Modeling Vapor Diffusion in Sublimation Tills of the Antarctic Dry Valleys: Implications for the Preservation of Near-surface Ice on Mars

Introduction: Interest in buried glacial ice has gained considerable attention in recent years due to its potential as an archive for long-term climate change and as a consequence of recent suggestions that call for relatively young, near-surface buried ice on Mars. Geochemical analyses of ice stored in stagnant debriscovered glaciers in the western Dry Valleys region of Antarctica (stable upland zone, [1]) may ultimately extend records back into the Miocene, well beyond that now possible from analyses of ice at Vostok and Dome C (e.g. [2]). At issue, however, is whether these stagnant debris-covered glaciers can maintain a core of glacier ice for millions of years, or whether ice sublimation would remove all traces of original glacier ice over these time scales (e.g., [3,4,5]). In order to address the question of the longevity of buried glacier ice in the Dry Valleys, we modeled summertime vapor flow through an ancient sublimation till that caps a buried glacier in central Beacon Valley (Fig. 1). The age of this underlying glacier ice is debated (e.g., [6]), with published ages ranging from ~300 ka [7], to > 2.3 Ma [5,8], to > 8.1 Ma [3]. In this paper we outline the range of climate conditions necessary to preserve the buried ice for millions of years. Our approach is to first calculate rates of summertime sublimation and vapor flow under existing climate conditions (atmospheric temperature and relative humidity, solar radiance, soil temperature and moisture) and then calculate sublimation rates for a range of plausible climate scenarios that may have occurred in this sector of Antarctica over the last several million years. Geologic setting: The buried ice in central Beacon Valley is stagnant (zero horizontal motion, [9]) and contains 3-wt% debris; it rests beneath a thin sublimation till that is on average 50-cm thick. Debris within the ice is commonly concentrated in bands up to 10-cm thick and includes clay-to-cobble-sized clasts of Ferrar Dolerite, Beacon Heights Orthoquartzite, and granite erratics foreign to Beacon Valley [8]. Sublimation of the ice has thus far produced the thin protective cap of sublimation till that mantles the ice (Fig. 1). Schaefer et al. [5] showed ice sublimation decreases with increasing till thickness and [8] found that the development of highcentered polygons at the till surface also exerts a strong control on ice sublimation. Initially rates of sublimation are highest at immature polygon troughs, but as troughs deepen via sublimation, they become preferred sites for windblown snow; this snow cover reduces underlying ice sublimation and in many cases leads to the formation of secondary ice [8]. To a first order, then, ice sublimation is controlled by the rate of ice loss at polygon centers (see also Fig. 1). Hence we have focused our analyses on sublimation processes at polygon centers.