荒漠河岸胡杨林生态系统涡度相关通量数据处理与质量控制方法研究

Data processing and quality control of eddy covariance in desert riparian forest

涡度相关通量数据处理与质量控制是保证各观测台站数据质量、实现数据共享与综合集成的前提。然而,由于各个背景样地差异比较大,目前这一方面还没有统一、精确的方法可用。因此本文以黑河下游额济纳旗胡杨林生态系统8月份的涡度数据为例,介绍了各处理方法对通量结果的影响,结果表明:相比其它各校正方法而言,野点值剔除及密度效应校正(WPL)对CO2通量的影响最大,其值分别降低了11%和42%;WPL校正对潜热通量的影响最大,其值增加了6.5%;超声虚温校正对感热通量的影响最大,其值降低了9.3%。就数据质量控制与评价的方法研究,发现该样地能量平衡闭合率为71%,分别有65%和72%的数据通过了湍流的平稳性检验和湍流发展的充分性检验,数据质量较好。通过该研究,以期为荒漠河岸林生态系统涡度相关数据的处理提供参考。

The eddy covariance technique is popular in long-term fluxes measurements across the interface between the atmosphere and a plant canopy by measuring the covariance between fluctuations in vertical wind velocity and flux mixing ratio. To make cross-site comparisons and global synthesis,the necessary prerequisite should be data processing and quality control. Unfortunately,there are no uniform and accurate methods in this aspect because of the lack of uniformity among sites. Hence a new uniform and specific method in some natural regions will be required soon. The data this paper used come from eddy covariance system which is comprised of： a three-dimensional sonic anemometer,an open-path infrared gas analyzer,soil heat flux plates and other instrument used measuring temperature,air pressure and humidity. The originally measured raw data can be sampled at up to 10 Hz and calculated through specially written software which calculates the surface fluxes of momentum,sensible and latent heat and carbon dioxide. Furthermore,selected data processing software is the Edi RE software from the University of Edinburgh which is fast,flexible,and user friendly. Above all,in this paper：taking eddy covariance data of Ejina station in August,2013 as example,several steps of data processing and quality control are analyzed in detail,which highlights influences of the steps. The results show that： comparing the influences of other steps,the CO2 flux are more affected by the air density correction（WPL correction）,which decreased by 42%;WPL correction also influenced most the latent heat flux,which increased by 6.5%;sensible heat flux are more affected by the sonic virtual temperature correction,which decreased by 9.3%;every step of the processing is very important on the results except time-lag correction. Besides,surface energy balance ratio is closed to 0.71 and almost seventy percent of the data passed the test of steady state and developed turbulent conditions. The contributive highlight of the paper is not result-oriented,but a focus on the detailed analysis of processing steps which can be helpful for realistic and specific data processing methods of eddy covariance in arid and semiarid environments.