The impact of ice sheet thermal memory in Antarctica’s response to climate warming

The magnitude of the Antarctic Ice Sheet's response to future climate scenarios in ice sheet models depends on the choice of initial and basal sliding conditions. Basal sliding cannot be directly measured but is instead commonly inferred from observed surface velocity or ice thickness assuming the ice sheet is in equilibrium with the modern climate. The inferred basal sliding field is also affected by assumptions of different model parameters and the ice rheology, which all impact the modelled ice sheet behaviour. Ice rheology is often treated as idealised, prescribed as uniform, or also inferred from velocity observations. Such approaches lead to either a non-unique problem or to compensating errors in the inferred fields due to intrinsic uncertainties in the observations.To reduce compensating errors and not assume equilibrium with the modern climate, we force our ice sheet initial geometry with long-term temperature variations (i.e., a thermal spinup), thus generating a thermal structure (and therefore ice rheology) that is consistent with the ice sheet's long-term climate history. We assess different approaches to combine the thermal spinup with initialisation procedures for the Antarctic Ice Sheet, analysing their match to observed borehole temperatures at ice core sites. By initialising Antarctic Ice Sheet simulations with a thermal spinup, we improve our model’s initial conditions reducing the mismatch between modelled and observed ice sheet geometries and the uncertainty around the ice sheet's basal conditions and ice rheology with respect to basal and englacial temperatures. Finally, we use the different obtained initial states to show the impact of the ice sheet’s thermal history compared with idealised temperatures or equilibrium conditions on its sensitivity to future warming.