Kirschner Steppe Land - Atmosphere Flux Exchange Characteristics Of Study

Grassland ecosystem is one of the most important terrestrial ecosystems. There are richgrassland resources in China, and it is necessary to study matter and energy exchangesbetween grassland ecosystem and atmosphere for understanding interaction between theunderlying surface and the atmosphere and global change.This paper presents the results of one year (2004) continuous measurements of surfacefluxes above stipa krylovii steppe in Inner Mongolia using eddy covariance method (EC),moreover, variational technique(VT) is used to estimate surface fluxes of sensible and latentheats in terms of the gradient observations including wind speed, air temperature andhumidity, respectively, at two heights (2m and 4m) by microclimatic gradient system, andcompared with EC and Bowen ratio energy balance methods (BREB).The main results are asfollows:1) The diurnal amplitude of NEE varied substantially within the growing season. CO2uptake (negative value denotes uptake) by the canopy is stronger around noon than anytime ina day along with photosynthetically active radiation (PAR).In the daytime, the fluxes ofsensible heat (H) and latent heat (LE) are mostly positive away from the surface, while atnight the fluxes of sensible heat are negative toward the surface and the fluxes of latent heatare nearly zero. The time on which the maximum surface fluxes appear is consistent with thetime of the maximum PAR.2) The monthly average diurnal courses of NEE show converse campanulate plots fromAugust to October. The CO2 uptake of September is the highest because of the plentifulrainfall, and August takes the second place, and October lowest. In addition, the sensible heatflux of May and the latent heat flux of September are at the top, respectively.3) By analyzing the results of the whole year 2004, the conclusions can be drawn: inJanuary, February and December, the cold weather of steppe restrains the metabolism of theedaphon, resulting in weak carbon dioxide flux. The growing season starts in late April andends in late October. It is obvious that the period from mid-August to late September is acarbon sink. The latent fluxes dominate in late August, September, early October, and earlyDecember, and the sensible fluxes dominate in the remainder months.4) During the growing season, the relativity between the daily sum of NEE and thefluxes of water and heat is not large, however, the change trends of the data points show thatthe absolute values of NEE increase along with the increase of the latent heat fluxes and NEEare translated from a majority of negative values into a majority of positive values along withthe increase of the sensible heat fluxes, namely, from carbon uptake to carbon release.5) VT used full information provided by boundary layer observation, the surface energybudget, and Monin-obukhov similarity theory. Compared with observed fluxes from ECmethod corrected by streamline coordinate transforming and WPL density effects, the changesof sensible and latent heat fluxes by VT method are consistent with those from EC method.The fluxes from VT method satisfied the surface energy balance equation better than thosefrom EC method. However, the sensible heat fluxes from VT method are mostly higher thanthe corrected fluxes from EC method in the daytime, and the variations of the latent heatfluxes from VT method lagged behind them from EC method. The momentum fluxes by VTand EC both have many apices, and smaller at dawn and dusk than the rest of a day, whichacts in accord with the diurnal variation of the wind speed in grassland. Moreover, VT methodmay solve the problems that the conventional BREB method becomes computationallyunstable and produces spurious large values when Bowen ratio is in the vicinity of -1.Compared with EC method and BREB method, VT method is of rationality and stability.