A long-term (2005-2016) dataset of hourly integrated land-atmosphere interaction observations on the Tibetan Plateau

. The Tibetan Plateau (TP) plays a critical role in influencing regional and global climate, via both thermal and dynamical mechanisms. Meanwhile, as the largest high-elevation part of the cryosphere outside the polar regions, with vast 20 areas of mountain glaciers, permafrost and seasonally frozen ground, the TP is characterized as an area sensitive to global climate change. However, meteorological stations are sparsely and biased distributed over the TP, owing to the harsh environmental conditions, high elevations, complex topography, and heterogeneous surfaces. Moreover, due to the weak representation of the stations, atmospheric conditions and the local land-atmosphere coupled system over the TP as well as its effects on surrounding regions are poorly quantified. This paper presents a long-term (2005-2016) in-situ observational dataset 25 of hourly land-atmosphere interaction observations from an integrated high-elevation and cold region observation network, composed of six field stations on the TP. These in-situ observations contain both meteorological and micrometeorological measurements including gradient meteorology, surface radiation, eddy covariance (EC), soil temperature and soil water content profiles. Meteorological data were monitored by automatic weather stations (AWS) or planetary boundary layer (PBL) observation systems. Multilayer soil temperature and moisture were recorded to capture vertical hydrothermal variations and 30 the soil freeze-thaw process. In addition, an EC system consisting of an ultrasonic anemometer and an infrared gas analyzer was installed at each station to capture the high-frequency vertical exchanges of energy, momentum, water vapor and carbon dioxide within the atmospheric boundary layer. The release of these continuous and long-term datasets with hourly resolution represents a leap forward in scientific data sharing across the TP, and it has been partially used in the past to assist in understanding key land surface processes. This dataset is described here comprehensively for facilitating a broader 35 multidisciplinary community by enabling the evaluation and development of existing or new remote sensing algorithms as 2 well as geophysical models for climate research and forecasting. The whole datasets are freely available at Science Data Bank (http://www.dx.doi.org/10.11922/sciencedb.00103, Ma et al., 2020) and, additionally at the National Tibetan Plateau Data Center (https://data.tpdc.ac.cn/en/data/b9ab35b2-81fb-4330-925f-4d9860ac47c3/).

supplementary material, include these instead as an appendix to main text? Response: Yes, the data availability figures are very useful as they provide an intuitive depiction of the availability of each variable, and the purpose we provide the heat maps was to facilitate data selection. Thanks very much for your suggestion, these figures have been included to main text as an appendix. 5) Manuscript still needs an overall uncertainty table. E.g. Table 2 but with uncertainties (sensor-specific, site-specific, operation-specific) for each measurement at each site. Greater than 95% operation of most sensors at most heights in 2009 at NAMORS does not mean those data are useful! Authors provide only presence / absence while users need performance uncertainties. Cautions about using (or not using) raw data are not sufficient. Response: Yes, the high value of annual operation of sensors only indicate the presence/absence of raw data but can not clearly show the accuracy and useful of the observations. We do understand your concerns and needs on the observations, and we highly value the comment you posted. Based on your suggestion, we have added the accuracy of each sensor-specific measurement at each site in Table 2. However, as the data quality flag of each measurement is not available for the moment, obtaining the overall uncertainty is full of challenge and difficulties. But, the first priority of our near and long-term task is to provide a continuous dataset with strict quality control procedure, bias correction and gap filling procedures applied to ensure the accuracy and reliability of the observations. Meanwhile, the data quality flag and overall uncertainty of each measurement will be provided. 6) Line 347: "missidng" ?? Response: We are sorry for the mistake. The word should be "missing" and this error has been revised.
Correspondence to: Yaoming Ma (ymma@itpcas.ac.cn), Zhipeng Xie (zp_xie@itpcas.ac.cn) and Binbin Wang (wangbinbin@itpcas.ac.cn) Abstract. The Tibetan Plateau (TP) plays a critical role in influencing regional and global climate, via both thermal and dynamical mechanisms. Meanwhile, as the largest high-elevation part of the cryosphere outside the polar regions, with vast 20 areas of mountain glaciers, permafrost and seasonally frozen ground, the TP is characterized as an area sensitive to global climate change. However, meteorological stations are sparsely and biased distributed over the TP, owing to the harsh environmental conditions, high elevations, complex topography, and heterogeneous surfaces. Moreover, due to the weak representation of the stations, atmospheric conditions and the local land-atmosphere coupled system over the TP as well as its effects on surrounding regions are poorly quantified. This paper presents a long-term (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016) in-situ observational dataset 25 of hourly land-atmosphere interaction observations from an integrated high-elevation and cold region observation network, composed of six field stations on the TP. These in-situ observations contain both meteorological and micrometeorological measurements including gradient meteorology, surface radiation, eddy covariance (EC), soil temperature and soil water content profiles. Meteorological data were monitored by automatic weather stations (AWS) or planetary boundary layer (PBL) observation systems. Multilayer soil temperature and moisture were recorded to capture vertical hydrothermal variations and 30 the soil freeze-thaw process. In addition, an EC system consisting of an ultrasonic anemometer and an infrared gas analyzer was installed at each station to capture the high-frequency vertical exchanges of energy, momentum, water vapor and carbon dioxide within the atmospheric boundary layer. The release of these continuous and long-term datasets with hourly resolution represents a leap forward in scientific data sharing across the TP, and it has been partially used in the past to assist in understanding key land surface processes. This dataset is described here comprehensively for facilitating a broader 35 multidisciplinary community by enabling the evaluation and development of existing or new remote sensing algorithms as Deleted: well as geophysical models for climate research and forecasting. The whole datasets are freely available at Science Data Bank (http://www.dx.doi.org/10.11922/sciencedb.00103, Ma et al., 2020) and, additionally at the National Tibetan Plateau Data Center (https://data.tpdc.ac.cn/en/data/b9ab35b2-81fb-4330-925f-4d9860ac47c3/). 40

