NDVI Based Change Detection in Sundarban Mangrove Forest Using Remote Sensing Data

As a global ecosystem, Sundarban mangrove forest plays a significant role by tackling enormous CO2, and other environmental impurities from air and water. It also protects Ganges Delta's people from water surges, and cyclones. The land cover of Sundarban is degraded day by day which is threaten for mangrove plants, and animals. To detect those changes for forest mapping, feature characterization and mapping of coastal erosion, remote sensing (RS) imagery have adopted based on Satellite e.g., Landsat images. The conventional change detection (CD) approaches include post-classification comparison (PCC), unsupervised CD, change vector analysis (CVA), single threshold algorithm (STA) etc. However, the single RS image classification result may not be satisfactory using the PCC method. Tiny noise and changes of land-use or land-cover heritage cannot be recognized apparently through unsupervised CD method. CVA losses information when it uses original feature space whereas STA possesses certain difficulties to compute the complete information. Consequently, this paper proposed a CD technique to quantify the temporary changes of Sundarban RS images from 2011 to 2019 based on NDVI (Normalized Difference Vegetation Index) for making an index of Floras for indicating the volume/amount of which area is highly vegetated and whether growing of vegetation is possible or not. The outcome of this experiment indicates progressive changes for high-vegetation (0.02%), vegetation (19.22%), regressive changes for water (17.30%), and no-vegetation (1.93%).