Monitoring change in forest cover in Karnataka, India using Landsat data

This research evaluates the potential for forest change monitoring in the state of Karnataka, India using Landsat data. Landsat imagery from 1986 and 2003 was analyzed using two different change detection techniques: image differencing applied to the Principal components of the data, the Normalized Difference Vegetation Index (NDVI), and Kauth-Thomas transformed data, and post-classification comparison of binary change/no-change maps. The study demonstrates how data acquired from Landsat provide an invaluable resource for monitoring forest cover dynamics using these change analysis methods. The NDVI image differencing method was slightly better in detecting vegetation loss than PC 2 and KT-G; however, post classification comparison performed better in detecting changes as compared to NDVI differencing. Analysis of vegetation change maps revealed links of deforestation and reforestation to land use and cultural practices in the Karnataka region. The most significant causes of deforestation were submergence of forests by reservoirs in the construction of hydroelectric dams and organized smuggling of valuable timber such as sandalwood, whereas reforestation was mainly due to ongoing forest plantation projects.