Introduction
The Tibetan Plateau (TP) is the world's highest and largest plateau with highly complex terrain and is referred as the "Third Pole of the World" (Qiu,2008). Moreover, the TP has the most extensive high-elevation distribution of cryosphere outside the polar regions. There are vast areas of mountain glaciers, snow, permafrost and seasonally frozen earth across the TP (Zhou and Guo, 1982;Kang et al., 2010;Cheng and Jin, 2013). Therefore, it also acts as the "Water Tower of Asia" 45 (Immerzeel et al., 2010). Numerous researches indicate that the TP plays an essential role in controlling regional and global climate through its thermal and mechanical mechanisms (Manabe and Broccoli, 1990;Yanai et al., 1992;Duan and Wu, 2005;Liu et al., 2007). It exerts a major control on atmospheric circulation at the local and continental scale (Ding, 1992;Ye and Wu, 1998;Li et al., 2018) through its latent heat release (Wu et al., 2016) and interactions between the Asian monsoon and mid-latitude westerlies (Yao et al., 2012). Meanwhile, the TP is highly sensitive to climate change (Pepin and Lundquist,50 2008; Kang et al., 2010;Chen et al., 2015). It is the driving force for regional environmental changes, and it amplifies environmental changes to global scale as well (Pan et al., 1996;Kang et al., 2010).
Land-atmosphere interactions over the TP play a crucial role in controlling the hemispheric atmospheric circulation pattern and climate evolution (Yang et al., 2004;Duan and Wu, 2005;Xiao and Duan, 2016;Li et al., 2018). Previous studies have revealed that accurate simulation of water and heat flux exchanges between the land surface and the atmosphere is a 55 pivotal step towards improving the predictability and the projection accuracy of the climate system (Sellers et al., 1997;Pitman, 2003); this can be achieved through a comprehensive and accurate understanding of the land-atmosphere interactions based on in-situ observations (Yang et al., 2009). However, compared with other terrestrial regions of the world, observational data are scarce over the TP, owing to its vast geographic area with steep terrain, varied landforms, complex and diverse climates, harsh environmental conditions. The sparse and biased spatial distribution of observation stations is hard to match the high 60 degree of landscape heterogeneity over the TP. In addition, high uncertainties in the satellite-retrieved land and atmospheric environmental variables of the TP impair the establishment of continuous, long-term regional-scale observations in remote areas of the TP. The lack of sufficient observational data limits our understanding of the interactions between the different earth spheres with heterogeneous land surface conditions and hinders the development of parameterization schemes in some critical physical processes of the land surface and atmospheric boundary layer, thereby, leading to associated uncertainties in 65 estimating the past, present, and future climate change and its impacts. Therefore, it is essential to improve the atmospheric observation capability on the TP and its surrounding areas and obtain accurate atmospheric physical parameters for the nearsurface and boundary layers over the TP, which can significantly contribute to the scientific understanding of the weather, climate and environmental changes, as well as their impacts, from regional scale (TP) to global scale.

