Geospatial Analysis of Land Cover Maps: A Case Study in Mumbai Metropolitan Region, India

Land Use Land Cover (LULC) classification is crucial for effective resource management, urban planning and environmental monitoring. Remote sensing based satellite imagery in association with machine learning has great potential for LULC classification. This study suggests a hybrid machine learning model that combines the ensemble learning and feature fusion approach of Support Vector Machine (SVM), Random Forest (RF) and Gradient Tree Boosting (GTB) as base learner and bagging based RF as meta-classifier on Sentinel 1 - VH and VV polarization band satellite imagery for the year 2024. Experimental results show that the hybrid model outperforms standalone models in terms of overall accuracy. Additionally, feature importance scores are computed. The generated LULC map is supported with the ESA world cover map. This proposed study is highly beneficial to the policy makers.