Accuracy Assessment of Different Open-Source Digital Elevation Model Through Morphometric Analysis for a Semi-arid River Basin in the Western Part of India

Accurate watershed delineation and drainage system extraction are critical aspects for the hydrological and geomorphological models. However, high-resolution digital elevation models (DEMs) are often not accessible without cost. This reason leads to assessing low-resolution open-source products that are freely accessible. Watershed morphology evaluation is very important for flood control and recovery planning and understanding basin hydrology. In this study, the Rel River Basin has been selected as the study area in northern Gujarat, India. DEM reliability is a significant element of such a study. The accuracy of DEM highly affects the accuracy of mapping the catchment area. This analysis has been therefore directed to compare various morphological watershed parameters depicted by the open-source DEMs, i.e., Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), Shuttle Radar Topography Mission (SRTM), and Cartosat and Advanced Land Observing Satellite (ALOS). A new approach has been introduced in this study to combine four DEMs (ALOS, SRTM, Cartosat, and ASTER) by averaging the pixels to enhance the accuracy of the calculated topography known as Average DEM. This study compares various open-source DEMs and newly generated Average DEM with digitized topographic data using morphometric parameters. Percentage error is used to compute the accuracy of the open-source DEMs and the Average DEM for every morphometric parameter. While comparing the open-source DEMs and Average DEM with the Survey of India (SOI) toposheet using mean percentage error (MPE) and the mean absolute percentage error (MAPE), Average DEM exhibited the least error. Hence, Average DEM has been considered more appropriate to derive the morphological characteristics for the Rel River Basin.