Estimation of snow depth over open prairie environments using GOES imager observations

We assess the potential for estimating snow depth using observations in the visible and infrared spectral bands from the imager instrument onboard the Geostationary Operational Environmental Satellites (GOES). The approach makes use of a correlation between depth of the snowpack and satellite‐derived subpixel fractional snow cover over non‐forested and sparsely forested areas. To retrieve the snow depth we propose a simple analytical formula approximating the statistical relationship between the snow depth and the snow fraction. The primary focus of this study was the US Great Plains and Canadian prairies area. Daily maps of snow depth at a spatial resolution of 4 km have been produced for this region for four winter seasons from late 1999 to the beginning of 2003. Validation of the algorithm developed was performed through comparison of the satellite‐based product with snow depth measurements made at first‐order synoptic stations, US Cooperative Network stations and Canadian climate stations. The accuracy of snow depth retrievals was found to be about 30% of the observed snow depth for snow depths below 30 cm.