Item

Abstract

Understanding ice shelf–ocean interaction is fundamental to projecting the Antarctic ice sheet response to a warming climate. Numerical ice shelf–ocean models are a powerful tool for simulating this interaction, yet are limited by inherent model weaknesses and scarce observations, leading to parameterisations that are unverified and unvalidated below ice shelves. We explore how different models simulate ice shelf–ocean interaction using the 2nd Ice Shelf–Ocean Model Intercomparison Project (ISOMIP+) framework. Vertical discretisation and resolution of the ocean model are shown to have a significant effect on ice shelf basal melt rate, through differences in the distribution of meltwater fluxes and the calculation of thermal driving. Z-coordinate models, which generally have coarser vertical resolution in ice shelf cavities, may simulate higher melt rates compared to terrain-following coordinate models. This is due to the typically higher resolution of the ice–ocean boundary layer region in terrain following models, which allows better representation of a thin meltwater layer, increased stratification, and as a result, better insulation of the ice from water below. We show that a terrain-following model, a z-level coordinate model and a hybrid approach give similar results when the effective vertical resolution adjacent to the ice shelf base is similar, despite each model employing different paradigms for distributing meltwater fluxes and sampling tracers for melting. We provide a benchmark for thermodynamic ice shelf–ocean interaction with different model vertical coordinates and vertical resolutions, and suggest a framework for any future ice shelf–ocean thermodynamic parameterisations.