Integrative modeling for enhanced flood risk forecasting and management in Semi-Arid area of Iran

Abstract

Precise flood forecasting, both preceding and succeeding an event, is paramount for enacting effective management strategies to minimize potential damages. This study developed a comprehensive framework for predicting pre- and post-flood operations in a semi-arid basin in Iran by MultiModel Integration; the Weather Research and Forecasting (WRF) model, the Hydrologic Component-Hydrologic Modeling System (HC-HMS) model, and the Hydraulic Engineering Center-River Analysis System (HEC-RAS) model. The WRF model was used for pre-flood operations, while a satellite product assessed post-flood damage. Among five precipitation prediction schemes, the Lin scheme showed the highest accuracy in forecasting 48-hour precipitation, achieving a True Skill Score (TS) of 0.93. The precipitation output from the Lin scheme was then inputted into the HC-HMS hydrological model. The coupled WRF-HC-HMS model demonstrated a simulation accuracy ranging from 0.33 to 0.93, as indicated by the Nash-Sutcliffe Efficiency (NSE) criterion. The hydrological model outputs were then incorporated into the HEC-RAS hydraulic model to generate two-dimensional flood inundation maps, with simulation accuracies between 0.60 and 0.83. Finally, MODIS satellite imagery was used to estimate pre- and post-flood damage in the study area. The integrated framework provides valuable insights for water resources and flood management decision-makers, enabling them to forecast 48-hour runoff/precipitation and issue flood warnings before an event. The generated flood hazard maps can also assist in estimating the area and extent of flood-affected zones. This holistic method improves the capability to prepare for and respond to flood disasters in the semi-arid basin.