Abstract
This paper describes the feasibility of mapping snow cover extent by means of time-series of SAR images of the sensor ENVISAT ASAR in the conditions of the Czech Republic. The methodology is based on evaluating a change in absorption of a radar signal in the snow pack due to the liquid water content.
Two images with same acquisition geometry are compared: the first one is a reference image acquired under dry snow or snow free conditions, the second one gathered during a snowmelt period. The main objectives were to suggest most suitable conditions for acqui-sition of reference images and to define a threshold value for wet snow classification with respect to a chosen reference image and a type of land cover.
Results were compared with maps of snow cover derived from MODIS and MSG data and with data from meteorological stations of the Czech Hydrometeorological Institute (CHMI). The same methodology was also tested on three TerraSAR-X images.
The results showed that the most suitable reference image should be acquired under dry snow conditions. The thresholds for detection of wet snow are in general around 0 dB or higher.
At the end of melting season with no remaining snow in lowlands it is necessary to mask out areas without snow cover in lower altitude.