This dataset includes the observational data that were collected by two sets of Cosmic-ray Soil Moisture Observation System (COSMOS), named crs_a and crs_b, which were installed near the Daman Superstation in the flux observation matrix from 1 June through 20 September 2012. The land cover in the footprint was maize crop, and the site was located with the cropland of the Daman Irrigation District, Zhangye, Gansu Province. Crs_a was located at 100.36975° E, 38.85385° N and 1557.16 m above sea level; Crs_b was located at 100.37225° E, 38.85557° N and 1557.16 m above sea level. The bottom of the probe was 0.5 m above the ground; the sampling interval was 1 hour. The raw COSMOS data include the following: battery (Batt, V), temperature (T, ℃), relative humidity (RH, %), air pressure (P, hPa), fast neutron counts (N1C, counts per hour), thermal neutron counts (N2C, counts per hour), sample time of fast neutrons (N1ET, s), and sample time of thermal neutrons (N2ET, s). The distributed data include the following variables: Date, Time, P, N1C, N1C_cor (corrected fast neutron counts) and VWC (volume soil moisture, %), which were processed as follows: 1) Quality control Data were removed and replaced by -6999 when (a) the battery voltage was less than 11.8 V, (b) the relative humidity was greater than 80% inside the probe box, (c) the counting data were not of one-hour duration and (d) then neutron count differed from the previous value by more than 20%. 2) Air pressure correction An air pressure correction was applied to the quality-controlled raw data according to the equation contained in the equipment manual. The procedure was previously described by Jiao et al. (2013) and Zreda et al. (2012). 3) Calibration After the quality control and corrections were applied, soil moisture was calculated using the equation in Desilets et al. (2010), where N0 is the neutron counts above dry soil and the other variables are fitted constants that define the shape of the calibration function. Here, the parameter N0 must be calibrated using the in situ observed soil moisture within the footprint. This procedure was previously described by Jiao et al. (2013) and Zreda et al. (2012) 4) Computing the soil moisture Based on the calibrated N0 and corrected N1C, the hourly soil moisture was computed using the equation from the equipment manual. This procedure was previously described by Jiao et al, (2013) and Zreda et al. (2012) For more information, please refer to Liu et al. (2016) (for multi-scale observation experiment or sites information), Zhu et al. (2015) (for data processing) in the Citation section.

The annual report (2008 and 2009) of the Zhangye water conservancy bureau included: (1) the water management staff statistics; (2) irrigation statistics; (3) projects status statistics; (4) project management statistics; (5) the technical and economic index of the irrigation area management; (6) water management tasks status statistics; (7) water management planning index. Those provide reliable information for water resources analysis in the middle stream.

The dataset of runoff plot observations was obtained in the Binggou watershed foci experimental area from Jun. 19 to Oct. 17, 2008. The runoff plot (38°03′, 100°13′, 3472m, with a slope of 20.16°) was 10m long, 5m wide and 80cm deep, with soil depth about 50cm and sandy clay and gravels beneath (50-80cm). The main vegetation type is scrub (about 20cm high) and grass (about 3cm high). Observation items included the surface flow, interflow (80cm down the land surface), and precipitation at a fixed point at the right of the runoff plot. One subfolder and two data files (directions on data observations and raw data) were archived.

The dataset of object spectral was obtained in the Linze station foci experimental area from May 25 to Jul. 11, 2008. The measurement instrument is ASD Spectroradiometer (350～2 500 nm) from BNU and the reference board (40% before Jun. 15 and 20% hereafter). The selected typical objects included maize field, soil, soil with known moisture and desert scrub. The measured quadrates included Wulidun farmland quadrates (May 28 and 30, Jun. 16 and 29 and Jul. 11), the desert transit zone strips (May 28 and 30 and Jun. 16) and Linze station quadrates (May 23 and Jul. 9) Besides, soil samples were collected inside Linze station quadrates on Jun. 24 and 30, 2008. Raw spectral data were archived as binary files, which were recorded daily in detail, and pre-processed data on reflectance and transmittivity were archived as text files (.txt). See the metadata record “WATER: Dataset of setting of the sampling plots and stripes in the Linze station foci experimental area” for more information of the quadrate locations.

