The No. 4 hydrological section is located at Wujin Heihe River Bridge (100.433° E, 39.065° N, 1431 m) in the midstream of the Heihe River Basin, Zhangye city, Gansu Province. The dataset contains recorded by the No.4 hydrological section from 10 June, 2012 to 10 August, 2012, and from 6 September, 2013 to 31 December, 2013. The width of this section is 58 meters and the cross-sectional area is unstable because of human factors. The water level was measured using an HOBO pressure range and the discharge was measured using cross-section reconnaissance by the StreamPro ADCP. The dataset includes the following parameters: water level (recorded every 30 minutes) and discharge. The missing and incorrect (outside the normal range) data were replaced with -6999. For more information, please refer to Li et al. (2013) (for hydrometeorological observation network or sites information), He et al. (2016) (for data processing) in the Citation section.

The research area is located in the middle section o the northern slope of the Tianshan Mountains. The research area extends from Wusu in the Tacheng District of Xinjiang in the west to Mulei County in Changji Prefecture in the east. It is approximately 500 km long from east to west. The vertical vegetation gradient on the northern slope of the Tianshan Mountains can be divided into six different belts: alpine cushion vegetation belt (>3400 m), sub-alpine meadow belt (3400~2700 m), mid-mountain forest belt (2700~1720 m), forest steppe belt (1720~1300 m), semi-desert belt (1300~700 m) and typical desert belt (<700 m). Based on the characteristics of the vertical vegetation belts on the northern slope of the Tianshan Mountains, five sedimentary sections with different elevations, different vegetation belts and different sedimentary ages were selected for analysis. Five mid-late Holocene sections were measured to calculate the composite dissimilarity index of sporopollen, and the index was used to explain the sporopollen diversity. The index was then combined with integrated multiple analysis data, such as particle size, magnetic susceptibility, and ignition loss, and the changes in biodiversity and environmental characteristics since the mid-late Holocene in the area were assessed. The data include the following: 1. Sporopollen grain number data for the Daxigou section (8-110 cm, a total of 52 layers were analysed for sporopollen grain number, 3640±60 a BP to 890±60 a BP) 2. Sporopollen grain number data for the Xiaoxigou section (0-90 cm, a total of 38 layers were analysed for sporopollen grain number, 3240±60 a BP) 3. Sporopollen grain number data for the Huashuwozi section (0-106 cm, a total of 52 layers were analysed for sporopollen grain number, 2170±185 a BP to 450±155 a BP) 4. Sporopollen grain number data for the Sichanghu section (10-84 cm, a total of 19 layers were analysed for sporopollen grain number, 1000±50 a BP to 665±65 a BP) 5. Sporopollen grain number data for the Dongdaohaizi section (0-190 cm, a total of 64 layers were analysed for sporopollen grain number, 4500±310 a BP to 305±130 a BP) For detailed descriptions of the data, please refer to the following study: "Palaeo-biodiversity at the Northern Piedmont of Tianshan Mountains in Xinjiang During the Middle to Late Holocene"

During the 2012 aerial remote sensing experiment conducted midstream, Li-Cor8100 was used to measure soil respiration every five days in the EC matrix area. Instrument: LI-Cor8100 Measuring Method: Soil respiration ring was made using PVC pipe with length of 10 cm. Before measuring soil respiration, soil respiration ring was inserted into the soil, 4 cm in soil and 6 cm above soil. Soil respiration measurement should be taken after standing for at least 24 hours science ring was inserted in soil. Sample measurement time is during 9-12 in the morning. Set of three replicates per plot. Marked according to EC site name. Data content: Data content includes header information, and once every five days repeated three times observations value and the average value. Measuring location: EC sites within the matrix core experiment area (No. EC01 to EC17), each plot set three repeat samples. For the superstation (EC15) plot set nine repeat samples. Measuring time: From 6 June to 20 August, 2012, once every five days for site EC01, EC03, EC05, EC10, EC11, EC12, EC13, EC14, and EC17; from 1 July to 20 August, 2012, once every five days for site EC02, EC04, EC06, EC07, EC08, EC09 and EC16. The time used in this dataset is in UTC+8 Time. Part of the observation points during the observation just irrigation, these times are not observable.

The project studying the evolution pattern and development trend of the arid environment in western China was a major research component of the project Environmental and Ecological Science for West China, which was funded by the National Natural Science Foundation of China. The leading executive of the project was Academician Zhisheng An from the Institute of Earth Environment of the Chinese Academy of Sciences. The project ran from January 2002 to December 2004. The data collected by the project include the following: 1. History and variability data for arid regions in western China: 1) Chinese Loess Plateau mass accumulation rate data (3600-0 kyr BP): Fields include age and mass accumulation rate (MAR) (txt file). 2) Chinese Loess Plateau grain size and magnetic susceptibility data (3600-0 kyr BP): Fields include age, stacked mean grain size, and stacked magnetic susceptibility (txt file). 2. Sporopollen content data of different loess strata since 12 kyr BP in the Yaozhou District of Shanxi Province (excel table): The distributions of 27 species of sporopollen (0-397 cm) from 67 different layers of loess samples are included. 3. 10Be record data (table) 10Be concentration, magnetic susceptibility and bulk density data of loess with different thicknesses (79.67- 0.09 kyr BP). 4. Simulation data on the modulation of the East Asian monsoon resulting from orbital variability driven by the uplift of the Tibetan Plateau: ah0-sum.nc nc file, hh0-sum.nc nc file, jfh0-sum.nc nc file, kdh0-sum.nc nc file, lfh0-sum.nc nc file, mask.nc nc file, phis.nc nc file.

