APPLICATION OF DIFFERENT SATELLITE IMAGE CLASSIFICATION TECHNIQUES FOR MAPPING LAND USE AND LAND COVER OF EAST GODAVARI RIVER ESTUARINE LANDSCAPE, ANDHRA PRADESH, INDIA
Coastal mangroves are highly diverse, and productive. Mangroves in India are under severe pressure from developmental activities, causing conversion of mangroves to other land use types. Hence, mapping the spatial extent of the coastal habitats is essential for long term conservation and management of mangroves especially in the East Godavari River Estuarine Ecosystem (EGREE), which is the second largest congregation of mangroves in India. In this context, land use mapping of the EGREE landscape was attempted using visual interpretation, supervised and unsupervised classification of Landsat 8 OLI satellite data to find the most suitable method in terms of effort and accuracy. Supervised classification was attempted in both conventional method and a modified method. Eleven land use/land cover (LULC) types viz. mangrove, agriculture, aquaculture, casuarina, coconut plantations, mud flats, open land, salt pans, sand bar and water bodies were delineated using these three classification techniques. Five additional LULC classes- industrial area, port, wetlands, Prosopis spp dominated habitat and riverine forests were also mapped using on-screen visual interpretation method. Among all the classification techniques used, visual interpretation method could yield highest accuracy (98.69%), followed by the modified version of supervised classification (91.36%). Conventional supervised and unsupervised classification gave accuracies of 78.53% and 73.3%, respectively. It was found that the mangroves cover 193 km2 area in EGREE landscape. The LULC map generated in this study can act as a baseline for monitoring this dynamic coastal landscape.
- There are currently no refbacks.