The importance of ice shelf melt rate distributions in modelling of Antarctic ice loss

Jowan Barnes, Hilmar Gudmundsson, Ralph Timmermann

Ocean-induced basal melting of ice shelves is a major cause of Antarctic ice loss. It is well known that the magnitude of this melt is an important factor to consider when modelling the future evolution of the ice sheet, and values of total melt volume or mean melt rates are often quoted for ice shelves. The finer detail of the distribution of this melt is usually given less attention. Here we make the argument for using coupled modelling frameworks such as Úa-FESOM, which employs unstructured triangular meshes in both the ice sheet and ocean components, providing added flexibility in resolving critical areas of ocean circulation and ice flow. We validate our model outputs by assessing correlation with Antarctic-wide observationally derived ice shelf basal melt rates, and a simple depth-based parameterisation of the type used in standalone ice sheet modelling. We show that the distribution of melt - independently of total melt volume - can significantly affect modelled ice flow, and we investigate the role of mesh resolutions in both components. Our results demonstrate the need for setups like Úa-FESOM, showing differences of 10% in the first year of ice sheet responses due to differing melt distributions and model resolutions.