Enhancement of Ice Melting Rates via Homogeneous Isotropic Turbulence

A significant portion of sea level rise comes from the melting of ice shelves at ice-ocean interfaces. Ice shelves are attached to and buttress tributary glaciers; however, once an ice shelf melts it is typically replaced by traveling glaciers in a continuous process. In Antarctica, the Circumpolar Deep Water (CDW) current carries warmer waters to and underneath ice shelves. The CDW along with buoyant meltwater plumes and temperature gradients generate turbulence, thereby facilitating rapid mixing. The rate at which cold meltwater surrounding ice is replenished by warmer water is thus significantly enhanced by turbulence and drives melting. This process is analogous to turbulent diffusivity, whereby interfacial dynamics, in this case melting, are accelerated compared to scenarios with negligible ambient flow.

Currently absent from the literature is a thorough quantification of the effect turbulence has on ice melting rates. To understand the underlying physics of turbulence on melting rates, we designed an experimental tank in which turbulence is generated by random jet arrays around the exterior sidewalls, thus generating homogeneous isotropic turbulence absent mean flow in the center of the tank. While in the ocean turbulence is typically found with waves and currents, baseline conditions can be established for ice sheet modeling applications with this fundamental experimental study. An ice sphere dyed with Rhodamine B was placed stationary in the center of the tank. To determine the respective contributions to melting by temperature gradients compared to turbulence, melting rates were quantified under conditions varying both ambient water temperature and turbulence intensities. The flow visualization technique laser-induced fluorescence (LIF) was used to quantify instantaneous melting. Simultaneous particle image velocimetry (PIV) measurements were used to measure velocity fields surrounding the ice. Statistics such as turbulent kinetic energy, dissipation, spectra, and integral scales were computed from PIV data. Ultimately, we determine melt rates in response to varying turbulence levels across a range of ambient temperature gradients that contribute to melting processes in dynamic environments.