Inter-comparison of Long-Term Simulations of Temperature and Precipitation over China by Different Regional Climate Models
Jin-Ming FENG,Cong-Bin FU
2007-01-01
Abstract:Because the regional climate model(RCM) includes the detailed surface characteristics and physical processes,it can simulate climatic fields more exactly than the global model does.Currently,significant achievements for RCMs have been obtained in simulating mean climate state,extreme events of climate and climate changes,however there are many problems and limitations.In the past,most works focused on the case study and the sensitivity experiments.The integration period of experiments was often short,although some long-term experiments were performed in the recent years.Also most of researches on regional climate simulation are carried out by a single model,the inter-comparison among different RCMs is scarce.In order to understand the performance of RCMs in simulating Asian,especially East Asian climate,the Regional Climate Model Inter-comparison Project(RMIP) for Asia was launched by START Regional Center for Temperate East Asia,Institute of Atmospheric Physics(IAP)/Chinese Academy of Sciences(CAS),in 2000.In phase II of RMIP,six models participate for 10-year climate simulation over Asia.The objective is to assess the statistical characteristics and capabilities of the models in simulating Asian monsoon climate systemically,including mean field and variability of meteorological elements.Among six models,there are five RCMs and one global variable resolution model.The RCMs include: RIEMS,developed by START Regional Center for Temperate East Asia,IAP/CAS;NJU RCM,based on MM5v3 and improved by Nanjing University,China;MRI JSM_BAIM,a Japanese Spectral Model developed by Meteorological Research Institute/Japan Meteorological Agency;RegCM2b,based on RegCM2,improved by Central Research Institute of Electric Power Industry,Japan;SNU RCM,based on MM5v3,improved by Seoul National University,Korea.CSIRO CCAM is a global variable resolution model,developed by Commonwealth Scientific and Industrial Research Organization Atmospheric Research,Australia.These models are all integrated from December 1988 to November 1998.The experiment domain contains a large part of Asian continent,the western Pacific,the Bay of Bengal and the South China Sea.The horizontal resolution of the models is 60 km,and the driving fields come from NCEP-II reanalysis data.The inter-comparison of temperature and precipitation over China simulated by six models,as well as the validation to the observation,are presented in the paper.The observed data at 730 stations in China are used for model validation.In some areas of western China,observation stations are very sparse,thus CRU(Climatic Research Unit,University of East Anglia) data are used.Results show that almost all of the models have the capability to reproduce the spatial patterns of both temperature and precipitation over China.But a general cold bias is found in the temperature simulations,while most models overestimate the precipitation over most areas.The location and intensity of the rainbelt center have the differences!from the observations.The models usually have a good performance in simulating the interannual variability of temperature over China,as well as the precipitation in winter.However the interannual variability of summer precipitation is poorly simulated.The deviation of the large-scale driving fields and the deficiency of validation data may be some reasons for the bias of simulation.In addition,the bias of stream field is an important reason for simulation errors of RCMs.Cumulus parameterization scheme has large influence on precipitation simulation.