Measurements and parameterization of near-surface turbulent fluxes over a grassland site in rural Beijing

Climate simulations are sensitive to the partitioning of surface energy fluxes and their parameterizations in models. A field experiment was conducted on the surface turbulent fluxes of momentum, moisture, heat, and carbon dioxide over a grass (0.65 m in height) site in rural Beijing from 19 July through 6 August 2010. The zero displacement (d) and the roughness length (z0), were estimated to be 0.44 m and 0.02 m respectively, and the footprint flux distribution analysis shows that the maximum source weight location is 157 m upwind from the mast. The diurnal variations of net radiation, latent heat, sensible heat and soil heat flux as well as carbon dioxide were quantified. The average energy closure ratio was determined to be 0.89 for the clear days during the experimental period and the surface albedo was 0.13. Three iterative schemes, and one recently improved non-iterative scheme, for modeling the fluxes were tested by using the measured friction velocity, temperature scale and sensible heat flux obtained from the field experiments. Our findings suggest that under unstable and weak stable stratification, Högström scheme (for unstable stratification), incorporated with Beljaars and Holtslag scheme (for stable stratification), and the non-iterative scheme performed better, whereas all four schemes gave unaccepted errors under strong stable condition. When the ratio of aerodynamic roughness length to temperature roughness length z0/z0h=100, the simulated sensible heat flux was found to be closer to measurements than z0/ z0h=10.