Simulations of a rainfall process in Southeastern China R with WRF and MM5

In an attempt to understand the relative strengths of two state-of-art mesoscale models: WRF (Weather Research and Forecasting Model) and Penn State/NCAR MM5, a nested domain was constructed to simulate a precipitation process induced by land-sea wind shear in Southeastern China with the two models. The results indicate that both WRF and MM5 capture the local circulation and the heavy rainfall area well, but precipitation amounts are overestimated relative to observations. The WRF model is better than the MM5 in precipitation strength, but worse in the location of maximum rainfall. In comparison with the observations, the simulation by both MM5 and WRF is larger than observation, but the simulation of WRF is closer to observation than MM5. Except that the distribution tendency of simulated wind field by MM5 is better than by WRF, the distribution trend of 2-m temperature and precipitation simulated by WRF is closer to observation than by MM5, but there exists phase difference between simulation by WRF and observation, for precipitation with lagged phase, and for temperature with phase ahead.