Studies of melting blue ice areas

In the near-coastal regions of Dronning Maud Land, Antarctica, below-surface ice-melt in blue-ice areas has been observed. The low scattering coefficients of the large-grained blue-ice, allow penetration of solar radiation, thus providing an energy source below the ice surface. The sub-surface meltwater is significant enough to show up on remote-sensing imagery in the form of ice-covered lakes. Adjacent snow-accumulation areas have much higher scattering coefficients and consequently limit solar radiation penetration in these regions. These snow and ice surfaces are generally below freezing, and little surface melting occurs. To assess the response of these melt features to changes in atmospheric forcings such as cloudiness, air temperature, and snow accumulation, a physically-based model of the coupled atmosphere, radiation, snow, and blue-ice system has been developed. The model is comprised of a heat transfer equation with a spectrally-dependent solar-radiation source term. The penetration of radiation into the snow and blue-ice depends on the surface albedo, and the snow and blue-ice grain size and density. Model simulations show that ice melt occurring in this area is sensitive to potential variations in atmospheric forcing. Under certain conditions more traditional surface melting occurs and, under other conditions, the existing melt processes can be shut down completely. In light of the sensitivity of this system to variations in atmospheric forcing and the ability to view melt-related features using remote-sensing, a tool exists to efficiently monitor variations in Antarctic coastal climate. We are currently analysing to what degree this process contribute to the mass balance of Antarctica The data set includes: 1) Meteorological data from Jutulgryta, Dronning Maud Land, 2) Satellite images from Jutulgryta, 3) NOAA AVHRR mosaic mapping blue ice areas all over the Antarctic continent, 4) albedo measurements from Jutulgryta, 5) GPS data on surface topography, 6) Snow temperature records, 7) model results