Ice melting in saltwater: laboratory experiments in the diffusive-convective regime

The melting of Antarctic ice shelves is driven by heat fluxes from the underlying ocean to the ice. The relationship between basal heat fluxes and ocean conditions is an active topic of research, as current state-of-the-art parameterizations perform relatively poorly in all but the fully-turbulent well-mixed regime^a. Indeed, discrepancies between observed and predicted melt rates under certain ice-shelves have been detected^b. Here, we  perform laboratory experiments of tabular ice cuboids melting in salty water. We aim to improve our understanding of the basal melting of ice shelves in the diffusive-convective regime, for which there is currently no parameterization^c. To this end, we investigate the melting rate and underlying fluid dynamics over a broad range of water salinity and temperature, without any external forcing. Our work uniquely complements field observations, which are difficult and sparse, and simulations, which most often approximate the dynamics for computational expediency.

We use a meter-scale tank, which we fill with saltwater and place a freshwater tabular ice cuboid to melt on top. A bottom heating plate is used to maintain the bottom saltwater temperature at a prescribed value and the setup is placed in a cold room. The depth-dependent seawater temperature, salinity, currents, and the melt rate are explored for different bottom water temperatures and initial salinity values. We use a moving temperature and salinity sensor, PIV data and shadowgraphy images of the retreating ice-water front to provide a relatively comprehensive data set from which we derive a mapping between the average melting rate and the flow statistics (kinetic energy density, dissipation rate, temperature gradient) of interest to polar oceanography.

We find that temperature and salinity vertical gradients in the system can create a layered system, depending on the conditions. In particular, we observe the formation of a freshwater layer insulating the ice plate, and slowing the melting, at relatively low temperature. When the bottom temperature is relatively large, the two-layer organisation disappears as convection becomes vigorous enough to penetrate and mix the freshwater layer with the ambient. 

References :

        a - Malyarenko, A., Wells, A. J., Langhorne, P. J., Robinson, N. J., Williams, M. J., \& Nicholls, K. W. (2020). A synthesis of thermodynamic ablation at ice–ocean interfaces from theory, observations and models. Ocean Modelling, 154, 101692.

        b - Kimura, S., Nicholls, K. W., \& Venables, E. (2015). Estimation of ice shelf melt rate in the presence of a thermohaline staircase. Journal of Physical Oceanography, 45(1), 133-148.

        c - Rosevear, M. G., Gayen, B., \& Galton-Fenzi, B. K. (2022). Regimes and transitions in the basal melting of Antarctic ice shelves. Journal of Physical Oceanography, 52(10), 2589-2608.