Land cover change detection of Hatiya Island, Bangladesh, using remote sensing techniques

Land cover change is a significant issue for environmental managers for sustainable management. Remote sensing techniques have been shown to have a high probability of recognizing land cover patterns and change detection due to periodic coverage, data integrity, and provision of data in a broad range of the electromagnetic spectrum. We evaluate the applicability of remote sensing techniques for land cover pattern recognition, as well as land cover change detection of the Hatiya Island, Bangladesh, and quantify land cover changes from 1977 to 1999. A supervised classification approach was used to classify Landsat Enhanced Thematic Mapper (ETM), Thematic Mapper (TM), and Multispectral Scanner (MSS) images into eight major land cover categories. We detected major land cover changes over the 22-year study period. During this period, marshy land, mud, mud with small grass, and bare soil had decreased by 85%, 46%, 44%, and 24%, respectively, while agricultural land, medium forest, forest, and settlement had positive changes of 26%, 45%, 363%, and 59%, respectively. The primary drivers of such landscape change were erosion and accretion processes, human pressure, and the reforestation and land reclamation programs of the Bangladesh Government. (c) 2012 Society of Photo-Optical Instrumentation Engineers (SPIE). [DOI: 10.1117/1.JRS.6.063608]