Zone II • Versions EN1
Abstract: Based on micrometeorological theory, eddy covariance technique (EC) is an approach for direct measurement of ecosystem-scale productivity, energy balance and greenhouse gas exchange. EC-generated flux data importantly support studies on ecosystem carbon & water cycling mechanism and the spatio-temporal patterns of carbon sink capacity and water use. Resulting from the first batch observations by the Chinese Flux Observation and Research Network (ChinaFLUX)founded in 2002, this dataset collected field measurements from eight sites during 2003–2005, including 4 forests, 3 grasslands and 1 cropland. The dataset follows the ChinaFLUX data processing protocols, and it provides important data basis for analyzing responses to environmental changes.
Keywords: eddy covariance technique, flux measurement, meteorological factors, carbon and water cycling, terrestrial ecosystem
|English title||Carbon and water fluxes observed by the Chinese Flux Observation and Research Network (2003 – 2005)|
|Corresponding data author||Yu Guirui (firstname.lastname@example.org)|
|Data producers||Station||Observer||Former Director||Current Director||Affiliation|
|Inner Mongolia||Hao Yanbin||Wang Yanfen||Wang Yanfen||College of Resources and Environment, University of Chinese Academy of Sciences|
|Dinghu Mountain||Zhang Deqiang||Zhou Guoyi||Zhou Guoyi||South China Botanical Garden, Chinese Academy of Sciences|
|Haibei||Zhang Fawei||Zhao Xinquan||Li Yingnian||Northwest Institute of Plateau Biology, Chinese Academy of Sciences|
|Changbai Mountain||Wu Jiabing||Han Shijie||Wang Anzhi||Institute of Applied Ecology, Chinese Academy of Sciences|
|Ouyang Zhu||Ouyang Zhu||Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences|
|Dangxiong||He Yongtao||Zhang Xianzhou||Zhang Yangjian||Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences|
|Xishuangbanna||Song Qinghai||Zhang Yiping||Lin Luxiang||Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences|
|Qianyanzhou||Dai Xiaoqin||Liu Qijing||Wang Huiming||Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences|
|Data Center||Zhang Leiming,|
|Yu Guirui||Yu Guirui||Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences|
|Time range||2003 – 2005|
|Observation scope||Eight ChinaFLUX terrestrial stations in China|
|Data amount||78.94 MB|
|Data service system||<http://www.cnern.org.cn/data/meta?id=40572>;|
|Sources of funding||National Key Research and Development Program of China (2017YFC0503801, 2016YFA0600104), Strategic Priority Research Program of the Chinese Academy of Sciences (XDA19020302), Science and Technology Service Network Initiative of the Chinese Academy of Sciences (STS Plan,KFJ-SW-STS-169).|
|Dataset composition||The dataset consists of the flux and meteorological measurement of eight ChinaFLUX stations, including Changbai Moutain, Qianyanzhou, Dinghu Mountain, Xishuangbanna, Inner Mongolia, Haibei, Yucheng and Dangxiong, each station corresponding to a subset. Each subset consists of half-hour and daily data products. The half-hour dataset of flux measurement includes CO2 flux, latent flux and sensible flux after quality control, while that of meteorological measurement includes air temperature, precipitation, global radiation and photosynthetic active radiation. The daily dataset includes integrated daily CO2 flux, latent flux and sensible flux.|
Long-term observation of terrestrial ecosystem has attracted worldwide scholarly attention. By measuring the carbon and water exchange between vegetation and atmosphere based on micrometeorological theory, eddy covariance technique (EC) realized direct monitoring of the ecological processes and functions, including productivity, energy budget and greenhouse gases (GHGs) at the ecosystem scale,1-3 and EC has been accepted as the core technical method for global flux observation network (FLUXNET).4EC-generated flux data have been widely utilized to validate different kinds of models and remote sensing observation. Moreover, the long-term coordinated observation on multi-processes and multi-factors supports analyses of ecosystem carbon-water-nitrogen cycling mechanism and spatial-temporal distributions of carbon sink/source.5-6Especially, the global-scale united flux observation breaks grounds in its extension of the observation from ecological phenomena or elements to global ecosystem functions.7
Sponsored by the Knowledge Innovation Project of the Chinese Academy of Sciences “Study on the carbon budget of terrestrial and offshore ecosystems in China”, the Chinese Terrestrial Flux Observation and Research Network was launched relying on Chinese Ecosystem Research Network (CERN) in 2001. In 2003, eight sites, including Changbai Moutain, Qianyanzhou, Dinghu Mountain, Xishuangbanna, Inner Mongolia, Haibei and Yucheng, started their coordinated measurement on ecosystem carbon and water fluxes in China,8 which filled a regional gap in the Asian Monsoon Zone. In 2014, a new ChinaFLUX was formed by combining ChinaFLUX with other flux stations in China.
