Habitat Based Identification Of Foredune And Incipient Foredune By Per Pixel And Sub Pixel Approach, A Case Study From Panaiyur Coast, Tamil Nadu, South India

Abstract

Coastal sand dunes, livelihood of different habitats, are the most important barrier in the coastal environment against coastal hazards. Dune flora acts as a binding material in budding the dunes (Incipient/Embryo) and for the accumulation of sand particles in building dunes and acts as protecting agent against erosion. Majority of studies on dunes are based on field, which characterizes the dunes based on distance from the present coast and height and a few authors have mentioned the vegetation present on it. Studies of coastal dune complex with remote sensing data were mainly done with visual interpretation of multispectral image after proper enhancements. Accurate mapping of incipient /embryo and fore dunes is possible with height information from high resolution satellite derived Digital Elevation Models (DEM). Coastal fore dunes and incipient dunes occupy a small portion of area in coastal environment therefore the identification of these individual dunes is difficult from multispectral images. To overcome the limitation of multispectral data for delineating these younger coastal dunes, this study attempts to delineate these dunes based on its plant habitats (unique/indicator vegetation species on dunes) by per pixel or sub-pixel classification. Sentinel 2A data has been radiometrically corrected and the vegetation has been extracted by Normalized Difference Vegetation Index (NDVI). Further Spectral Angle Mapper (SAM) and Linear Spectral Unmixing (LSU) have been carried out to delineate the dune habitats. Based on field validation studies, a few species such as Ipomoea pescaprae, Spinifex Littoreus and Launaea Sarmentosa are particularly identified on embryo dunes and fore dunes were further used as an indicator to delineate these dunes from satellite image.