Basic Evaluation of Beijing Normal University Earth System Model (BNU-ESM) Version 1

Thursday, 18 December 2014
Duoying Ji1, Lanning Wang2, Jinming Feng2, Qizhong Wu2 and Huaqiong Cheng2, (1)Beijing Normal University, College of Global Change and Earth System Science, Beijing, China, (2)Beijing Normal University, Beijing, China
An earth system model has been developed at Beijing Normal University (Beijing Normal University Earth System Model, BNU-ESM). The model is based on several widely evaluated climate model components and is used to study mechanisms of ocean-atmosphere interactions, natural climate variability and carbon-climate feedbacks at interannual to interdecadal time scales. The atmospheric, oceanic, sea ice and land components are based on the Community Atmospheric Model version 3.5 from National Center for Atmospheric Research, the Modular Ocean Model version 4p1 from Geophysical Fluid Dynamics Laboratory, the Los Alamos sea ice model version 4.1 from Los Alamos National Lab, and the Common Land Model developed at Beijing Normal University, respectively. In this study, results for the CMIP5 historical simulation are evaluated to demonstrate the model’s performance in terms of the mean model state and the internal variability. It is illustrated that BNU-ESM can simulate many observed features of the earth climate system, such as the climatological annual cycle of surface air temperature and precipitation, annual cycle of tropical Pacific sea surface temperature, the overall patterns and positions of cells in global ocean meridional overturning circulation. For example, the globally averaged surface temperature has a bias of -0.17 °C compared with CRU TS 3.1 and HadISST merged datasets over the period of 1976-2005, the globally averaged precipitation has a bias of 0.18 mm/day compared with GPCP observations over the period of 1979-2005. The El Niño-Southern Oscillation (ENSO) simulated in BNU-ESM exhibits an irregular oscillation between 2 and 5 years with the seasonal phase locking feature of ENSO. In terms of terrestrial carbon cycle, simulated land gross primary production (GPP), net primary production (NPP) and net biosphere production (NEP) are consistent with several observational estimates. Important biases with regard to observations are presented and discussed, including warm SST discrepancies in the major upwelling regions, an equatorward drift of midlatitude westerly wind bands, tropical precipitation bias over the ocean that is related to the double Intertropical Convergence Zone (ITCZ) and substantial less soil organic carbon stocks simulated in the model.