Study of contaminated snow cover using remote sensing in the Eastern Himalayas of Arunachal Pradesh, India

Abstract

Snow is considered contaminated when any foreign materials are deposited/mixed with it, which can accelerate melting and significantly impact the snow cover's radiative balance. Such an enhanced melting rate results in a reduction in freshwater sources at the catchment level. In optical remote sensing, snow contamination is widely studied using a normalizing difference index called the snow contamination index. This is based on the finding that the impact of snow contamination diminishes with wavelength and is most noticeable in the visible spectrum (0.3—0.7 μm). However, the study of snow contamination using optical remote sensing is hindered in the Himalayan terrain due to enduring cloud cover in the region. Synthetic Aperture Radar (SAR) data such as Sentinel-1 can be used to ensure all-weather monitoring of such areas. This study focuses on the SAR backscattering behavior at the C-band of clear and contaminated snow for March 2022 in a part of the Eastern Himalayas of Arunachal Pradesh, India. An attempt has been made to utilize Landsat-9 and Sentinel-1 to study the snow contamination. The SAR backscattering for snow conditions (clear/contaminated) is studied using thresholds obtained from the Landsat-9 snow cover map. The SCI and SAR backscattering statistical analysis shows a negative correlation (R² > 0.6) at a 95% confidence level. It is observed that in the microwave region of the C-band, contaminated snow has a comparatively higher backscattering value than clear snow. However, in the visible wavelength, the contaminated snow has a lower reflectance value than clean snow. Such behavior of the snowpack in the microwave region of the C-band is explained using the physical properties of the snowpack and the dominant scattering mechanism over the surface. The key findings of this study suggest that SAR backscattering is affected by snow contamination due to changes in the local incidence angle, snow wetness, and surface roughness. This research provides critical insight into snow contamination using microwave remote sensing, which can be the first step toward developing an index for radar observations.