The distribution and magnitude of winter-season water equivalent stored in the snowpack across the Canadian tundra are presently unknown. Quantifying this storage, as well as the degree of interannual variability, is essential, because the tundra snowpack is a persistent landscape feature and a key parameter in water, energy, and biogeochemical cycling. Tundra snow cover represents the solid storage term in the water balance, affects the energy balance through changes in surface radiative properties, and controls the ground thermal regime and vegetation dynamics. Emphasis must be placed on understanding climate change impacts on terrestrial Arctic and sub-Arctic snow cover because of observed and simulated warming conditions combined with coincident increases in winter precipitation (Watson and Core Writing Team, 2001; ACIA, 2004). However, regional simulations of snow cover and water balances by global and regional climate models are hampered by the lack of appropriate snow cover data for comparison and validation.The determination of past and present snow cover conditions relies heavily on in situ datasets that are limited by recent reductions in manned meteorological stations, differences in instrumentation, and biases in measurement techniques (Yang and Woo, 1999). Furthermore, in situ sampling provides a measure of snow depth, snow water equivalent (SWE), or snowfall at discrete point locations which are often not representative of regional snow cover conditions required for large scale monitoring and modelling. For instance, conventional snow cover observations across the Canadian tundra are extremely sparse and biased to the Arctic coast.The non-uniform snow distribution found in high-latitude environments is primarily a function of complex terrain and landscape features which control wind redistribution. Despite a high degree of heterogeneity, snow distribution patterns can be modelled based on terrain characteristics and meteorological conditions. Slope length, angle, aspect and ground cover, along with wind direction, magnitude and fetch, can all be used to model snow distribution patterns successfully (e.g. Essery and Pomeroy, 2004). This approach can provide a perspective on regional snow cover, but it is still limited by the need for forcing data to generate model simulations and requires independent datasets for model evaluation.Remote sensing techniques have long been employed to monitor terrestrial snow cover. Visible wavelength imagery has been