Abstract
The present study utilises semi-automatic classification of cloud-free multispectral Landsat satellite images to find changes in land use land cover (LULC) in Yamunanagar district, Haryana (India) at decadal intervals for the years 2000, 2010 and 2020. The images are classified according to the classification scheme of the National Remote Sensing Centre, India to ensure compatibility with other global/regional LULC datasets. The normalised difference vegetation index (NDVI) is applied to separate green vegetation from other LULC features. The forest was extracted from NDVI images using the forest boundary of Yamunanagar forest division. The major changes in LULC are observed in agriculture, water, forest and built-up areas. Both, fields observed and classified LULC features showed that the presence of different LULC types can be best described qualitatively and quantitatively using satellite data. The error matrix shows an overall 88.69% accuracy is achieved in LULC features obtained from image classification. The Kappa coefficient calculated is 0.9 which is rated as almost perfect.
Get full access to this article
View all access options for this article.