Potential of Information Fusion of Optical and SAR Data for Snow Cover Characterization

Abstract

Himalayan snow contributes significantly in towards water resource in several countries in Asia. In mountain hydrology, timely information on the spatial and temporal variability of the extent of wet snow is significant in snowmelt runoff models. Recent advances in remote sensing technology have enabled us to study the extent of snow cover using different techniques based on remote sensing data. The all-weather capability of Synthetic Aperture Radar (SAR) makes it extremely useful especially in difficult alpine terrains. Due to the penetrability of microwave signals, dry snow often exhibits scattering from the underlying soil surface. Snow exhibits very high reflectance in the visible and near infrared spectrum. Thus, using either optical or active microwave data, identification of wet and dry snow for characterization of snow surface is a complicated problem. This paper investigates the potential of optical and SAR imagery for identification of dry and wet snow in alpine regions by considering a synergistic multi-sensor framework. In this study, the wet snow is identified using Sentinel-1A and ALOS-2 PALSAR-2 and simultaneously the dry snow is determined using Landsat-8 multispectral data. The study is conducted around a glacier area in Sarsot near Gulmarg in Jammu and Kashmir, India.