Mapping distribution of Sundarban mangroves using Sentinel-2 data and new spectral metric for detecting their health condition

AbstractThe present study explores the suitability of Sentinel-2 images for mapping the physiological conditions of Sundarban mangroves and identifying suitable wavebands for the purpose. This work finds two shortwave infrared (SWIR) bands of Sentinel-2 appropriate and proposes a new spectral metric, namely Discriminant Normalized Vegetation Index (DNVI) based on these two bands. The spectral profile of mangroves is validated with the hyperspectral and high spatial resolution Airborne Visible Infrared Imaging Spectrometer-Next Generation (AVIRIS-NG) data. DNVI emphasized the importance of SWIR bands in segregating different health conditions of similar and dissimilar mangrove species assemblages. Classification using Support Vector Machine supervised learning model discriminated mangroves with an overall accuracy of 93.9%. The article highlighted the competence of high spatial resolution Sentinel-2 images and the proposed spectral metric can be exploited for mapping mangrove assemblages and their health a...