The dataset of snow properties measured by the Snowfork was obtained in the Binggou watershed foci experimental area from Dec. 5-16 2007, during the pre-observation period. The aims of the measurements were to verify applicability of the instruments and to acquire snow parameters for simultaneous airborne, satellite-borne and ground-based remote sensing experiments and other control experiments. Observation items included: (1) physical quantities by direct observations: resonant frequency, the rate of attenuation and 3db bandwidth (2) physical quantities by indirect observations: snow density, snow complex permittivity (the real part and the imaginary part), snow volumetric moisture and snow gravimetric moisture. Five files including raw data and processed data are kept, data by the Snowfork on Dec 5, data by BG-A MODIS on Dec 6 and 7, data in BG-B, BG-C, BG-D and BG-E on Dec 10, and data in BG-D with the microwave radiometer on Dec 14 and 16.

The dataset of automatic meteorological observations was obtained at the A'rou freeze/thaw observation station from Jul. 25, 2008 to Dec. 31, 2009, in Wawangtan pasture (E100°28′/N38°03′, 3032.8), Daban, A'rou. The experimental area, situated in the valley highland of south Babaohe river, an upper stream branch of Heihe river, with a flat and open terrain slightly sloping from southeast to southeast and hills and mountains stretching for 3km is ideal for a horizontal homogeneous underlying surface. Observation items included multilayer (2m and 10m) of the wind speed, the air temperature and air humidity, the air pressure, precipitation, four components of radiation, the multilayer soil temperature (10cm, 20cm, 40cm, 80cm, 120cm and 160cm), soil moisture (10cm, 20cm, 40cm, 80cm, 120cm and 160cm), and soil heat flux (5cm & 15cm). The raw data were level0 and the data after basic processes were level1, in which ambiguous ones were marked; the data after strict quality control were defined as Level2. The data files were named as follows: station+datalevel+AMS+datadate. Level2 or above were strongly recommended to domestic users. As for detailed information, please refer to Meteorological and Hydrological Flux Data Guide.

The dataset of airborne Polarimetric L-band Multibeam Radiometers (PLMR) was acquired on 26 July, 2012, located in the middle reaches of the Heihe River Basin. The aircraft took off at 9:10 am (UTC+8) from Zhangye airport and landed at 13:40 pm, with the flight time of 4.5 hours. The flight was performed in the altitude of about 2300 m and at the speed of about 220-250 km during the observation, corresponding to an expected ground resolution of about 700 m. The PLMR instrument flown on a small aircraft operates at 1.413 GHz (L-band), with both H- and V-polarizations at incidence angles of ±7.5°, ±21.5° and ±38.5°. PLMR ‘warm’ and ‘cold’ calibrations were performed before and after each flight. The processed PLMR data include 2 DAT files (v-pol and h-pol separately) and 1 KMZ file for each flying day. The DAT file contains all the TB values together with their corresponding beam ID, incidence angle, location, time stamp (in UTC) and other flight attitude information as per headings. The KMZ file shows the gridded 1-km TB values corrected to 38.5 degrees together with flight lines. Cautions should be taken when using these data, as the RFI contaminations are often higher than expected at v-polarization.

