Ocean mixing and heat transport beneath Antarctic ice shelves

Abstract

Ocean-driven melting of floating ice shelves is currently the main process controlling Antarctica’s contribution to sea level rise. However, despite major advances in the study of ice shelf-ocean interaction over recent decades, key processes governing basal melting remain poorly understood, limiting our ability to accurately parameterise basal melting in numerical models used for sea level projections. This thesis aims to address these gaps by combining observational data analyses and numerical modelling techniques to provide insight into sub-ice shelf mixing and heat transport processes across a range of temporal and spatial scales. We begin by examining tidal currents and their influence on basal melt rate variability, using a multi-year dataset of oceanographic mooring measurements paired with autonomous phase-sensitive radar (ApRES) melt rate observations beneath Ronne Ice Shelf. While variations in near-ice current speed are identified as the dominant driver of basal melt rate variability at the study site, changes in thermal driving also contribute notably on spring-neap and longer timescales. Analysis of the tidal current vertical structure shows that the influence of ice base friction is enhanced in the semidiurnal frequency band, consistent with the site’s proximity to the semidiurnal critical latitudes. This leads to pronounced differences between near-ice and mid-water column semidiurnal tidal ellipses. Moreover, the extent to which the near-ice semidiurnal tidal ellipse deviates from the free-stream ellipse varies in time, and this temporal variability is reflected in the change in tidal current magnitude near the ice base relative to deeper in the water column. These findings challenge assumptions underlying a commonly used tidal melt rate parameterisation, which neglects the latitude- and time-dependence of frictional effects on tidal currents. Next, we investigate small-scale turbulent mixing processes within the ice shelf-ocean boundary current using idealised large-eddy simulations (LES). This study examines the interplay between ice base slope, mixing across the pycnocline, and basal melting, focusing on slopes appropriate to the grounding zone of small Antarctic ice shelves and to the flanks of relatively wide basal channels. The simulations reveal an unexpected relationship between the gradient Richardson number in the pycnocline and the ice base slope angle. Combined with a linear approximation to the density profile, as observed in the simulations, this relationship leads to a square root scaling between melt rate and slope angle. This suggests that the linear scaling reported by previous studies may not hold in the steepest parts of ice shelves, which dominate their overall melting. The derivation of the melt-slope scaling provides a potential framework for incorporating slope dependence into parameterisations of mixing and melting beneath ice shelves. Finally, this thesis examines heat transport beneath rapidly melting ice shelves in the Amundsen Sea using a high-resolution regional ocean model configured with realistic ice and seabed geometries but idealised forcing. Analysis of the sub-ice shelf heat budget reveals that advective heat flux, rather than turbulent heat flux, sustains basal melting. Moreover, across most of the domain, vertical advective heat transport dominates over horizontal transport despite vertical velocities being at least one order of magnitude smaller than the horizontal velocities. These findings, while requiring validation through further simulations, highlight the need to reconsider heat transport assumptions in basal melt rate parameterisations. Together, the three studies presented in this thesis provide a multi-scale analysis of Antarctic ice shelf-ocean interaction, from small-scale turbulence to cavity-scale circulation, and advance our understanding of some of the key physical processes that drive basal melting.