Weather research and forecasting (WRF) model performance for a simulation of the 5 November 2009 heavy rainfall over southeast of Thailand

Abstract

In seasonal northeast monsoon, heavy rainfalls caused many natural disasters in the southern east-coast of Thailand. A thorough study of simulations of heavy precipitation phenomena could provide a better understanding of their nature, inducing the better forecasting of similar cases. The cumulus parameterization technique was used to determine the appropriate cumulus physics for a model. Three cumulus parameterization schemes, such as, the Kain-Frisch (KF), the Grill-Devenyi (GD), and the Betts-Miller-Janjic (BMJ) schemes were applied for the Weather Research and Forecasting (WRF-ARW version) modeling system. Their performances were evaluated by validation of the simulated rainfall by mean of calculating the categorical skill scores, such as, frequency bias (BIAS), critical success index (CSI), and equitable threat scores (ETS). The rainfall data observed from the Tropical Rainfall Measuring Mission (TRMM), and the Thai Metrological Department (TMD) were used to verify the rainfall simulation results. During the heavy rainfall events in this study, the maximum rainfalls simulated by KF, GD and BMJ schemes were recorded 146, 120, and 38 mm, respectively, while TRMM showed 187 mm of the maximum rainfall. The specific location of the intense rainfalls and their magnitude were excellently simulated in KF scheme. The positions of the maximum rainfall simulated by KF and GD schemes were also more accurate than the positions simulated by BMJ scheme. The categorical skill scores evaluation revealed that the KF scheme outperforms the GD and BMJ schemes if heavy precipitation is used as the threshold.