Flood Hazard Assessment of Transportation Infrastructure Using Remote Sensing and Machine Learning

Climate change is causing more intense and frequent extreme weather events, such as heavy rain, which can damage vital transportation infrastructure like roads, railways, and bridges. Therefore, it is essential to design and construct transportation infrastructure that can withstand current and future climate-driven flood frequency and amplitude increases. Flood detection using remote sensing approaches like Synthetic Aperture Radar (SAR) can help in monitoring natural hazards such as floods. In this study, the authors investigate the use of Sentinel-1 SAR imagery as a powerful tool for flood detection and monitoring transportation infrastructure. The study combines InSAR coherence and RGB composition methods to visualize multi-temporal changes and detect changes on the terrain surface via a temporal color image. The results were used to monitor Little River Watershed (LRW) flood events in central Oklahoma, and a Support Vector Machine classifier was used to extract flooded areas in the false RGB images. Overall, the proposed method showed promising results in accurately detecting flooded areas and can be useful for monitoring natural hazards and transportation infrastructure.