As the first dataset of coordinated flux measurement in China, this dataset is composed by the ecosystem carbon and water fluxes and routine meteorological records from the most initially established eight sites. The data indicators mainly include CO2 flux, sensible flux, latent flux, friction, air temperature, soil temperature, soil moisture, precipitation, global radiation, net radiation and photosynthetic radiation at both half-hour and daily scales. This dataset was released in October 2013, and since then has received wide attention at home and abroad, but there is still a lack of systemic description of it. This is published as an endeavor of ChinaFLUX to promote data sharing and data (re)use. ChinaFLUX will further strengthen the coordination with field stations to promote long-term, multi-site data sharing while dedicated to protecting data property by means of data publishing.
2.1 Data sources
Based on top-level design and scientific appraisal, we considered the spatial distribution of the terrestrial transects and the field stations of CERN, and conducted observation system design, instrument selection, site investigation,, on-site installation and instrument debugging. Before the end of 2002, the first six sites were established, including Changbai Moutain, Qianyanzhou, Dinghu Mountain, Xishuangbanna, Haibei and Yucheng sites, while Inner Mongolia and Dangxiong sites were completed in May 2003. The eight sites constituted the first version of ChinaFLUX (Tables 1 & 2, Figure 1).
|CBS||Changbai Mountain Site||Temperate broad-leaved Korean pine mixed forest site in Changbai Mountain||128°06'E||42°24'N||Forest|
|QYZ||Qianyanzhou Site||Subtropical evergreen plantation flux site in Qianyanzhou||115°03'E||26°44'N||Forest|
|DHS||Dinghu Mountain Site||Subtropical monsoon mixed forest flux site in Dinghu Mountain||112°30'E||23°09'N||Forest|
|XSBN||Xishuangbanna Site||Tropical forest flux site in Xishuangbanna||101°16'E||21°54'N||Forest|
|NMG||Inner Mongolia Site||Temperate grassland flux site in Inner Mongolia||116°18'E||44°08'N||Grassland|
|HBGCT||Haibei Site||Alpine grassland flux site in Haibei||101°20'E||37°40'N||Grassland|
|DX||Dangxiong Site||Alpine meadow flux site in Dangxiong||91°03'E||30°29'N||Grassland|
|YC||Yucheng Site||Temperate cropland flux site in Yucheng||116°38'E||36°58'N||Cropland|
|Site code||Dominant species||Canopy height1)|
|CBS||Pinus koraiensis, Tiliaamurensis, Quercus mongolica, Fraxinus mandshurica , etc.||26||6.1||Upland dark brown forest soil|
|QYZ||Pinus massoniana, Pinus elliottii, Cunninghamia lanceolata and Schimasuperba etc.||12||3.5||Typical red earth|
|DHS||Castanopis chinensis , Schimasuperba, Cryptocarya chnensis, Pinus massoniana, etc.||17||5.6||Lateritic red-earth, yellow-earth,|
and mountain shrubby-meadow soil
|XSBN||Terminalia myriocarpa , Pometia tomentosa, Barringtoniamacrostachya, Gironnierasubaequalis, Mitrephoramaingayi, Garciniacowal, Knemaerratica, Ardisiatenera, Mezzettiopsiscreaghii, Dichmpetalum gelonioides, etc.||36||6.0||Lateritic and red lateratic soil|
|NMG||Agropyron cristatum , Stipa grandis, Cleistogenes squarrosa and Carex duriuscula, etc.||0.5||1.4||Chernozem soil|
|HBGCT||Potentilla fruticosa , Kobresia humilis, Stipa aliena, Poa orinosa, Helictotrichon tibeticum, Elymus nutans, Aster flaccidus, Polygonum viviparum, and Leontopodium nanum, etc.||0.3||2.8||Alpine meadow soil,|
alpine scrubby meadow soil,
and swamp soil
|DX||Kobresiapygmaea , Stipa capillacea, Carex montiseverestii, etc.||015||1.8||Meadow soil with sandy loam|
|YC||Winter wheat and summer maize||Winter wheat:0.9|
Summer maize: 2.5
|Winter wheat: 6.5|
|Aquox and salt aquox; surface soil is rich in light-mid loam|
Notes:1). The canopy height of grassland and cropland indicates the maximum vegetation height during the growing season; 2). LAI denotes the maximum LAI (leaf area index) during the growing season.
