This research develops a lightweight, physics-motivated elevation-based framework for stratifying monsoon flood risk in Rohingya refugee camps. Using eight years of flood records, it validates a 3 m elevation threshold, achieves 90.2% temporal validation accuracy, and identifies 11 high-risk camps containing 276,510 people. The framework requires no model training, GPU, or internet connection, offering a practical and low-cost tool for humanitarian preparedness and pre-positioning of aid.