3
To mitigate the scarcity of observational data and to improve our understanding of the coupled local land-atmosphere 70 system and its effects, a series of atmospheric field experiments have been carried out over the TP since the 1970s. For example, the first Qinghai-Xizang Plateau Meteorology Experiment (QXPMEX) (Tao et al., 1986), the Global Energy and Water Cycle Experiment (GEWEX) Asian Monsoon Experiment (GAME)/Tibet intensive observation (Wang, 1999) Although some meteorological data can be requested from the National Tibetan Plateau Data Center (TPDC) in recent years (http://www.tpdc.ac.cn), only the daily mean values are provided, which are commonly lack of consistent and information of 90 standard data processing methods. Furthermore, the temporal resolution of these daily mean value is too coarse for the land surface and climate modeling community, for which at least hourly values are required to run models and to evaluate detailed physical models. To overcome the above issues, a continuous and long-term integrated observational dataset of landatmosphere interaction with high temporal resolution is now provided (Ma et al., 2020). The underlying observation network is composed of six stations over the TP. At each station, the following measurements are available: meteorological gradient, 95 surface radiation, EC and soil hydrothermal. This dataset is released in a unified format that can be easily accessed and used by many communities, aiming to facilitate the consistency and continuity in scientific understanding of the interactions among the multi-sphere coupled systems over the TP. We expect this dataset will be widely used in studying the environment of the "Third Pole", especially by the atmosphere, hydrology, ecology and cryosphere communities. We also hope this dataset will promote the sharing, opening and value-added exploitation of the in-situ land-atmosphere interaction observations over the TP. 100 In this paper, we introduce and provide access to the long-term hourly dataset of the integrated land-atmosphere interaction observations over the TP. The integrated land-atmosphere interaction observation network is first described in Section 2. Section 3 deals specifically with a description of the meteorological, solar radiation, EC and soil hydrothermal data, and presents an overview of the observation infrastructure, highlighting differences and similarities between the stations with respect to the observation items, and their variations at diurnal, daily and monthly scales. The availability of this dataset is 105 documented in Section 4 and a final summary is presented in Section 5.

Site descriptions
The integrated land-atmosphere interaction observation network in this study consists of six field stations ( The MAWORS station was located in the region where the atmospheric circulation was influenced by the westerly wind 115 all year round. Soil at this station was predominately sandy soil and gravel with sparse and short grass-covered. Large scale modern glaciers are distributed around the station (the standard deviation of elevation within a kilometer around the station is 152.92 m, which is the highest among the six stations as shown in Table 1) and exert great influence on the local weather and climate. The observations from this station are of great significance for the study of interactions between westerly winds and monsoon and their effects on land-glacier-atmosphere changes, as well as changes in snow and ice resources. 120 The NADORS station was built in a flat and open mountain valley in the northwestern TP (with the lowest standard deviation of elevation). The land use type here is Gobi Desert with very short grasses (about 1-2 cm) on the sandy soil and gravel surface. It is located at the convergence zone of the Indian monsoon and westerly wind, where these two atmospheric circulations interact intensively, making the NADORS as an excellent location for the study of westerly-monsoon interactions on the desert landscape. 125 The BJ site is located in a flat, open prairie except for the north, where there stand low hills (the standard deviation of elevation is 15.14 m). The site is well-vegetation-covered and the dense grasses are relatively high with height up to 5 cm. Soil at the site is predominantly sandy silt loam. The BJ site is an ideal place to observe the land-atmospheric interactions on the alpine meadow ecosystem.
The NAMORS station is located on the banks of Lake Nam Co, with the Nyainqentanglha Mountain behind. The land is 130 covered by alpine meadows and the soil type is predominantly sandy silt loam, but the gravel content is high at 30-40 cm below the ground. As lake has a significant influence on the atmospheric circulation in this region, and plays a certain role in regulating temperature variation and precipitation, etc. Thus, this station is an ideal place to measure the land-atmosphere interactions in the water-land-mountain mesoscale system. The SETORS station lies in a mountain valley close to the forested southeastern TP (the terrain is highly heterogeneous, but is not as complex as the MAWORS). It's surrounded by a dense vegetation cover (mainly temperate needle-leaf trees and alpine meadows). The shallow soil here is well developed and the water-holding capacity of the soil is greatly enhanced due to the presence of organic matter, while the deep soil is predominantly gravel. The observations from the SETORS station are important for studying the water and heat transport along the alpine valleys by the South Asian monsoon, the alpine forest-145 glacier-atmosphere interactions, and the transport of hydrothermal components of the vertical belt in the mountainous regions.
This high-cold region observation network is an essential component of the meteorological observation platform over the TP, carrying out land surface processes observations in areas that are typical in geography while currently lack of in-situ observations. This network serves as key locations for field observations and experiments: in particular, for monitoring the interactions between geological processes and climate; for collecting first-hand, high-resolution records of modern 150 environmental variations; and for monitoring land surface processes and atmospheric processes. The observation system at each station primarily includes the following four categories of measurements: meteorological variables either from the PBL tower or the AWS, solar radiation components, eddy covariance fluxes and soil hydrothermal conditions. The meteorological instruments consist of up to 5 levels of wind speed and direction, air temperature, relative humidity instruments, surface air pressure, and precipitation. The surface radiation components include the incoming and outgoing shortwave and longwave 155 radiations. The open-path EC turbulent flux measurement system is used to sample the high frequency vertical turbulent fluxes of the sensible heat flux, latent heat flux and carbon dioxide flux. Vertical profiles of soil temperature and soil moisture content are measured by multilayer temperature probes and water content reflectometers (5 or 6 layers). A list of the observation items and instruments in detail can be found in Table 2. To ensure the accuracy and reliability of the observations, periodic inspection, maintenance and calibration are carried out by professional engineers. Meanwhile, all stations are manned except for the cold 160 winter season, and the instruments are checked and data are collected and processed regularly.

