P1F.20
On numerical simulation of warm season precipitation diurnal cycle
Li Zhou, Department of Meteorology, University of Hawaii, Honolulu, HI; and Y. Wang and K. Souma
Realistic simulation of warm season precipitation is a great challenge to most weather and climate model models. Almost all models have systematic biases in simulating both the phase and amplitude of warm season precipitation diurnal cycle. The difficulty stems from mainly the nature of precipitation which involves convective processes and turbulent mixing and as well as the remote effects associated with atmospheric wave dynamics. In this study, the International Pacific Research Center (IPRC)'s regional climate model is used to investigate how the cumulus parameterization affects the diurnal cycle simulation and compare the performance of different parameterizations for convective transport/mixing in simulating the precipitation diurnal cycle in East Asian summer monsoon region. It is found that the current cumulus parameterization simulates too fast vertical mixing/transport while the eddy-diffusivity parameterization produces too slow vertical mixing/transport. As a result, without cumulus parameterization with a horizontal resolution of 12 km, the later afternoon peak in the precipitation diurnal cycle is about two hours too later and the amplitude is too weak, while with a massflux convective parameterization, the peak appears about 2 hours too early and the amplitude is too strong. With an eddy-diffusivity scheme to enhance the vertical mixing once the column becomes convectively unstable and with no cumulus parameterization, we are able to reduce the amplitude bias but not the phase bias. It is proposed that a combined mixing/transport scheme of an eddy-diffusivity approach and a massflux parameterization for convectively unstable conditions could be the resolution to improve the warm season precipitation diurnal cycle. Furthermore, with this framework, we will be able to allow resolution-dependent vertical mixing for the organized eddies, including the cumulus convection, which is mainly treated by the massflux approach.
* Corresponding author: Dr. Yuqing Wang, IPRC/SOEST, University of Hawaii, POST 409G, 1680 East-West Road, Honolulu, HI 96822
Poster Session 1F, Poster Session Numerical Modeling
Tuesday, 29 April 2008, 3:30 PM-5:00 PM, Palms ABCD
Previous paper Next paper