Geospatial Land Use/Land Cover (LULC) classification of the upper Wainganga Basin, Madhya Pradesh

Abstract

This study investigates the Land Use/Land Cover (LULC) dynamics of the Upper Wainganga sub-basin (8425.72 km²) in Madhya Pradesh, India, using multi-temporal satellite imagery of 2024. Data from Sentinel-2B satellites were processed in ERDAS Imagine 2014 and QGIS 3.38.0. An unsupervised classification based on the maximum likelihood algorithm identified five major land cover categories: forest (45.93%), agriculture/other vegetation (32.01%), fallow/open/barren (15.81%), built-up (4.87%), and water bodies (1.36%). Accuracy assessment using 500 reference points yielded an overall accuracy of 71.6% and a Kappa coefficient of 0.61, indicating substantial agreement. Forest, water bodies, and built-up areas showed high classification reliability, while confusion between agricultural and fallow lands occurred due to spectral overlap. The results demonstrate the effectiveness of remote sensing and GIS techniques in mapping of LULC, providing valuable insights for sustainable land-use planning and resource management in the basin.