Index-Based Identification of Surface Water Resources Using Sentinel-2 Satellite Imagery

Abstract

Water is a key variable for the sustainability of the world. Surface water resources are of prime importance for all living creatures. Thus these resources should be monitored for proper water planning. In this study, we aimed to identify which spectral water index will represent water body better when the data of new imaging satellite, namely Sentinel-2, are used. Since these indices were originated from old Landsat missions, it is important to investigate the performances of these indices with other data resources. Catalan and Yedigoze dam reservoirs were considered as the study area. Normalized Difference Water Index (NDWI) and Modified Normalized Difference Water Index (MNDWI) were utilized as spectral indices. The obtained results showed that NDWI presented water body better than MNDWI when using Sentinel-2 data. As a statistical metric, the kappa values of NDWI and MNDW were obtained as 0.88 and 0.83 respectively. The results have revealed that remote sensing technology and remotely sensed images are important resources for surface water monitoring and management.