Intercomparison Of Three Near Sea Surface Turbulent Flux Parameterization Scheme In Community Atmosphere Model Version 3(NCAR CAM3)

To improve the ability of the general circulation models in climate simulation, it is especially important to better parameterize the physical processes in numerical models. In this paper, a new developed turbulent flux parameterize scheme is incorporated into NCAR CAM3.0. This scheme takes account of the difference between aerodynamic roughness length and the heat roughness length, and has the features of simple formula and no iterative loop needed. Meanwhile, the COARE scheme, as a reference, is also introduced into the model to test simulation ability of the new scheme, because COARE algorithm has been widely used in recent years.The climatic mean states of precipitation, latent heat flux, sensible heat flux, sea level pressure, and wind stress are compared against the reanalysis data NCEP. The results show:①The simulated magnitude of precipitation is improved in GCAM and CCAM, especially more realistic monsoon precipitation. There is, however, a tendency for the tropical precipitation maxima to remain in the Northern Hemisphere throughout year in both OCAM and GCAM, and this is in sharp contrast with most observational datasets, which shows a clear seasonal migration of ITCZ precipitation across the equator.②The latent heat flux simulated by OCAM produce positive difference in warm current and negative difference in cold current compare with NCEP data. This bias had been reduced in GCAM. To the zonally averaged latent heat flux, GCAM is much better agreement with the observation than OCAM, except in high latitude.③Although only the near sea surface turbulent flux parameterization scheme had been revised in GCAM, it have an indirect positive influence on the mainland. The magnitudes of zonally averaged sensible heat fluxes modeled by three schemes are closed to each other, and there is not obvious improvement in GCAM. The possible reason is that latent heat flux is the main consumer of surface available energy, and sensible heat flux is relatively small. The largest regional differences which occur in polar region might not be tested by reanalysis data NCEP because of the data's uncertainty. This biase needs to be verified by further observation.④GCAM have produced a modest improvement in the simulated sea level pressure field. It should be pointed out that GCAM produce a little higher biase in some region during the boreal summer season.⑤As for the sea surface wind stress, GCAM reduce the bias which is obvious in OCAM. ⑥The compute efficiency have been improved in GCAM.GCAM have improved the simulated climate, including more realistic precipitation, latent heat flux, sensible heat flux, sea surface wind stress and sea level pressure, meanwhile, GCAM cut down the calculate time. To introduce the Gao scheme into CAM3.0 is feasible.