71st Annual Meeting of the APS Division of Fluid Dynamics
Volume 63, Number 13
Sunday–Tuesday, November 18–20, 2018;
Atlanta, Georgia
Session E35: Geophysical Fluid Dynamics: Cryosphere
|
Hide Abstracts |
Chair: Andrew Wells, University of Oxford
Room: Georgia World Congress Center B407
|
|
Sunday, November 18, 2018
5:10PM - 5:23PM
|
E35.00001: Supercooled convective instability driven by double diffusion and ice crystal growth
Andrew Wells, Margaret Lindeman, Mary-Louise Timmermans
In the polar oceans in winter, a combination of cooling, ice growth and brine rejection creates water masses close to the insitu freezing point. We investigate the potential for convection at the base of the mixed layer driven by differential diffusion of heat and salt, supercooling and ice crystal growth, in conditions where both of the temperature and salinity stratifications are statically stable. We consider an initially warmer, fresher layer of salt water overlying a cold, salty layer, but with both at their salinity-dependent freezing temperature. A theoretical model is developed for energy and salt transfer allowing for ice crystal growth, in a limit with plentiful nucleation sites and rapid quenching of supercooling to a state of local thermodynamic equilibrium. Taking the asymptotic limit of small deviations in salinity, the resulting solutions show that rapid diffusive cooling leads to ice crystal growth at the base of the upper layer. We describe flow using the Bousinesq Navier-Stokes equations with a mixture density depending on temperature, salinity and ice concentration, and use a linear stability analysis is to determine the characteristic time and length scales for the onset of convection across a range of ocean salinities. |
Sunday, November 18, 2018
5:23PM - 5:36PM
|
E35.00002: Transition to turbulence in the oceanic boundary layer beneath ice shelves
Catherine A. Vreugdenhil, John R. Taylor
We use large-eddy simulations of the ocean beneath ice shelves to examine the transition to turbulence and the melt rate. The rectangular domain is bounded from above by the ice shelf base where a dynamic melt condition is imposed. An imposed far-field current provides a shear flow while the far-field temperature and salinity are restored to chosen values. For the majority of the simulations, the ice base is assumed to be horizontal, but in a series of additional runs the effect of small slopes is considered. We examine a range of far-field current strengths and temperatures to identify a fully turbulent regime and a relaminarising regime. The Obukov length scale, a comparison of the strength of shear flow turbulence to the buoyancy flux at the base of the ice, is successfully used to describe this transition to turbulence. We use parameters that are comparable to realistic Antarctic ice shelves to provide a real-world comparison. |
Sunday, November 18, 2018
5:36PM - 5:49PM
|
E35.00003: Direct Numerical Simulations of Forced Turbulence within an Ice-Shelf Ocean Boundary Layer
Leo Middleton, John R. Taylor, Catherine Vreugdenhil, Peter Davis
Ice sheets in Antarctica and Greenland have the potential to cause catastrophic sea level rise. However, their fate is highly uncertain. Part of this uncertainty lies in the interaction between ice shelves, the floating extensions of ice sheets, with a warming ocean. Here, we use idealized direct numerical simulations (DNS) to investigate melting beneath an ice shelf. The simulations are based on recent measurements made beneath the George VI ice shelf in West Antarctica (Kimura et al., 2015). The measurements suggest that turbulence associated with double-diffusive convection may influence the melt rate of the ice shelf. We investigate this scenario by forcing turbulence in the far field, away from the ice, where the temperature and salinity are relaxed towards imposed background values. A dynamic boundary condition allows the melt rate and the associated heat and salt fluxes to respond to the turbulence. The fluid density can be stabilizing or de-stabilizing depending on the parameters and initial conditions, and the melt rate changes accordingly. A comparison of the results with simulations of shear-driven turbulence and the observations suggests improvements to existing parameterization schemes.
|
