RAPID GEOMORPHIC CHANGE IN AN UNCHANGING POLAR DESERT, ANTARCTICA

The McMurdo Dry Valleys, Antarctica, is the largest ice-free region on the continent, located along the edge of the western shore of the Ross Sea at about 78˚S. It is a polar desert with annual air temperatures of about -20^oC and summer air temperatures occasionally warming a few degrees above 0^oC. With no rainfall, little snowmelt, and glaciers frozen at their base, the rate of geomorphic change is slow. Indeed, high elevations landscapes are thought to have been largely stable for the past 14 million years. In this context, it was surprising to encounter, in January 2009, a rapidly incising stream in Garwood Valley, one of the McMurdo Dry Valleys. Field inspection showed incision of up to several meters in places and later aerial lidar differencing showed not only deep incisions in the main stream channel but also incisions in tributary channels as part of the geomorphic response.

Investigations in the upper elevations of the valley revealed two glacier-dammed lakes, each episodically draining by melt-erosion through the ice-dam formed by the toe of each glacier. Measurements and time-lapse imagery at one ice dam showed the lake draining in a few hours after breakout. We hypothesize that a similar glacier-dammed lake flooding event took place during an unusual warm period in the austral summer of 2001-02. The flood waters thermally eroded a narrow ice-dam between the glacier and valley wall draining the lake to another glacier-dammed lake below, which it too thermally eroded under the toe of that glacier sending the water down valley. The flood waters eroded through the ice-cemented permafrost mantling a layer of massive ice below. This ice was most likely emplaced during the last ice age when the Ross Ice Shelf entered the valley. Subsequent aeolian drift covered the ice, preserving it, when the Ross Ice Shelf withdrew. When the flooding stream waters reached this buried ice, it eroded quickly causing bank collapse in the main channel as well as incision in the tributary streams during the following years as the fluvial system adjusted to the new equilibrium. Calculations of flood volume and rates of incision will be presented. The warm event in 2001/02 thought to have triggered the flood may be unique in the past ~10,000 years as there is no evidence of prior stream incision.