2.2 Data collection
The same instruments were applied in each site for measuring the ecosystem fluxes and meteorological elements. The data measurement and collection was completed automatically.
Composition and models of the instruments: The sensors, analyzers of the instruments and their manufacturers are presented in Table 3.
|Measurement system||Measured elements||Model of sensor or analyzer||Manufacturers|
|Routine meteorological system||Air temperature||HMP45C||VAISALA|
|Eddy covariance system||Sonic anemometer||CSAT3||CAMPBELL|
|CO2 and H2O densities||LI7500||LI-COR|
|Data collection and transmission||Meteorological elements||CR10X/CR23X||CAMPBELL|
Instrument installation and deployment: Considering the local terrain and vegetation height at each site, different sensors and analyzers were installed on the flux towers for continuous, real-time measurement of CO2, H2O, energy fluxes and meteorological elements (Table 4).
|Routine meteorological system||Air temperature*1||32||39||27||42||2||2||2||2|
|Eddy covariance system||CO2 and H2O fluxes||40||39||27||48.8||2||2||2||2|
|Canopy height||26.0||12.0||20.0||40.0||0.40||0.60||0.15||0.8(winter wheat)|
Notes: 1 and 2 indicate the two different heights for air temperature at one site.
Data collection and transmission: The raw data observed by Eddy covariance system and routine meteorological system were recorded at 10Hz and 30 min, respectively. Both were collected and saved with the dataloggers as indicated in Table 4. According to the local conditions, the data was downloaded through the cable (Changbai Mountain site, Qianyanzhou site, Dinghu Mountain site, Xishuangbanna site and Yucheng Site) or by using memory card (Inner Mongolia site, Haibei site and Dangxiong site) from the datalogger for subsequent quality control, standard processing and product formation.
2.3 Data processing and product formation
ChinaFLUX developed standard protocols to process the raw observed data of ecosystem carbon and water fluxes (Figure 2).
Data quality control and quality assurance: Widely-recognized methods were applied for data quality control and quality, including raw data analysis,2 sonic temperature correction,9 coordinate rotation,10 WPL correction,11 frequency loss correction,12 canopy storage estimation,13 steady state test and turbulent characteristics analysis,14 friction velocity threshold filtering,15outlier removal,16 and energy balance closure evaluation.17Among them, the double coordinate rotation was applied to grassland and cropland, while the planar fit method was applied to forest. The canopy storage of CO2, H2O and heat was estimated with the single point method, which only considered forests with high canopy.
Gap filling: Short gaps (< 2 hour) in the missing flux and meteorological data were filled through interpolation. Large gaps ( >= 2 hour) in the meteorological data were first filled with records from their adjacent meteorological field operated by CERN. Remaining gaps would be filled by the mean diurnal variation method.
