Identifying the Optimal Numerical Scheme within the WRF Model for Precipitations Leading to Floods and Its Uncertainties

Providing an accurate depiction of precipitation structure and ensuring precise and dependable forecasts of precipitation are crucial, yet considerably challenging, elements for hydrological and hydraulic simulations. Moreover, the selection of schemes within the WRF numerical system significantly influences forecasting outcomes. This research primarily aims to assess the sensitivity of WRF numerical system to the physical parameterization of convection, boundary layer, and microphysics segments. It seeks to identify the optimal configuration for the study area to accurately simulate heavy rainfall events leading to floods and to evaluate the model's performance in predicting rainfall at 6, 12, and 24-hour time steps. The WRF model is established using two nested domains with a scaling ratio of 1 to 3. Findings indicated that alterations in the convection and boundary layer configurations result in the most substantial differences between simulated and observed precipitation. Hence, selecting appropriate physical parameterization for convection and boundary layer outweighs the significance of microphysics. Evaluation across various time steps suggested that the model's daily predictions can be utilized with higher confidence levels. Additionally, for the Poldokhtar basin, the combination of WDM5, Grell 3D, and Asymmetrical Convective Model v2 schemes respectively for microphysics, convection, and boundary layer sections, exhibited superior performance in precipitation simulation. Based on the results, this configuration is advisable for future studies aiming to simulate and predict heavy rainfall-induced floods.