| Based on the data of the eddy covariance observations and gradient observations at the XILINHOT National Climate Observatory in2009and2010,this paper deals with the pre-process and analysis of the data of the prairie land surface eddy correlation; With the reference of the eddy correlation method and the Bowen ratio energy balance method, This paper also compares several flux parameterization scheme which commonly used in the normal model,and analyze the bulk transfer coefficients, as well as the aerodynamic parameters variation. The main conclusions are as follows:(1) Because the terrain changes affect the measurement accuracy of the Eddy Observing System, this paper adopts the secondary coordinate rotation to correct the data. In winter, instrument icing, snow and sandstorms also affect the accuracy of measurement and cause abnormal values of high-frequency data, so the value of AGC>90must be removed. At night, when the wind is at a slow speed, the turbulent exchange will almost stagnate; under this condition, the data measured by instrument are not accurate, so the data of U<1.5m/s must be removed.(2) Two aerodynamic methods (BD and HB) and Bowen Ratio Energy Balance method(BREB) were applied to calculate the surface fluxes of turbulent momentum flux, sensible heat flux and latent heat flux. And this paper also analyzes the above fluxes with the reference of concerning results calculated by the Eddy correlation method (EC).The results showed that:The above three methods to calculate the turbulent fluxes diurnal and seasonal variation trend are almost the same; but there exists certain value bias. The value of the momentum flux are overestimated when adopting the BD method and HB method, and the deviation of non-growing season which is greater than the growing season; when using the BREB method, the sensible heat flux were overestimated while the BD method underestimated the sensible heat flux, and the growing season deviation is greater than non-growing season. The results of the HB method and the EC one are mostly closed, showing that and the deviation of growing season is less than non-growing’s; the latent heat flux is generally underestimate by BD method and BREB method, where the deviation in the growing season is higher than non-growing season. HB method is almost close to EC method when estimating the latent heat flux, and its deviation of growing season is just only-2.277W/m2. As showed above, HB method to calculate the energy flux closure rate is much higher than the others,and it has a better applicability in the grasslands underlying surface in Mongolia.(3) Bulk transfer coefficient characteristics and aerodynamic parameters showed that, bulk transfer coefficients for momentum(Cd) and heat(Ch) appear obviously the same seasonal variation and different diurnal variation. The seasonal variation was the strongest in summer, and the weakest in winter. The diurnal cycle of Cd showed single peak, while the Ch showed double peak value. With the increase of height the Cd decreased, but the Ch only decreased with the increase of height in summer. When the atmosphere was unstable, both Cd and Ch were decreased with the increase of wind speed. While in the stable atmospheric condition, Cd looks opposite against the wind speed if the flow under6m/s, then turns to along with the wind speed if the flow over6m/s. But Ch has no obvious trend with the wind speed. As the surface was colder than the air, Cd and Ch increase with the decrease of the temperature difference between the surface and the air. On the contrary, the larger the temperature difference the stronger the Cd, and the Ch was inversely proportional to the variation of temperature difference. Due to the underlying surface height of grass, the roughness appear seasonal variations that increase at first and then decrease, the value was maximum in summer (0.045m), lowest in the winter (0.003m). |
PDF Full Text Request