all regimes is desirable. Continuum models, commonly used for large-scale sea ice simulations, rely on parameterizations to approximate subgrid-scale processes such as floe interactions, wave attenuation, and dilation. Although high-resolution (<2 km) continuum models improve the representation of LKFs and deformation statistics, they remain fundamentally limited by their reliance on simplified, or ill-posed rheologies and the continuum assumption, which cannot reconcile velocity discontinuities inherent in granular materials like sea ice. Discrete element models (DEMs), on the other hand, explicitly resolve particle-scale interactions and naturally capture fracture and granular behaviour, but their computational cost has historically restricted their application to small-scale scenarios.

We addressed this gap by developing the granular floes for discrete Arctic rheology (GODAR) model, a DEM specifically designed to simulate the mesoscale evolution of sea ice mechanics. GODAR tracks the time evolution of contact normals between floes, enabling us to derive generalized equations that relate dilation to prognostic variables such as shear and normal stress, open water fraction, and floe size distribution. These results demonstrate that GODAR effectively captures both the granular physics and fracture-driven dynamics underpinning LKFs. By seamlessly integrating microscale processes into macroscale behaviour, GODAR offers a powerful framework for bridging the limitations of continuum models. Its insights provide a pathway to improved parameterizations, advancing both the scientific understanding of sea ice dynamics and the operational forecasting capabilities necessary for safe navigation and climate modeling.