Uncertainties of Cumulus Convective Parameterization Schemes over East Asia in RegCM3

Cumulus convective parameterization is important for the accurate simulation of precipitation, especially in East Asian monsoon region due to the complex terrain and complicated air-sea-land interactions. A suitable cumulus convective parameterization scheme (CPS) is a key issue for the monsoon climate simulation. In this study, ensemble sensitivity experiments in 1994 summer monsoon climate (from May to August) are performed to evaluate the uncertainties of different CPSs in simulating the East Asian monsoon climate. Four CPSs (the Anthes-Kuo, Betts-Miller and Grell scheme with two closure schemes) are used and 16 ensemble experiments for each scheme are conducted with the initial fields at 6-hour intervals. The results show that RegCM3 reproduces well the observed basic climate variation patterns in monthly scale, but there exist obvious differences in simulating the climatic anomaly, particularly precipitation over East Asia, by using different CPS schemes. It is also shown that the spreads in all CPSs are small in South China, while Betts-Miller scheme and Grell with F-C closure scheme have larger spreads in North and Northwest China, implying that Anthes-Kuo and Grell-1 CPSs have good performance in simulating the East Asian monsoon rainfall. The comparison of the precipitation simulation results with the observations also shows that Anthes-Kuo scheme is more suitable for East Asia.