CLOUD RESOLVABLE REGIONAL CLIMATE SIMULATION OVER THE YANGTZE RIVER DELTA IN SUMMER DURING 2003

The regional, 3 km-resolution climate model WRF was used to simulate the extreme regional climate in the Yangtze River Delta in the summer of 2003, and compared with the station observations and TRMM precipitation data. It shows that the WRF model reasonably simulates the main features of precipitation in the Yangtze River Delta during that time, and the simulated monthly average precipitation, the central area of heavy rainfall and precipitation intensity are all close to the observation. The model also reasonably simulates the temporal evolution of many heavy rain events in the Yangtze River Delta during the Meiyu season, with very high time correlation coefficients between the simulated and observed time series of regional average daily precipitation in Yangtze-Huaihe Rivers basin, Yangtze River basin and southern Zhejiang area. Meanwhile, the WRF model reasonably simulates the probability characteristics of light, moderate and heavy rain, the three different categories of daily precipitation, and the simulated probability distribution of heavy rain is closest to the observation. In addition, the model also reasonably simulates many heavy rain events during the Meiyu season in the Yangtze River Delta, and the occurrence time and area of simulated heavy rainfall and the southward and northward shifts of rain bands are very close to the observation, but with a little bias in precipitation intensity. It reasonably represents the regional climate characteristics of high-temperature weather and significant temperature bias between the south and the north of Yangtze River, which are caused by the cooling effect of many heavy rainfall events during Meiyu season in the Yangtze River Delta in 2003. The distribution and largest-value zones of simulated min. and max. temperature are both close to the observation, and the simulated min. temperature is closer to the observation than the max. one. The simulated time evolutions of daily max. and min. temperature are closer to the observation than the simulated precipitation, with higher time correlation coefficients.