Application of HEC-RAS and Satellite Imagery in Flood Risk Estimation for Rural Area in Thailand

The world is facing unavoidable disaster challenges, especially floods in Thailand, these inundations have been becoming increasingly difficult to manage each year. Therefore, a preliminary approach to addressing this issue is to identify areas that may experience future flooding. This study used HEC-RAS model, was calibrated by Manning's N to compare with Sentinel-1 satellite imagery and then were evaluated by Accuracy, F1 and F2 metrics to simulate and assess the affected areas for rural area in Thailand. As the results, no evidence of data overfitting was observed after the calibration between flood maps generated by HEC-RAS and satellite imagery. Moreover, the F metrics and accuracy were not exceeded by 60%, demonstrating a reasonable level of model performance. Subsequently, such model was employed to simulate flooding scenarios with high water levels 1 and 2, was found to be the identified agricultural lands as the most significantly affected area with a high likelihood of impact further. Urban areas were the second most affected. Consequently, these findings benefit the substantial impact of flooding on the livelihoods of agricultural communities and emphasize the consequent damage to household assets.