Understanding Calving by Combining Discrete and Continuous Methods

Abstract:

Calving is a significant challenge in the ice sheet modelling community, both in terms of the nature of the parameterization in continuum models (the ‘calving law’) and in terms of the remeshing and interpolation of continuum model domains following calving. The CALISMO project combines the finite element model Elmer/Ice with the Discrete Element model HiDEM. Elmer/Ice has the remeshing capabilities required to simulate calving events, while HiDEM simulates fracture directly, allowing us to investigate event-scale calving in detail. Applied to Store Glacier, West Greenland, we find that HiDEM results correspond well with observations, providing an informal validation of the model. Importantly, it was not necessary to tune the model physics to this specific glacier. Our results suggest that calving is initiated in regions of high extensional stress, and that fractures tend to propagate orthogonally to the maximum principal stress (i.e. Mode I fracture). How best to implement such a calving law into a continuum ice sheet model remains an open question.