Investigating the causes of different performance of WRF model in forecasting rainfall of different weather systems from a Dynamic Perspective: A case study

One of the main challenges in numerical weather prediction models is the correct rainfall forecasting, especially for springtime convective precipitation. The dynamical recognition of the "rainfall factors" leads us to select "more suitable schemes" for more accurate rainfall prediction. The WRF model has a wide range of physical parameterization schemes. The choice of each group of these schemes can change the model outputs, significantly. But sometimes, there is not enough accuracy in the prediction even using different schemes. Therefore, it is necessary to select the appropriate schemes considering the factors causing precipitation. Here, to investigate the effect of dynamical factors causing the atmospheric instability and precipitation on the appropriate physical scheme selection, two atmospheric systems for springtime precipitation with different dynamic factors have been selected. In the first case, a false forecast of a heavy rainfall leading to floods in the central regions of the Iran occurred, and in the second case, the heavy rainfall in these areas did not forecast correctly; wrong warnings of these false predictions led to management problems for the related decision makers. In this research, initially, the two systems were studied from the dynamical perspective and then, they were simulated using the WRF mesoscale model with nine different configurations. The ERA-Interim was used as the model input data for the region of interest (33°N-36°N, 48°E-54°E). The simulations were done in two nested domains with a horizontal resolution of 45 and 15 km for the first and second domain, respectively. The proper configuration was selected for each atmospheric system according to the dynamical factors and comparing the model outputs with the observations (from some stations in Tehran, Qom, Markazi, Hamedan, and the west of Isfahan provinces). The results show that the type of convection scheme has a significant impact on the prediction of rainfall amount and this scheme extremely depends on the instability factor, which initiates from "the upper and the lower atmospheric levels conditions". Based on the results, for the first case, which its main cause of precipitation is due to the potential vorticity (PV) streamer penetration to the upper level of the troposphere, the Tiedtke convective scheme and the Kessler microphysics scheme are better than the other schemes that overestimated rainfall. In the second case, the occurrence of severe rainfall with the main mechanism of daily convection was not predicted by eight configurations. But in the ninth configuration, the Grell-Freitas convection scheme could successfully predict the convective heavy precipitation due to the better capture of daytime convection over the area of study rather than other used convective schemes.