Designed and implemented a production-grade land cover classification workflow for Chittagong District using Sentinel‑2, topography, spectral indices, and Random Forest, aligned with typical environmental GIS/RS analyst responsibilities.

• Built a cloud-masked Sentinel‑2 composite using Cloud Score+, calculated key indices (NDVI, NDBI, MNDWI, BSI), and integrated JAXA ALOS elevation and slope as additional predictors.

• Trained and optimized a Random Forest classifier using manually collected ground control points, applying 10‑fold cross-validation and hyperparameter tuning to maximize accuracy across four classes (Urban, Bare Soil, Water, Vegetation).

• Applied spatial smoothing and majority filtering to reduce noise, delivering a high-quality land cover product suitable for urban planning, natural resource monitoring, disaster-risk analysis, and climate-related land-use change studies.

  GitHub: https://github.com/mohammadoney/Land-Cover-Classification-of-Chittagong