Winter snow cover variability on East Antarctic sea ice

• Published in: Journal of Geophysical Research, Washington, DC Washington, DC, 1998-10, Vol.103 (C11), p.24837-24855
• Main Authors: Massom, R. A., Lytle, V. I., Worby, A. P., Allison, I.
• Format: Article
• Language: English
• Subjects: Earth, ocean, space; Exact sciences and technology; External geophysics; Marine; Physics of the oceans; Sea ice; Snow. Ice. Glaciers
• ISSN: 0148-0227; 2169-9275; 2156-2202; 2169-9291
• DOI: 10.1029/98JC01617

Analysis of the first detailed data set of snow characteristics collected over East Antarctic sea ice in winter confirms that on small scales, snow on Antarctic sea ice is highly variable in both thickness and properties. High‐amplitude cyclical variability in atmospheric forcing related to the passage of storms is responsible for the high degree of textural heterogeneity observed. Changes in snow properties were examined over a 3‐week period, during which a largely icy snow cover, formed at near‐freezing temperatures, metamorphosed to snow in which facetted crystals and depth hoar dominated, as the air temperature plummeted. Even on flat ice, significant localized thickening of snow occurs in the form of barchan dunes. Although we observed great variability in snow thickness and properties on local scales, overall snow thickness distribution and the complex textural assemblage of snow types are similar from region to region. Similar observations were made by Sturm et al. [1998] in West Antarctica. Large‐scale similarities are also apparent in mean snow density, grain size, and bulk snow salinity, although high variability is again found across individual floes. Rapid depth hoar formation is a ubiquitous process that greatly affects the density, texture, grain size, and effective thermal conductivity of the snow cover. The observed heterogeneity results in varying snow effective thermal conductivities. The mean bulk effective thermal conductivity, computed from the proportion of observed snow types, is 0.164 W m−1 K−1, significantly lower than values typically used in large‐scale sea ice modeling but similar to that derived by Sturm et al. [1998] in a near‐simultaneous experiment in the Bellingshausen and Ross Seas. It varies from 0.097 to 0.383 W m−1 K−1 in different snow pits. The findings support those of Sturm et al. [1998] that periodic flooding and subsequent snow ice formation, which are also ubiquitous processes, effectively diminish the degree to which basal snow processes create inhomogeneities in the snow pack.