Capability assessment of Sentinel-1 data for estimation of snow hydrological potential in the Khanabad watershed in the Hindu Kush Himalayas of Afghanistan

Abstract

The water resource in arid countries like Afghanistan is mainly dependent upon the rivers driven by snowmelt runoff. The Khanabad watershed in Afghanistan contributes significantly to the river network system in this country. This study investigates the role of freely available Sentinel-1 dual polarimetric and interferometric SAR data to determine snow hydrological potential in the Khanabad watershed. The objectives of this study are to assess the Snow Water Equivalent (SWE) estimates derived from the Sentinel-1 data with respect to field observations and ERA5-Reanalysis data for February and March 2019. The snow depth and snow density estimates are derived using the interferometric and polarimetric methods to determine the SWE. The snow depth is determined using differential interferometry with snow-free scenes and snow-covered scenes acquired by Sentinel-1 data. The density is empirically modeled using the permittivity estimates derived from bi-temporal VV-channel Sentinel-1 data corresponding to snow-covered scenes. The comparison between the SWE results observed from Sentinel-1 data with respect to field observations showed high agreement determined by the overall correlation of 0.877 and a Root Mean Square Error (RMSE) of 37 mm. The higher RMSE was attributed to the uncertainties introduced by the presence of liquid water content in the March 2019 dataset.