Evaluation of WRF Model for Simulation of Precipitation and Flood Forecasting in Karun 4 Basin

In this study, hourly values ​​of precipitation and air temperature in Karoon 4 basin in southwestern of Iran were simulated with WRF numerical model to evaluate the accuracy of the model for flood prediction. The flood event was selected in March 2016. For global boundary conditions of the model, global data with a resolution of 0.75 were used. The WRF model was implemented using eight different configurations, including a convective schema, two planetary boundary layer schemes, two microphysical schemes, a surface layer schema, and two shortwave radiation schemes to obtain a suitable configuration for simulating temperature and precipitation. The results showed that in the simulation of hourly precipitation, the combination of MYJ boundary layer design with other micro-physics and short-ray projections yields better results than YSU schema. The best values of IOA (matching coefficient) between observed and estimated precipitation of the model was 0.77, 0.76, 0.74 and 0.52 in Shahrekord, Saman, Koohrang and Lordegan, respectively, by combining MYJ boundary layer, Lin physics and Dudhia short radiation. While in simulating hourly air temperature, the YSU boundary layer schema combination with other schema showed better performance. So that with this combination, the heighest conformity coefficient (0.47) was obtained between the estimated and observed hourly temperature. The estimated rainfall adjusted by MYJ boundary layer configuration, Lin physics, GODDARD short radiation and MYJLG has performed better prediction for peak dischage than the other schemas, so that the lowest Nash coefficient and RMSE were -0.293 and 37 respectively, with this combination. Therefore, the combination of MYJ boundary layer schema, Lin cloud microphysics schema, and GODDARD radiation schema appear to be the best for estimation of precipitation and temperature and consequently for prediction of the March 2016 flood in the Karun 4 basin.