Comparing steady and unsteady hydraulic modeling for flood analysis with satellite imagery in Kermanshah

Floods rank among the most destructive natural disasters worldwide, with their frequency and intensity amplified by climate change. This study presents an integrated approach combining hydraulic modeling and multi-sensor satellite data to improve floodplain mapping accuracy during the April ۳, ۲۰۱۹ flood event in Kermanshah, western Iran. The research focuses on the confluence of the Gharasoo, Merek, and Razavar rivers, where combined flows created significant flood risks for Kermanshah city. The study makes several important methodological contributions to flood modeling. First, it demonstrates the value of simultaneous analysis of both steady and unsteady state conditions, providing a more comprehensive understanding of flood dynamics compared to conventional single-state approaches. Second, the integration of optical (Sentinel-۲, Landsat) and radar (Sentinel-۱) remote sensing data effectively overcomes the limitations of individual sensors, particularly in addressing cloud cover issues. Third, the implementation of Google Earth Engine enables near-real-time flood monitoring capabilities, significantly enhancing operational response potential. Finally, the development of robust validation metrics specifically adapted for flood model assessment represents an important step forward in model verification methodologies. HEC-Geo RAS simulations predicted extreme conditions with water levels rising up to ۶ meters and flow velocities reaching ۳m/s. Validation results showed strong agreement between unsteady state modeling and satellite observations (F۱=۰.۷۳, F۲=۰.۷۲), while steady-state conditions exhibited lower correlation (F۱=۰.۴۱, F۲=۰.۲۸). The model effectively tracked flood progression from inception to peak, while satellite imagery provided rapid regional coverage despite occasional cloud obstructions.

کلیدواژه ها:

Flood plain ، Google Earth Engine ، Hec Geo Ras ، Satellite Images ، water level