Analysis of CGCM control runs from CMIP-II

Coupled Ocean Atmosphere GCMs are a vital tool to study climate change and the simulations by such models are used in evaluating future climate scenarios. The Intergovernmental Panel on Climate Change (IPCC) Second Assessment Report (Houghton, 1995) summarized in its evaluation of climate models that "the need for flux adjustment is expected to diminish as the component atmospheric, oceanic, land-surface and sea ice models are improved". Since then, there have been many improvements in the models e.g. increased resolution, higher computational speeds resulting in longer and more simulations, and improved parameterizations. A significant change has been the increased number of models that do not employ any flux adjustment schemes.

Control run simulations with constant CO2 from 12 different coupled Ocean Atmosphere GCMs (with sea-ice) participating in the Coupled Model Intercomparison Project (CMIP II) are analyzed. Half of the models studied do not employ any "flux adjustment", while the rest employ some flux adjustment scheme. Monthly mean data for surface air temperature, sea level pressure, and precipitation are compared with the NCEP and ERA15 reanalyses to quantify errors. Flux-adjusted and non flux-adjusted models are investigated for any systematic differences in the simulations. The model data are analyzed over different regions of interest and over seasonal and inter-annual time scales to isolate strengths and weaknesses of models in simulating present-day climate. We find that Coupled GCMs simulate the climatology of surface air temperature better than that of precipitation, with an intermediate skill in sea level pressure simulation. The flux corrected models seem to perform better albeit marginally.