Revealing Earth’s Blueprint: Influence of Global Surface Water Probability in Flood Classification Using Synthetic Aperture Radar

Accurate prediction and classification of natural disasters are crucial but are often hampered by the unavailability of timely and accurate data, especially in challenging weather conditions. Flood prediction and classification, in particular, suffers from such limitations. However, synthetic aperture radar images play a pivotal role in overcoming these challenges due to their ability to capture data in all weather conditions. In this study, we utilized synthetic aperture radar images along with global water probability data for flood classification. We used only 0.6% of the dataset for training, while the remaining 99.4% was allocated for testing. Our analysis reveals a notable improvement in flood classification accuracy, rising from 0.799 without global water probability to 0.8427 when the water probability feature was included. This underscores the crucial role of global surface water probability in improving flood classification accuracy, emphasizing its value in satellite-based flood disaster management and response strategies.