Machine learning based land use land cover analysis and prediction in Google Earth Engine towards achieving sustainable development in Indian Western Himalayan region

Abstract The Indian Western Himalayan (IWH) region experiences persistent threats from natural hazards mainly influenced by land use and land cover (LULC) changes resulting from rapid urbanization. The present study aims to perform machine learning-based LULC classification for the years 2016, 2019, and 2022 and the prediction of future LULC for 2035 using Multi-layer Perceptron Artificial Neural Network-Markov Chain model (MLP-ANN-MCM). LULC map was prepared using Sentinel-2 imagery in Google Earth Engine (GEE), and eight LULC classes were identified with dominance of dense vegetation, scrubland, and barren land. Rapid urban growth and declines in natural habitats were observed from the LULC transition analysis. The kappa statistics validate the accuracy of the model for LULC prediction. Predicted LULC for 2035 indicates an increment in built-up and agricultural areas. The study promotes sustainable development and resilience by emphasizing effective land use planning.