Measurement of ice stability in the high elevation Dry Valleys, Antarctica

The high-elevation Dry Valleys of Antarctica are a unique location for studying possible formational mechanisms and stability of permafrost. In particular, the hyper-arid and cold conditions do not allow for the presence of liquid water, meaning the exchange of water between the atmosphere and subsurface is through vapour diffusion. The hyper-arid conditions also result in significant desiccation of the subsurface, thus forming a dry permafrost layer - soil which has a temperature always below 0°C (cryotic), but which is dry (not ice-cemented). The high-elevation Dry Valleys of Antarctica are the only known location on Earth with a dry permafrost layer overlying the ice-cemented ground; dry permafrost is common on Mars. We have made direct calculations of the stability of subsurface ice in University Valley (1730 m; 77°S 51.8', 160°E 43'), by measuring the temperature and humidity (air, surface, subsurface) at different locations along the valley at high temporal resolution (30 min). We then calculated the vapour pressure at each location, which shows the movement of water vapour throughout the year. Based on atmospheric measurements, the mean atmospheric vapour pressure is lower than the mean vapour pressure over ice in the subsurface, and thus the subsurface ice should be quickly retreating. Yet ice-cemented ground is observed throughout the valley. This may be explained by our surface measurements which suggest that the ground is snow covered for more than half of the year. This results in the subsurface seeing an upper boundary condition which has a significantly higher vapour pressure than if the subsurface was directly responding to atmospheric conditions. Vapour flux calculations based on our temperature and humidity measurements show that while using the atmospheric conditions clearly shows the subsurface ice to be unstable, using the surface measurements results in a vapour flux, and ice retreat rate, which is over two orders of magnitude lower and effectively zero. Similar measurements at three other locations in the valley support the importance of snow cover. We will discuss our measurements, calculations, and models for ice stability and distribution.