Nonlinear regression method was applied to fill large gaps ( >= 2 hour) in the CO2 flux data,18-19while the Arrhenius model was utilized to fill the nighttime gaps.20 If the site suffered from drought stresses (e.g., Qianyanzhou site, Inner Mongolia site and Dangxiong site), Q10 in the Arrhenius model was modified as the linear regression of both soil temperature and soil moisture.21 The rectangular hyperbola function was adopted to fill the daytime gaps within a time window of 7 days.
For the large gaps ( >= 2 hour) in energy flux data, the marginal distribution sampling method was adopted.15
CO2 flux partitioning: The marginal distribution sampling method was utilized to partition the net CO2 flux into gross ecosystem productivity and respiration.15 First, the Arrhenius model, which was used for gap filling previously, was used to determine the coefficients of the aspiration equation based on nighttime data, so as to estimate the daytime and nighttime ecosystem respiration. After that, both the interpolated daytime CO2 flux and daytime respiration were summed to obtain the daytime ecosystem gross productivity. The nighttime gross productivity was assumed as zero due to the lack of plant photosynthesis during nighttime.
3.1 Naming scheme and data size
This dataset covers data observed from eight sites, corresponding to eight subsets that total66 excel files and 78.94 MB (Table 5). The half-hour scale data files are named in the format "AAATL30MIN_GFYY.xls" and "AAAQX30MIN_GFYY.xls", respectively, where AAA indicates site code, TL means flux, QX means meteorology,30MIN means time scale of the data, GF means gap filled data, and YY means observation year (e.g., 2003 was abbreviated as 03).
For daily scale data, a similar naming scheme is applied, except that “30MIN” was replaced by “DAILY” in these cases.
|Site code||Year||Half-hour scale||Size(MB)||Daily scale||Size(KB)|
3.2 Data items
Tables 6 and 7 present items in the flux data file and the meteorological data file at half hour scale, respectively. Table 8 presents items in the daily flux data file.
|CO2 Flux||Number||mgm-2 s-1||Corrected CO2 flux without gap filling||0.03017|
|Latent flux||Number||W m-2||Corrected latent flux without gap filling||−0.28914|
|Sensible flux||Number||W m-2||Corrected sensible flux without gap filling||−9999*|
Notes:-9999 indicates that the original data was either lost during measurement or was removed during data quality control.
|Air temperature*||Number||℃||Air temperature under canopy||−18.53|
|Air temperature*||Number||℃||Air temperature above canopy||−18.81|
|Global radiation||Number||W m-2||Mean global radiation||277.2|
|Photosynthetic radiation||Number||μmol m-2 s-1||Mean photosynthetic radiation||483.2|
Notes: For grassland and cropland, both air temperatures indicate air temperature above canopy, while the installation height for the former was lower than that of latter. Specific installation heights are listed in Table 4.
|NEE||Number||g C m-2 d-1||Daily accumulated net ecosystem exchange||-0.03325|
|GEE||Number||g C m-2 d-1||Daily accumulated gross ecosystem exchange||-0.52953|
|Re||Number||g C m-2 d-1||Daily accumulated ecosystem respiration||0.49628|
|LE||Number||Kg H2O m-2d-1||Daily accumulated latent heat flux||0.02922|
|H||Number||Kg H2O m-2d-1||Daily accumulated sensible heat flux||0.46578|
Notes: “-99999”indicates that the daily accumulation could not be accounted because of too many unfilled gaps for that day.
