Permafrost

organogenic horizons of soils for the long period. Taking into account the scales of thermokarst
distribution and high carbon content in buried horizons this influence is rather essential.
Key words: Role, thermokarst, global balance of carbon
138
Origin and Stability of Ground Ice in the Dry Valleys, Antarctica
Ronald S. Sletten 1 , Birgit Hagedorn 1 , Warren W. Dickinson 2 and Bernard Hallet 1
(1.University of Washington, Department of Earth and Space Sciences & Quaternary Research Center,
Box 351310, Seattle, WA, 98195 USA; 2.Department of Geology, Victoria University of Wellington,
Wellington, New Zealand)
Abstract: Subsurface ice occurs with one meter in most soils in the McMurdo Dry Valleys,
Antarctica. The ground ice exists as massive ice or as pore ice. The existence and persistence of
this ice is of interest since ablation models suggests the ice should be absent in all but the
youngest soils (or in all soils older than 10,000 yrs) and yet ice is ubiquitous under much older
surfaces. Since ground ice drives development of many landscape features, understanding the
factors that lead to its formation and stability are of interest for understanding polar landscapes
and the history of ice sheet advance and retreat. Furthermore, the Dry Valleys provide our best
analog for understanding the presence of ground ice on Mars. This presentation highlights
various types of ground ice in the Dry Valleys as well as their chemical and physical
characteristics.
Ground ice can be broadly classified as: (1) pore ice cementing soil/sediment; (2) modern
or remnant buried glacial ice; and (3) syngenetic ice. Buried lake ice may also exist but is not
presented here. Each of these types display considerable variation in physical and chemical
characteristics; the origin of the ice may be difficult to determine. Pore ice is the most common
form of subsurface ice. Typically it is present within a few decimeters of the soil surface in
areas displaying contraction cracks and polygonal patterned ground. Our data on air and soil
temperature and humidity together with our model of sublimation, as well as published models,
suggest that this ice should have sublimated long ago given the age of the soils. The persistence
of ice-cement reveals that the models are incomplete. We are investigating the potential role of
snow-cover and episodic meltwater recharge in the long-term occurrence of subsurface ice in
cold hyper-arid soils.
One of the most debated aspects of ground ice is the age of buried glacial ice, which is
found in numerous locations at elevations of 400 m and higher. Three distinct assemblages have
been identified: moraines cored with ice from a former polar glacier, deposits from temperate
glaciers of unknown age, and active debris-covered glaciers. The surface of all three forms
displays polygonal patterned ground; however, relative to this form of patterned ground found
in sediments, the micro-morphology is distinct, consisting of low-relief central domains
bordered by deeper troughs, presumably due to the near-surface ablation of ice sublimating
along the contraction cracks.
The third type of ground ice is syngenetic ice. This type of ice is least common and its