Meteorological observations
To fully characterize the meteorological conditions and their vertical distributions in the surface layer, instruments were installed at several heights on a multi-layer PBL tower (QOMS, NAMORS, SETORS and BJ). For stations without the PBL 165 tower, meteorological variables are recorded by a one-layer AWS at MAWORS and two-layer AWS at NADORS. The layer 6 arrangements of sensors are not the same at these 6 stations. For example, five layers of wind speed and wind direction anemometers, air temperature and humidity probes were installed at QOMS and NAMORS; four layers of sensors were installed at SETORS; for BJ, three layers of wind and two layers of air temperature and relative humidity probes were available during 2006-2014, while four levels of these measurements were provided during 2015-2016 (see Table 2  The climatological averaged wind speeds at diurnal, daily and monthly scales are shown in Figure 2  detected, the air temperature data provided at present are in raw format without any post-processing applied. Consequently, careful inspection is crucial when air temperature observations are required. In subsequent work, stricter data quality controls will be applied to detect problematic data and quality flags will be provided for each observational element.

Humidity 215
The heights of the humidity sensors are the same as those of the air temperature probes. Besides the relative humidity, up to four layers of water vapor pressure observations are also available at MAWORS (1. The relative humidity showed obvious diurnal variations, peaking in the afternoon (Figure 2g). Compared with the magnitude of diurnal variations in summer, the diurnal range of relative humidity at SETORS in winter and spring was much greater, reaching 50%, and the maximum value of the average diurnal cycle of relative humidity was about 80%, which was also significantly higher than those at other stations. In contrast, the diurnal variability during the monsoon season was much smaller than that at BJ, QOMS, NAMORS and MAWORS. The monthly relative humidity was lowest at NADORS, however, 225 there was a marked increase in summer due to the transition of mid-latitude westerlies to the Asian summer monsoon.
Differences in humidity among the six stations presented in the diurnal and daily relative humidity records were clearly reflected in the seasonal variations at the monthly scale.

Air pressure
Barometers produced by Vaisala were installed at each station. Compared with variations in wind speed, air temperature 230 and relative humidity, the diurnal and seasonal variations in air pressure were not obvious (Figure 2j-l), and pressure remained at a relatively stable level throughout the year. Air pressure is elevation-dependent amongst the six stations, with the highest value at SETORS and the lowest value at NAMORS, while a consistent diurnal and seasonal variations were found both at QOMS and NADORS of similar altitude.

