Topographic controls on subpixel change detection in western Himalayas

Abstract Over Himalayas, the detection of seasonal snow cover variations via remote sensing is challenging due to the presence of rugged terrain and shadow effects. Moreover, traditional per-pixel-based change detection methods less competent to identify the inner-pixel subtle variations with coarse resolution satellite imagery. In the present work, we have proposed the subpixel-based change detection (SCD) algorithm to examine the impact of topographic correction (TC) on the SCD algorithm. For experimental purposes, the SCD has been implemented over a part of the Himalayas (Himachal Pradesh, India) using two satellite sensor datasets i.e. AWiFS and Landsat-8 OLI/TIRS. Experimental outcomes have confirmed the effectiveness of topographically corrected SCD (86.17% with AWiFS and 85.38% with Landsat-8) as compared to topographically uncorrected SCD (81.82% with AWiFS and 84.19% with Landsat-8). The proposed framework is proven to be significant in terms of both information extraction from shadowed areas and detection of inner-pixel subtle variations. This study has many applications especially in the field of climatology, glaciology, and hydrology.