This dataset contains the flux measurements from site No.3 eddy covariance system (EC) in the flux observation matrix from 3 June to 18 September, 2012. The site (100.37634° E, 38.89053° N) was located in a cropland (maize surface) in Yingke irrigation district, which is near Zhangye, Gansu Province. The elevation is 1543.05 m. The EC was installed at 3.8 m high, and sampled at 10 Hz. The EC was installed at a height of 3.8 m; the sampling rate was 10 Hz. The sonic anemometer faced north, and the separation distance between the sonic anemometer and the CO2/H2O gas analyzer (Gill&Li7500A) was 0.2 m. Raw data acquired at 10 Hz were processed using the Eddypro post-processing software (Li-Cor Company, http://www.licor.com/env/products/ eddy_covariance/software.html), including spike detection, lag correction of H2O/CO2 relative to the vertical wind component, sonic virtual temperature correction, angle of attack correction, coordinate rotation (2-D rotation), corrections for density fluctuation (Webb-Pearman-Leuning correction), and frequency response correction. The EC data were subsequently averaged over 30 min periods. Moreover, the observation data quality was divided into three classes according to the quality assessment method of stationarity (Δst) and the integral turbulent characteristics test (ITC), which was proposed by Foken and Wichura [1996]: class 1 (level 0: Δst<30 and ITC<30), class 2 (level 1: Δst<100 and ITC<100), and class 3 (level 2: Δst>100 and ITC>100), representing high-, medium-, and low-quality data, respectively. In addition to the above processing steps, the half-hourly flux data were screened in a four-step procedure: (1) data from periods of sensor malfunction were rejected; (2) data before or after 1 h of precipitation were rejected; (3) incomplete 30 min data were rejected when the missing data constituted more than 3% of the 30 min raw record; and (4) data were rejected at night when the friction velocity (u*) was less than 0.1 m/s. There were 48 records per day; the missing data were replaced with -6999. Moreover, suspicious data were marked in red. The released data contained the following variables: data/time, wind direction (Wdir, °), wind speed (Wnd, m/s), the standard deviation of the lateral wind (Std_Uy, m/s), virtual temperature (Tv, ℃), H2O mass density (H2O, g/m^3), CO2 mass density (CO2, mg/m^3), friction velocity (ustar, m/s), stability (z/L), sensible heat flux (Hs, W/m^2), latent heat flux (LE, W/m^2), carbon dioxide flux (Fc, mg/ (m^2s)), quality assessment of the sensible heat flux (QA_Hs), quality assessment of the latent heat flux (QA_LE), and quality assessment of the carbon flux (QA_Fc). In this dataset, the time of 0:30 corresponds to the average data for the period between 0:00 and 0:30; the data were stored in *.xlsx format. For more information, please refer to Liu et al. (2016) (for multi-scale observation experiment or sites information), Xu et al. (2013) (for data processing) in the Citation section.

The dataset of TIR spectral emissivity was obtained in the arid region hydrology experiment area and A'rou foci experiment area. Observations were by: (1) Spectral emissivity obtained from 102F at 2-25um in cooperation with the handheld infrared thermometer (BNU) for the surface radiative temperature and one au-plating board for downward atmospheric radiation. The radiative transfer equation and TES methods were applied to retrieve emissivity. The grassland and the concrete floor were measured on May, 27, 2008, the wheat field and the maize field at ICBC resort on May, 29, 2008, the concrete floor (multiangle measurements) at ICBC resort on Jun. 3, 2008, the bare soil and the maize leaf in Yingke oasis maize field on Jun. 22, 2008, the maize and wheat canopy in Yingke oasis maize field on Jun. 23, 2008, the rape field in Biandukou experimental area on Jun. 24, 2008, the alfalfa, the saline land, the grassland and the barley land on Jun. 26, 2008, the wheat field and the maize field in Yingke oasis maize field on Jun. 29, 2008, the desert bare land and vegetation (Reaumuria soongorica) in No. 2 Huazhaiai desert plot on Jun. 30, 2008, the rape field and the grassland in Biandukou experimental area on Jul. 6, 2008, and the grassland and the bare land (multiangle) in A'rou experimental area on Jul. 14, 2008. The cold blackbody calibration (*.CBX/*.CBB), the warm blackbody calibration (*.WBX/*.WBB), the ground objects measurements (*.SAX), au-plating board measurements, and the downward atmospheric radiation (*.DWX) were all needed during observation. Moreover, the spectral radiance and emissivity were also archived. The response function of various bands could be acquired by 102F. And then emissivity of 2-25um could be retrieved. Two results of emissivity were developed: one was direct from 102F and the other was retrieved by ISSTES (Iterative spectrally smooth temperature-emissivity separation). Spectral resolution for raw data and proprecessed data was 4cm-1. (2) Spectral emissivity obtained from BOMAN at 2 -13μm in cooperation with the blackbody barrel and the blackbody from Institute of Remote Sensing Applications and the blackbody (BNU). The desert was measured on Jun. 30 and Jul. 1, 2008, A'rou foci experimental area on Jul. 14, 2008, indoor observations on the deep and shallow layer soil, vegetation, small stones, two maize plants from Yingke No.2 (YKYZYMD02) field and one maize plant and bare land from No. 3 (YKYZYMD03)field on on Jul. 16, 2008, Linze experimental area on Jul. 17, 2008, and gobi on Jul. 18, 2008. The sample site, coordinates, time and photos were all archived. During each observation, BOMAN was preheated and the blackbody was set at the predicted target temperature, which would be changed after the infrared radiation of the blackbody was measured by BOMAN. And then the target infrared radiation, the downward atmospheric radiation (reflected by the au-plating board) and the infrared radiation of the blackbody would be measured one by one. Raw data were archived in Igm, and after processed by FTSW500, the result was Rad (radiation). Finally, Rad would be changed into txt files by Matlab programs.

