sami9029/Spatiotemporal-surface-water-dynamics-of-ARB-Ethiopia-using-Sentinel-1-2-and-ML-on-GEE

An automated GEE-based framework for spatiotemporal surface water mapping (2016–2025) in the Awash River Basin, Ethiopia. Integrates Sentinel-1 SAR and Sentinel-2 MSI imagery with four machine learning classifiers (RF, SVM, GTB, CART). Training and validation samples are automatically extracted from JRC-GSWD and Dynamic World datasets . # Spatiotemporal Surface Water Dynamics of the Awash River Basin, Ethiopia (2016–2025): Multi-sensor Fusion of Sentinel-1 and Sentinel-2 Data with Machine Learning Algorithms ## 📌 About This repository contains the Python implementation of an automated surface water mapping framework built on the Google Earth Engine (GEE) for the Awash River Basin (ARB), Ethiopia . The framework fuses Sentinel-1 SAR and Sentinel-2 MSI imagery with automatically collected training and validation samples from the JRC Global Surface Water Dataset (JRC-GSWD) and Dynamic World (DW), respectively. Four machine learning classifiers — RF, SVM, GTB, and CART — are comparatively evaluated. SVM achieved the best performance (OA = 98.6%, Kc = 0.971). ## Study Area Awash River Basin (ARB), Ethiopia Area 114,123 km² Period 2016 – 2025 (Winter, Spring, Summer, Autumn) ## Data Sources Sentinel-1 SAR (VH, IW) Sentinel-2 MSI JRC Global Surface Water (JRC-GSWD) Dynamic World (DW) SRTM DEM ## Classifiers Evaluated 1. Random forest (RF), 2. Support vector machine (SVM), 3. Gradient tree boost (GTB), and 4. Classification and regression tree (CART) ## 📖 Citation If you use this code, please cite: ## 👤 Authors Samuel Negussie Bekele, Mulugeta Musie Abdi Water Resources Engineering Department, Adama Science and Technology University Contact Email- saminegu6@gmail.com GitHub](github.com) ## 🙏 Acknowledgements - Google Earth Engine team for the cloud computing platform - ESA for Sentinel-1 and Sentinel-2 open-access data - JRC for the Global Surface Water Dataset - Google / WRI for the Dynamic World dataset