Snow Cover Monitoring Using Multi-temporal Envisat/asar Data

ABSTRACT A method has been developed to apply multi-temporal Advanced Synthetic Aperture Radar (ENVISAT/ASAR, C-Band) images to snow cover monitoring and mapping in mountainous areas. A multi-temporal dataset that includes sequences of ascending and descending ASAR wide swath and beam mode IS2 scenes acquired over Switzerland was investigated. The images were geometrically corrected to remove relief distortions, producing geocoded terrain corrected (GTC) image products. Backscattering coefficients (Beta, Sigma and Gamma nought) are typically calculated using a nominal local incidence angle value using a simplified ellipsoid geometry. We apply a more rigorous approach that models the backscattering coefficients normalised for local illuminated area (projected into the look direction for γ° retrieval), producing radiometric terrain corrected (RTC) image products. The snow cover was monitored by calculating the difference between the backscattering coefficients (γ°) of each ASAR image and the reference backscattering coefficient of a synthetic dry snow or snow free image. Similar to conventional fixed thresholding (e.g. 3 dB), a reference winter dry snow image was compared to each new ASAR image. Analysis of the resulting time series shows strong seasonal trends in backscatter behaviour, likely caused by variations of liquid water content in the snow cover. Meteorological data (MeteoSchweiz), NOAA images and snow cover maps from the Swiss Federal Institute for Snow and Avalanche Research supported interpretation and validation of the results.