The analysis of the ability of RIEMS2.0 to simulate the two extreme climate events in the sum- mers of 1997/1998 in China.
FU Congbin
2010-01-01
Acta Meteorologica Sinica
Abstract:Regional Integrated Environment Modeling System Version 2.0(RIEMS2.0) is a regional climate model that is developed starting from RIEMS1.0 by the Key Laboratory of Regional Climate-Environment Research for Temperate East Asia,the Institute of Atmospheric Physics,Chinese Academy of Sciences,China.In order to test RIEMS2.0's ability to simulate short-term climate,we perform ensemble simulations with different physics process schemes,as well as ensemble simulations with two cumulus convective parameterization schemes(Grell and Kain-Fritsch) under different initial conditions.The model is used to perform ensemble simulations on the two continuous extreme climate events,which are the severe drought with high temperature in the northern China in the summer (June,July and August) of 1997 and severe flood in the Yangtze River valley in the summer of 1998(simulation period from 1 March 1997 to 31 August 1998).The simulated results in the summer of 1997/1998 are compared with the observed data,as well as the precipitation data from the Global Precipitation Climatology Center(GPCC) and surface(at an elevation of 2 m) air temperature data from the Climate Research Unit(CRU).It is found that RIEMS2.0 ensemble simulations can reproduce spatial distributions of the precipitation and surface air temperature,and of the differences between those for the summer of 1997 and for 1998.The simulated results are well correlated with the observed data with the correlation coefficients for surface air temperature greater than those for the precipitation. However,except for the summer of 1998 in the humid sub-regions,the precipitation is overestimated as a whole,which is on the large side in the semi-arid sub-regions and on the small side in the humid sub-regions.Meanwhile,surface air temperature is overestimated in the arid,Jianghuai and Jiangnan sub-regions,and underestimated in the semi-arid and humid sub-regions.For the ensemble simulations with different physical process schemes,there are less biases for surface air temperature in the semi-arid and humid sub-regions, and greater bias in the arid,Jianghuai and Jiangnan sub-regions.For either precipitation or surface air temperature,the ACC(anomaly correlation coefficients)/RMSE(root mean squared error) from ensembles are greater/less than the averaged one over those from individual ensemble members.Though ACCs and RMSEs from the ensembles aren't better than those from every individual ensemble member,ensembles are able to decrease the model's uncertainty and improve the simulation precision in a certain degree.There is nice consistency in the simulations on both the precipitation and surface air temperature between different explicit moisture schemes and cumulus convective parameterization schemes used,especially in the humid sub-regions.Furthermore,comparisons of the simulations of the precipitation and surface air temperature between the different physical processes employed are able to disclose the importance of choosing suitable physical processes,which is very helpful to further model development and application.The simulations can be improved via selecting suitable schemes based on the regional characteristics.As a matter of fact,by doing so,RIEMS2.0 repro- duced the spatial distribution of the precipitation and surface air temperature for these two continuous extreme climate events in the summers of 1997/1998,and disclosed the sub-regional characteristics.Though there exists some difference among the simulated results of ensemble members,the model shows good stability.The model's performance on the precipitation and surface air temperature simulation can be improved with suitable physics process schemes selected.