Land use / land cover changes monitored by NDVI index in Rangamati, Bangladesh for the last four decades

Abstract

The study of land use/land cover dynamics has been increasingly important in the research of earth surface natural resources. The normalized difference vegetation index (NDVI) is a widely used method for observing land use/land cover change detection. The surface land resources are easily interpreted by computing their NDVI. This study aimed at analyzing Land Use/Land Cover (LULC) changes between 1977 and 2019 in the Rangamati district, Bangladesh using reclassify the NDVI values of the Landsat satellite image and identifying the main drivers to change LULC by household survey. Five different years of Landsat images were used to extract the NDVI values January of 1977, 1989, 2000, 2011 and 2019. The NDVI values are initially computed using the user define method to reclassify the NDVI map to create land use land cover map and change detection. The highest NDVI value was found in 1977 (0.88) which indicates healthy vegetation at that time and thereafter it followed a decreasing trend (0.79 in 1989, 0.74 in 2000, 0.71 in 2011 and 0.53 in 2019) which shows a rapid vegetation cover change in the study area. Analysis of the household survey revealed that population growth, migration from plain land, rapidly urbanization, Kaptai Dam, migration policy of government, high land price, unplanned development, development of tourism industry, firewood collection and poverty have been identified as the major drivers of LULC changes in the study area. Furthermore, analysis of NDVI confirms that the forest vegetation area is being decreased and settlement area and sparseness of vegetation are being increased. The accuracy of the NDVI-based classified images is assessed, using a confusion matrix where overall classification accuracy and Kappa coefficient are computed. The overall classification accuracy was 84% - 90% with corresponding Kappa statistics of 80% - 88% for TM and OLI-TIRS images, respectively. The study serves as a basis of understanding of the LULC changes in the southeastern part of Bangladesh.

Bangladesh J. Agri. 2019-2021, 44-46: 127-140