Incorporation of frazil ice into a sea ice-ocean model

Lead Research Organisation: University College London
Department Name: Earth Sciences


This project relates to the priority topic in SOFI theme 1, 'Climate, ..', 'Parametrisation of detailed sea ice physics to be suitable for ocean GCMs (studentship; WP 1.8 also WP 9.1-9.4).' Rapid changes have been observed in the sea ice cover of the Arctic Ocean, with record summer lows of sea ice extent in 2005 and 2007. Accurate predictions of the future sea ice cover require adequate representations of physical processes. This project will address uncertainty in the representation of winter sea ice formation in Global Climate Models (GCMs). The winter sea ice cover of the polar oceans insulates the ocean from the atmosphere, so that conductive heat losses to the atmosphere are reduced from 100-200 Wm-2 to 5-10 Wm-2. Open water regions arise due to mechanical formation of leads, which are long, narrow cracks in the ice cover; and wind blowing ice away from a coast/land-fast ice boundary to form coastal polynyas. During winter, the open water in leads and coastal polynyas is exposed to typical air temperatures of -40oC and quickly becomes supercooled, i.e. its temperature is lowered to below the freezing point of sea ice water. Rapid formation of frazil ice, millimetre-sized, disc-shaped ice crystals, quenchs the supercooling through the release of latent heat. Heat loss through leads and polynyas is rapid (e.g. 1000 Wm-2) and they are sources of new sea ice and salt, released into the ocean as the ice forms. While leads occupy only 1-2% of the area of the sea ice cover in winter, they contribute approximately half of the heat loss into the atmosphere from the ocean. Adequately accounting for the heat loss/salt rejection in leads and polynyas has a leading order impact on predictions of ice thickness, extent and deep water formation. The aim of this Ph.D. project is to develop a new model of the production of frazil ice crystals suitable for the sea ice/ocean component of a GCM. This will be done by building upon the frazil dynamics model of Holland and Feltham [2005], which was developed to describe the production and precipitation of frazil crystals in the ocean cavity beneath an ice shelf. This frazil model accounts for advection and turbulent diffusion, with the population dynamics encapsulated in size-class interaction terms. Inter-class transfers occur by frazil growth, melting and secondary nucleation. Growth and melting result in salt release/dilution. The frazil model contains a prescription for precipitation of frazil crystals to an upper surface based on the local level of turbulence and the frazil buoyancy. The precipitation model depends upon a Richardson number, modified to take account of the variable effective molecular viscosity of the frazil-water mixture, evaluated at the centre of the bursting layer next to the surface. The student will adapt the frazil dynamics model to the situation of frazil ice formation within open water regions in a sea ice cover, which will involve an examination of fundamental model assumptions. This model will be built and tested in stages, using laboratory and field studies. The frazil model will be coupled to a vertical ocean mixing model with an imposed horizontal advection and compared with estimates of the ocean-air heat flux measured over leads and polynyas. Frazil ice production, heat loss, and salt release rates will be calculated and related to GCM variables, e.g. sea surface temperature, and air temperature, in order to determine a parameterisation for sea ice and ocean models. The POL Arctic shelf seas model will be used to test the frazil model. The project is in the context of a coherent programme at CPOM and POL, which provide a climate of excellence in research, and will benefit from weekly meetings with expert supervisors. UCL provides a programme of transferrable skills, monitoring and assessment. Holland, PR & DL Feltham, Frazil dynamics and precipitation in a water column with depth-dependent supercooling, J. Fluid Mech., 530, 2005.


10 25 50