Brine Percolation, Flooding And Snow Ice Formation On Antarctic Sea Ice

Dissertation (Ph.D.) University of Alaska Fairbanks, 2001 Modelling studies of brine percolation, flooding, and snow ice formation on Antarctic sea ice were undertaken to (1) determine the influence of brine transport processes on the salinity, porosity, and stable isotopic composition of snow ice a...

Full description

Bibliographic Details
Main Author: Maksym, Ted
Other Authors: Jeffries, Martin O.
Format: Doctoral or Postdoctoral Thesis
Language:unknown
Published: 2001
Subjects:
Online Access:http://hdl.handle.net/11122/8621
Description
Summary:Dissertation (Ph.D.) University of Alaska Fairbanks, 2001 Modelling studies of brine percolation, flooding, and snow ice formation on Antarctic sea ice were undertaken to (1) determine the influence of brine transport processes on the salinity, porosity, and stable isotopic composition of snow ice and the underlying ice, (2) explain the range of salinities and isotopic composition observed in ice cores, and to provide a better estimate of the contribution of snow ice to the thickness of the winter pack ice, (3) better understand the microstructural controls on brine percolation and its effects on the properties of sea ice, and (4) understand the effects of meteorological forcing on snow ice formation and development of the ice cover. Snow ice thickness is most dependent on snow accumulation rates. Once snow ice begins to form on a floe, most of the subsequent thickening is due to snow ice formation. Results show that percolation in winter sea ice is most likely an inhomogeneous process. Flooding most likely occurs rapidly through localized regions of high permeability, such as in large, open brine drainage channels or cracks. Simulations of the freezing of a flooded slush layer show that focussing of thermohaline convection may form porous drainage channels in the ice and snow. These channels allow rapid desalination of the slush and exchange of H218O depleted brine with sea water. Significant positive shifts in delta18O are possible in the slush layer. This process can explain the range of delta18O observed in ice cores. Based on these results, a cutoff of delta18O < -2� is recommended for snow ice identification in the Ross, Amundsen, and Bellingshausen seas. Such a cutoff puts the amount of snow ice observed at 6--18% of the ice thickness. Although flooding appears to occur through spatially restricted regions of the ice, the precise nature of the flow and factors controlling onset of percolation are unclear.