4.1 Data quality control
Based on generally accepted techniques and methods for global flux observation, ChinaFLUX established its standard protocols based on technical flows widely recognized by the global flux observation circle,8for field measurement, data collection, quality assurance (QA), quality control (QC), gap filling and flux partitioning to support the long-term and continuous multi-site coordinated observation (Figure 2). According to a systemic comparison among ChinaFLUX, JapanFlux and KoFlux organized by AsiaFlux, the calculated fluxes showed a great consistency.22
4.2 Data quality assessment
Using the general methods for flux data quality assurance, the spectrum analysis indicated that the power spectra of 3D wind velocity, CO2, H2O and temperature within the inertia subrange abided by the -2/3 law, and the co-spectrum of CO2, H2O, temperature with vertical wind velocity abided by the -4/3 law.23 The energy closure analysis indicated that the energy balance ratio (EBR) varied from 0.57 to 0.95, averaged 0.73,24similar to the variation range of global flux observations.17 The uncertainty analysis showed that the annual CO2 flux was mainly influenced by the threshold of friction wind speed set during the nighttime data processing. Under the impact of the uncertainty of data processing, annual carbon exchange and annual ecosystem respiration had a relative deviation of 3.88％–11.41％and 6.45％–24.91％, respectively.25
At the half hour scale, the mean valid data coverage for CO2 flux, latent heat flux and sensible heat flux was 42.6％±4.6％, 54.1％±8.8％ and 54.7％±9.0％ across the different sites and years, respectively. However, when the daytime data was concerned only, the ratio for CO2 flux increased to more than 60％–70％ (Table 9). Flux data missing could be attributed to two major reasons: one was site-specific, including power breakdown, instrument failure, maintenance or disturbance, and the other resulted from data control, including outlier removal, nighttime data filtering, of which nighttime data filtering and quality control accounted for the main reasons of the lower data validity.
|Site code||Year||CO2 flux||Latent heat flux||Sensible heat flux|
This dataset was publicized simultaneously by the synthesis center of CERN and the synthesis center of ChinaFLUX. Users can login in the Data Source and Service website (http://www.cnern.org.cn) to submit on-line requests for dataset access, or visit Science Data Bank (http://www.sciencedb.cn/dataSet/handle/600).
As there has been a lack of consensus regarding the data processing and quality control of the eddy covariance flux measurements, five points are suggested here for the usage of this dataset.
(1) Subject to a combined influence of terrain, vegetation and climate conditions, no consensus has been reached as to relevant data processing techniques or methods.
(2) In order to minimize the bias from different data processing methods, this dataset was produced following the ChinaFLUX protocols. As a result, the data presented here might deviate from those provided by respective sites.
(3) With the theoretical and technical development of data processing, ChinaFLUX will update its protocols, which may also results in data deviations.
(4) Affected by the instrument status and quality control measures, Some data are missing. Considering possible uncertainties resulting from data interpolation, model validation and modification are advised first for checking valid data coverage, where dates with less missing data should be selected. For the studies on ecosystem carbon and water exchange processes, flux data without interpolation should be selected.
(5) For the detailed data control and processing, please refer to References [3, 8 & 23].
Each field station of ChinaFLUX provided important support to ournetwork operation and data accumulation. Particularly, staff of the following stations made great contribution to on-site equipment maintenance and data collection: Changbai Moutain, Qianyanzhou, Dinghu Mountain, Xishuangbanna, Inner Mongolia, Haibei and Yucheng.
Baldocchi DD, Hicks BB&Meyers TP. Measuring biosphere-atmosphere exchanges of biologically related gases with micrometeorological methods. Ecology 69(1988): 1331-1340.
Aubinet M, VesalaT, PapaleD et al. Eddy Covariance: A Practical Guide to Measurement and Data Analysis Series. Netherlands: Springer, 2012.
Guirui Yu& Xiaomin Sun. Principals of Flux Measurement in Terrestrial Ecosystem (2nd edition). Beijing: Higher Education Press, 2018.
BaldOCchi D, Falge E, Gu L et al. FLUXNET: A New Tool to Study the Temporal and Spatial Variability of Ecosystem-Scale Carbon Dioxide, Water Vapor, and Energy Flux Densities. Bulletin of the American Meteorological Society 82(2001): 2415-2434.
BaldOCchi D. Breathing of the terrestrial biosphere: Lessons learned from a global network of carbon dioxide flux measurement systems. Australian Journal of Botany 56(2008): 1-26.
Yu GR, Chen Z, Piao SL et al. High carbon dioxide uptake by subtropical forest ecosystems in the East Asian monsoon region. Proceedings of the National Academy of Sciences of the United States of America 111(2014): 4910-4915.