Precipitation 235
Precipitation is measured at all stations except for MAWORS, either with tipping buckets or weighting gauges. At BJ and NADORS, the cumulative precipitation is recorded, while the total half-hourly precipitation is recorded at NAMORS, QOMS and SETORS. For the cumulative precipitation, negative growth resulting from the evaporation from the rain gauge can seriously affect the measurement accuracy. Moreover, large errors can be introduced in the precipitation time series by windinduced under-catch, wetting loss, evaporation loss, and underestimation of trace precipitation amounts; it is difficult to apply 240 bias correction to account for these losses (Goodison et al., 1998). While precipitation data are extremely valuable, accurate measurement is notoriously difficult due to the large errors mentioned above, particularly in cold regions such as the TP.
Therefore, in the released datasets, the precipitation data are provided in raw format without any post-processing, which might potentially be underestimated, thus further bias correction or data selection is necessary before the precipitation observations are used.

EC data 270
The EC technique was applied to provide high-quality and continuous surface turbulent flux data for momentum, sensible, and latent heat. The EC system comprises a sonic anemometer (CSAT3, Campbell Scientific, Inc.) and a fast-response gas analyzer (LI-7500 open-path gas analyzer, Li-COR). All of the turbulence data were processed and quality-controlled using the TK3 software package (Mauder and Foken, 2011); the main processing procedures were as follows: excluding physically invalid values and spikes, revising the time delay of the high-frequency water vapor and carbon dioxide sampling, planar fit 275 coordinate rotation, correction of the loss of frequency response, correction of the ultrasonic virtual temperature and density fluctuations. The quality of each turbulent flux data series was evaluated by using the stationarity test and integral turbulence characteristics test. By combining the quality flags for stationarity and the integral turbulence characteristics test, a final quality flag (1-9) was assigned to each specific turbulent flux value except those for BJ, where classes 0-2 were used. Classes 1-3 (or 0 at BJ) indicate good quality suitable for fundamental research purposes, and classes 4-6 (1 at BJ) indicate suitability for 280 general use, such as long-term analysis. Classes 7-9 (2 at BJ) should be discarded. The multi-year diurnal variation and seasonal variation of sensible and latent heat flux were calculated based on the data with medium or higher quality.

Sensible heat flux
As can be seen from the diurnal variations of sensible heat flux in Figure 4a, the sensible heat fluxes at all stations were negative at night. During the period from March to October, the atmospheric heating effect on the ground at NADORS was 285 the strongest during the night, while the magnitude of diurnal variation in the sensible heat flux was the lowest here among the six stations from April to September. The variations in sensible heat flux (Figure 4a-c) show that prior to the monsoon season, and the sensible heat flux was the main consumer of surface available energy, then the diurnal variation in sensible heat flux decreased significantly with the onset of summer monsoon and was comparable to the latent heat flux. In other words, sensible heat flux exchanges prevail during the pre-monsoon periods. The timing of the onset of decreasing sensible heat flux following 290 the spring maximum varied, occurring earliest at SETORS and NAMORS, followed by BJ and NADORS. Influenced by the interactions between the midlatitude westerlies and the summer monsoon, the summer sensible heat fluxes were significantly lower than those in spring at all stations.

Latent heat flux
In contrast to the bimodal pattern of the seasonal variations in sensible heat flux, the seasonal variation in latent heat flux 295 revealed a unimodal pattern, that is, the latent heat flux was small during the pre-monsoon period, and when monsoon outbreaks, it increased rapidly as precipitation became frequent and the surface soil turned wet. The latent heat flux then increased

Carbon dioxide flux
The carbon dioxide flux is an important component of the atmospheric carbon balance and is a very important variable in the study of global climate change. As one of the key components of the EC monitoring system, the observed carbon dioxide 305 fluxes at each station are provided through the density correction and frequency response correction applied by the TK3 software package (Mauder and Foken, 2011). A previous study has reported that the self-heating of the infrared gas analyzer in the open-path EC system can cause notable differences in temperature between the observation path and the ambient air, which may result in signal distortion (Burba et al., 2008); therefore, it is necessary to apply a specific correction to the carbon dioxide flux data to eliminate the heating impact and to accurately reveal the intensity of carbon dioxide exchange in the TP 310 ecosystem (Zhu et al., 2012). However, the heating effect of the instrument was not been considered in the carbon dioxide flux data provided in this manuscript, more detailed information can refer to the studies of Burba et al. (2008) and Zhu et al. (2012).
When these data are used in studies of carbon dioxide exchange or related works (for example, estimating the net ecosystem production and its components), this specific correction of the data is needed to fully account for the impact of instrumental heating on observations. 315

