A study was conducted to classify rice growth stages in near real-time using machine learning on Google Earth Engine (GEE) for the Phulanakhara distributary of the Puri canal irrigation system. Sentinel-1 SAR imagery was used, with ground truth validation. Random Forest (RF) and Support Vector Machine (SVM) were used, with RF achieving 94.3% accuracy. The results classified early, timely, and late-transplanted rice at 808.47 ha, 2,163.71 ha, and 768.92 ha, respectively, demonstrating the effectiveness of RF algorithm for rice stage classification.
