Application of hyperspectral data for Development of spectral library of mangrove species in the Sunderban Delta

Remote sensing has played a crucial role in mapping and understanding of the spatial pattern of mangrove forests and changes in its areal extent caused by natural disasters and anthropogenic forces. So far traditional pixel-based classification of multispectral imagery has been widely applied for broad mapping of mangrove covers. But the recent and more advanced hyperspectral data taken from sensors (like Hyperion) is expected to demonstrate the potential for reliable and detailed characterization of mangrove forests including species level classification. This paper demonstrates the potential of hyperspectral imagery for species level identification of mangroves in the Henry Island of Sunderban Biosphere Reserve, West Bengal. After pre-processing of hyperspectral data, the spectral signature of each species have been extracted from the Hyperion data after which a spectral library has been developed comprising the seven dominant mangrove species of the region namely Excoeocaria Agallocha, Avicennia Officinalis, Ceriops Decandra, Avicennia Marina and Phoenix Paludosa, Brugueira Cylindrica, Aegialitis.