Evaluating the Efficiency of WRF-Hydro Model for Development of Flood Forecasting Systems (Case study: Kashkan Watershed)

Coupling of meteorological models with distributed hydrological models and reducing uncertainties related to downscaling is one of the major challenges in flood prediction. This study with the aim of predicting floods, demonstrates possibility of coupling the WRF model with the WRF-Hydro model, which is a hydrological-hydraulic, distributional, and physical model. For calibration, the Kashkan watershed is chosen as the study area due to its high flooding potential. According to observational flood hydrograph in study area, two severe flood events in three hydrometric stations were selected for calibration. Based on sensitivity analysis, the parameters of infiltration control, surface retention depth, surface roughness, and channel Manning’s roughness are considered as the most effective parameters. The use of these parameters will make it possible to calibrate the volume and shape of flood hydrograph to achieve accurate results. Afterward, Step by step method is used for determining the optimal values of parameters are identified in the sensitivity analysis. Poldakhtar station, located at the terminus of the basin is chosen for calibration for one event. In the next step, the calibrated model was evaluated based on the optimal parameters set for other stations and events. The evaluation results of the WRF-Hydro with optimized parameters verify the capability of the WRF-Hydro for simulating flood in the steep slope watershed after calibration. Moreover, the main characteristics of hydrograph like volume, timing and peak flow rate are simulated accusatively.