How Well Can a Climate Model Simulate an Extreme Precipitation Event: a Case Study using Transpose-AMIP Experiment

A high-impact extreme precipitation event over the Yangtze River valley in the mid-summer of 2016 is simulated using the atmospheric component of the Climate System Model of Chinese Academy of Meteorological Sciences (CAMS-CSM). The Transpose-Atmospheric Model Intercomparison Project (T-AMIP) type experiment is applied to investigate the model’s capability in reproducing the hourly evolution of rainfall and related synoptic circulation in the extreme event. Evaluation shows that the CAMS model forced by the historical SST well simulates the spatial distribution and diurnal variation of warm season rainfall over China, which gives the confidence to assess the rainfall processes in the extreme event. Analyses of the T-AMIP results indicate that the model realistically reproduces the heavy rainfall centers along the Yangtze River valley, but the frequency-intensity structure still shows similar biases as in the long-term run. The maximum hourly intensity is greatly underestimated in the model. The simulation of two typical heavy rainfall cases within the period of this extreme event is further compared. Comparison illustrates that the model shows different performance in cases with different precipitating systems. The zonal propagation of heavy rainfall centers during the first two days, which is related with eastward movement of low-level southwest vortex, is well reproduced. But for another case related with a smaller vortex, the observed movement of precipitating systems is not well simulated, and the model produces an artificial steady heavy rainfall center over the upwind slope of a mountain. These results imply that the simulation of extreme precipitation events is influenced by the model resolution and the related response to topography forcing.