Time series analysis of dual-pol COSMO-SkyMed images for monitoring snow cover in alpine areas

Abstract

Time series of dual-polarized COSMO-SkyMed (CSK) images are exploited for detection of seasonal snow cover in Alpine areas. For the first time a complete time series of CSK images acquired during snow fall and melt period in winter 2010–2011 is addressed to verify the snow cover mapping capabilities of X-band radar images under different conditions (from dry to wet snow). The algorithm for snow detection is based on a multi-temporal approach with the concept that free water in the snowpack attenuates the X-band synthetic aperture radar (SAR) signal and wet snow can be classified by comparing images acquired under wet snow and snow-free conditions. Thresholds to make this distinction are compared across all the images to check sensitivity to different winter conditions and land-use classes. The impact on the snow cover area (SCA) detected is verified by also exploiting both polarizations in the form of Cross-pol ratio, ratio of VH channel with the reference image, and Depolarization factor, ratio between VH and VV channel of the same image. Snow maps from CSK images compared with LANDSAT ETM+ snow maps indicate a constant underestimation in the detection of snow extent especially during winter season thus showing a scarce sensitivity of X-band signals to snow in dry conditions. The presence of VH polarization indicated, however, an increase in the snow detection variable between 10 and 15%.