Ground surface temperature
Ground temperatures at NADORS, SETORS and BJ are provided in this dataset. The variations in ground temperature show the weakest diurnal variations at BJ and the strongest at SETORS, where the ground temperature during the night was highest among the six stations. On the daily scale, the daily mean ground temperature at BJ was lower than that at SETORS 320 and NADORS throughout the year, although its amplitude of the diurnal cycles was larger than that at the other two stations owing to the lower night-time temperatures (Figure 5a). Daily mean and monthly ground temperatures at BJ dropped below 0 °C during all months from October to April.   at BJ showed that it was usually relatively small and had evident diurnal and seasonal variations.

Data availability
Raw data were converted from binary mode to ASCII mode, and then key variables were extracted and saved as commaseparated values (.csv format). The CSV format was chosen as it is one of the most widely supported structured data format in scientific applications. The plausible value check, time consistency check, and internal consistency check were applied to 350 ensure the accuracy and reliability of the observations. However, to retain the observations in their original form as much as possible, there is no further process taken except for replacing outliers with missing value (NaN). Data consistency check procedures were applied to ensure the accuracy and reliability of the observations, but the data quality flag is not available for the moment. For turbulent flux data, classes 1-3 (0 for BJ station) were recommended for fundamental research, such as surface energy balance analysis. Classes 4-6 (1 for BJ station) can be used in continuously-running systems or for long-term analysis. 355 Some time series of observations should be used with caution (for example, the soil hydrothermal data in SETORS), as anomalous changes or values were detected. In this case, further procedures such as bias correction or data selection are required. The local time was used in all the data files (UTC+8). All datasets presented and described in this article have been released and are available to free download from the Science Data Bank (http://www.dx.doi.org/10.11922/sciencedb.00103, Ma et al., 2020) and the TPDC (https://data.tpdc.ac.cn/en/data/b9ab35b2-81fb-4330-925f-4d9860ac47c3/). Special 360 compressed files were designated for each station with four categories: turbulent flux data (FLUX), gradient meteorological data (GRAD), soil hydrothermal data (SOIL) and radiation data (RADM). Meanwhile, the data integrity of each variable was also provided every year, with the value of 100 indicates complete continuous data, with no missing data. The heat maps shown in the Appendix are used to provide the data integrity information. These figures are very useful as they provide an intuitive depiction of the availability of each variable, facilitating data selection when analyzing land-atmosphere interactions and 365 structure of PBL, driving land surface models, or evaluating model results. (hourly) to date. Therefore, this fine-resolution data product can help to promote comprehensive scientific understanding of the interactions among the multi-sphere coupled systems over the TP and even the globe; to quantify uncertainties in satellite and model products; to assess the biases and gaps existing between the model simulations and reality; and to facilitate the development and improvement of land surface process models in cold regions. We believe that the datasets presented in this paper will contribute to these research areas and that they will be widely used in model development and evaluation. 380  indicates complete continuous data, with no missing data. WS and WD represent wind speed and direction, respectively, 395 followed by heights of each level with the underline symbol as connection; Ta refers to the air temperature; Relative humidity and water vapor pressure are expressed using RH and Vapor, respectively. Figure A2. Same as Figure A1, but for the surface radiations. Rsd and Rsu represent the incoming and outcoming solar radiation, respectively; Rld and Rlu refer to the downward and upward longwave radiation. The net radiation is expressed using Rn. 400 Figure A3. Same as Figure A1, but for the soil hydrothermal observations. Ground temperature is represented by Tg, and the soil temperature and soil water content are expressed with Ts and SWC, respectively; SHF refers to the soil heat flux. Figure A4. Same as Figure A1, but for the turbulent flux observations. H represents sensible heat flux and LE represents 405 latent heat flux, and the CO2 flux is expressed with Fc.

Summary
Author contributions. YMM, ZYH, ZPX and BBW led the writing of this article and endorse the responsibility of the experimental site and the instruments. YMM and ZPX drafted the manuscript and ZPX led the consolidation of the dataset, prepared the data in the standardized format described in this paper and wrote this manuscript together with all co-authors. 410 Competing interests. The authors declare that they have no conflict of interests.