Assessing flood susceptibility with ALOS PALSAR and LiDAR digital terrain models using the height above nearest drainage (HAND) model

Addressing the escalating issue of flood susceptibility, exacerbated by climate change and urban expansion, this study presents a comprehensive evaluation of the Height Above the Nearest Drainage (HAND) model in Barreiros, Brazil—a locale recurrently affected by floods. Employing rigorous calibration techniques, the HAND model was fine-tuned using Digital Terrain Models (DTMs)—Light Detecting and Ranging (LiDAR) at 5 m and Advanced Land Observing Satellite Phased Array Type L-band Synthetic Aperture Radar (ALOS PALSAR) at 12.5 m—and validated using historical flood data. Metrics such as error percentages, Receiver Operating Characteristic (ROC), Area Under the Curve (AUC), Specificity, Sensitivity, and Precision were employed for validation. Results reveal that the LiDAR-based DTM provided superior drainage delineation, resulting in enhanced flood simulation accuracy. An overall effectiveness of 74.5% was achieved, further validated by ROC/AUC analyses. The study also highlights the model's tendency to overestimate flood risks in plains and suggests caution in its application. These findings have critical implications for flood risk assessment, urban planning, and resource allocation.