METHOD AND ACCURACY OF ESTIMATING SNOW DEPTH USING SENTINEL-1 SAR DATA IN GUNMA PREFECTURE IN JAPAN

Abstract

 Using Sentinel-1 SAR data, the estimation method of snow depth was studied and its accuracy was verified at snow depth observation sites in the upper reaches of the Tone River in Gunma Prefecture, Japan. The estimation of snow depth was confirmed to be possible based on the variation of backscattering coefficients of VV and VH polarization with snow depth in the forested area in this region. The estimation accuracy of the snow depth below 400 cm based on the backscatter coefficients for a 500 m meshed slope covered with trees was about 20 to 50 cm, indicating that this method can be used to evaluate the snow cover situation over a wide watershed area for estimating snowmelt.