Monitoring Water Extent and Volume Changes in the Grand Ethiopian Renaissance Dam Reservoir using Multi-Temporal Sentinel-1 SAR with Sentinel-2 Optical Validation

Monitoring the changes in the Renaissance Dam Lake isn't just a dry hydrological exercise. It seems to be vital for understanding broader regional impacts. This study utilizes observed multi-temporal Sentinel-1 SAR data from 2018 and real satellite acquisitions from July 2025 to map changes, focusing on both the reservoir's extent and water volume. The analysis was performed using two parallel workflows: SNAP 11/ArcGIS Pro and a methodologically flexible, fully scripted Python process. A Sentinel-2 optical scene from June 2025 served as an independent validation source. When examining spatial change detection using VV/VH log-ratio imagery, the expansion patterns along the reservoir margins were remarkably consistent across both platforms. The classification results, which pitted Python's XGBoost against ArcGIS's SVM, also largely converged. That said, some minor disagreements arose in the shallow shoreline areas, likely where the models struggled to distinguish between saturated soil and actual water. Results indicate the reservoir has reached a total cumulative storage of 58-61 billion cubic meters, representing approximately 78.4% to 82.4% of its ~74 design capacity (according to the Official GERD Website), with an average filling rate of 8.3 to 8.7 /year. These findings aligned quite well with the data being reported elsewhere. The study demonstrates that an operationally robust SAR workflow provides reliable monitoring in regions where ground data are sparse. Ultimately, it seems Sentinel-1 is a fairly dependable tool for monitoring large reservoirs, provided the preprocessing steps are kept consistent. It's a good reminder that the specific software platform might be less critical than the care put into the underlying data pipeline.