Soil respiration observation was carried out for the typical vegetation ground in the lower reaches of the Heihe River Basin during the aviation flight experiment in 2014. The observation started on 23 July, 2014 and finished on 2 August, 2014. 1. Observation time Days from 23 July to 2 August, 2014 (25 July, 2014 excepted) 2. Samples and observation methods Large areas with relatively homogeneous vegetation (greater than 100 m * 100 m) were chosen as the observation samples. And combined the flux tower sites distribution of the lower reaches, five field samples closed to the sites were selected The observation sites sampled including Populus and Tamarix mixed forest, Populus, Tamarix group, bare ground and melon quadrats. 3-5 plots were observed for each samples. The PVC soil rings were installed one day before observation and kept about 5 cm out of the ground (the inner diameter of the PVC is 19.5 cm, the outer diameter is 20.0 cm, and the height is 12.0 cm). Minimal the effects to the surface of vegetation and withered matter when install the rings. In order to avoid fluctuations of the soil respiration value by the PVC rings, soil respiration rate was obtained when it returned to its original state (about 24h after the rings install). The observation time for each day was from 8:00 to 12:00 when soil respiration is relatively stable and can represent the whole day in this time. The Li-8100 Open Path soil carbon flux automatic analyzer was used (Model 8100-103) once for each plot. Cycles of observation for all plots of the five samples were completed for every morning. The soil respiration values of the samples were obtain by averaging the values of plots of the samples. 3. Observation instrument Li 8100 4. Data storage The observation recorded data were stored in excel and the original Soil respiration data were stored in 81x files.

The dataset contains vegetation type and plant structure in the middle reaches of the Heihe River Basin, which was used to validate products from remote sensing. It was generated from investigating the land cover strips of CASI and SASI the middle reaches of the Heihe River Basin between 25 June and 6 August in 2012. Instruments: High-precision handheld GPS (2-3 m) and digital camera were used as main tools in the survey. Measurement method: Vegetation range in the middle reaches of the Heihe River Basin and survey route could be decided with the help of Google Earth. Wuxing village in Xiaoman town was selected to survey detailed and other places were investigated as far to reach as possible. Main methods were to write down the longitude and latitude, phenology of the plant structure, take photos for the vegetation. Dataset contains: longitude and latitude, vegetation type, area and phenology. Observation Place: CASI flight area in artificial oasis in the middle reaches, CASI stripe flight area in the middle reaches and Zhangye district. Date: From 25 June and 6 August in 2012.