YU GR, CHEN Z, ZHANG LM et al. Recognizing the scientific mission of flux tower observation networks—Lay the solid scientific data foundation for solving ecological issues related to global change. Journal of Resources and Ecology 8 (2017): 115-120.
Yu GR, Wen XF, Sun XM et al. Overview of ChinaFLUX and evaluation of its eddy covariance measurement. Agricultural and Forest Meteorology137(2006): 125-137.
SchotanusP, NieuwstadtFTM&DeBruinHAR.Temperature measurement with a sonic anemometer and its application to heat and moisture fluxes. Boundary-Layer Meteorology 26(1983): 81-93.
WilczakJM, OncleySP, StageSA.Sonic anemometertilt correction algorithms. Boundary-Layer Meteorology, 2001, 99: 127-150.
WebbEK, PearmanGI&LeuningR. Correction of the flux measurements for density effects due to heat and water vapor transfer. Quarterly Journal of the Royal Meteorological Society 106(1980): 85-100.
Moore CJ. Frequency response correction for eddy correction systems.Boundary-Layer Meteorology, 1986, 37: 17-30
Hollinger DY, Kelliher FM, Byers JN et al. Carbon dioxideexchange between an undisturbed old-growth temperate forest and the atmosphere. Ecology 75(1994): 134-150.
Foken T, Leuning R, Oncley SP et al. Corrections and data quality. In: Aubinet M et al. (eds) Eddy Covariance: A Practical Guide to Measurement and Data Analysis. Dordrecht: Springer, 2012: 85-131.
Reichstein M, Fagle E, BaldOCchi D et al. On the separation of net ecosystem exchange into assimilation and ecosystem respiration: review and improved algorithm. Global Change Biology 11(2005): 1424-1439.
Papale D&Valentini R. A new assessment of European forests carbon exchanges by eddy fluxes and artificial neural network spatialization. Global Change Biology 9(2003): 525-535.
Wilson K, Goldstein A, Falge E et al. Energy balance closure at FLUXNET sites. Agricultural and Forest Meteorology 113(2002): 223-243.
Falge E, BaldOCchi D, Olson R et al. Gap filling strategies for defensible annual sums of net ecosystem exchange. Agricultural and Forest Meteorology 107(2001a): 43-69.
Falge E, BaldOCchi D, Olson R et al. Gap filling strategies for long term energy flux data sets. Agricultural and Forest Meteorology 107(2001b): 71-77.
Lloyd J&Taylor JA. On the temperature dependence of soil respiration. Functional Ecology 8(1994): 315-323.
Reichstein M, Tenhunen JD, Roupsard O et al. Severe drought effects on ecosystem CO2 and H2O fluxes at three Mediterranean evergreen site: revision of current hypotheses? Global Change Biology 8(2002): 999-1017.
TAKAGI K, HIRATA R, WEN XF et al. Inter-comparison of eddy flux calculation and QC/QA procedures of three flux networks (ChinaFLUX, JapanFlux and KoFlux) under AsiaFlux. AsiaFlux Newsletter 26(2008): 8-11.
Wen X F, Yu G R, Sun X M et al. Turbulence flux measurement above the overstory of a subtropical Pinus plantation over the hilly region in southeastern China. Science in China Series D 48(2005): 63-73.
Li ZQ, Yu GR, Wen XF, et al. Energy balance closure at ChinaFLUX sites. Science in China Series D 48(2005): 51-62.
1. Yu G, He Y, Wang Y et al. Carbon and water fluxes observed by the Chinese Flux Observation and Research Network (2003 – 2005). Science Data Bank. DOI: 10.11922/sciencedb.600 (2018).
How to cite this article
Zhang L, Luo Y, Liu M et al. Carbon and water fluxes observed by the Chinese Flux Observation and Research Network (2003 – 2005). China Scientific Data 4(2019). DOI: 10.11922/csdata.2018.0028.zh