The dataset of the structure parameter measurements for afforested forest was obtained in the Zhangye city foci experimental area. To be specific, the full height was measured by the altimeter 12km east of Zhangye city on May 30, the LAI by LAI-2000 in Linze grassland foci experimental area on Jun. 5, measurements in Yingke oasis on Jun. 14 and 17, afforested forest parameters in Jiulongjiang forest farm on Jun. 28 and LAI by LAI-2000 in Huazhaizi desert steppe foci experimental area on Jul.1, 2008. 6 files in total were archived, zhangyefroestpark.xls (31K, 5-30), Linzestation.xls (16K, 6-4), yingke1.xls ( 21K, 6-14), yingke2.xls (21K, 6-17), yingke3.xls (16K, 7-1), and jiulongjiangforestfarm.xls ( 22K, 6-28) respectively.

The dataset of vegetation cover fraction observations was obtained by the self-made instrument and the camera at a height of 2.5m-3.5m above the ground in the Yingke oasis, Huazhaizi desert steppe and Biandukou foci experimental areas on May 20, 24, 25, 28 and 30, Jun. 11, 14, 15, 21, 23, 24, 27 and 30, and Jul. 2, 2008. Observations were carried out in Yingke oasis maize field, Yingke oasis wheat field, Huazhaizi desert No. 1 and 2 plots, the rape field, the barley field and grassland in Biandukou. A pole with known length was put in each photo to determine the size of the photo. GPS data was used for the location and the technology LAB was used to retieve the coverage of the green vegetation. Besides, surrounding environment was also recorded. The dataset included the primary collected vegetation images and retrieved fraction of vegetation coverage.

The data set contains land cover data sets from the Yellow River Source, the Yangtze River Source, and the Lancang River from 1992 to 2015. A total of 22 land cover classifications based on the UN Land Cover Classification System were included. NOAA AVHRR, SPOT, ENVISAT, PROBA-V and other vegetation classification products were integrated. In China, (1) first, combined with the 1:100,000 vegetation classification (2007) of China, quality correction and control were performed, and (2) the vegetation classification of China emphasized the combination with climate zones, when correcting CCI-LC, climate divisions and the corresponding vegetation types were combined, and the data label was comprehensively revised.

This dataset is the LAI observation in the artificial oasis experimental region of the middle stream of the Heihe River Basin. The observation period is from 24 May to 20 September 2012 (UTC+8). Measurement instruments: LAI-2000 (Beijing Normal University) Measurement positions: Core Experimental Area of Flux Observation Matrix 18 corn samples, 1 orchard sample, 1 artificial white poplar sample Measurement methods: To measure the incoming sky radiation on the canopy firstly. Then the transmission sky radiation are mearued under the canopy for serveral times. The canopy LAI is retrieved by using the gap probability model.

The first dataset of ground truth measurements synchronizing with TerraSAR-X was obtained in the Daman foci experimental area on 4 June, 2012. The satellite image was in StripMap mode and HH/VV polarization with an incidence angle of 22-24°, and the overpass time was approximately at 19:00 UTC+8. The second dataset of ground truth measurements synchronizing with TerraSAR-X was obtained in the Daman foci experimental area on 15 June, 2012. The satellite image was in StripMap mode and HH/VV polarization with an incidence angle of 22-24°, and the overpass time was approximately at 19:00 UTC+8. The third dataset of ground truth measurements synchronizing with TerraSAR-X was obtained in the Daman foci experimental area on 26 June, 2012. The satellite image was in StripMap mode and HH/VV polarization with an incidence angle of 22-24°, and the overpass time was approximately at 19:00 UTC+8. The measurements were conducted at a sampling plot southeast to the Daman Superstation with an area of around 100 m × 100 m, which was dominantly planted with maize. Steven Hydro probes were used to collect soil moisture and other measurements with an interval of 5 m. For each sampling point, two measurements were acquired within an area of 1 m2, with one for the soil covered by plastic film (point name was tagged as LXPXXA) and the other for exposed soil (point name was tagged as LXPXXB). Concurrently with soil moisture sampling, vegetation properties were measured at around 10 locations within this sampling plot. Observation items included: Soil parameters: volumetric soil moisture (inherently converted from measured soil dielectric constant), soil temperature, soil dielectric constant, soil electric conductivity. Vegetation parameters: biomass, LAI, vegetation water content, canopy height, row distance and leaf chlorophyll content. Data and data format: This dataset includes two parts of measurements, i.e. soil and vegetation parameters. The former is as shapefile, with measured items stored in its attribute table. The measured vegetation parameters are recorded in an Excel file.

In order to understand the temporal and spatial variation characteristics of temperature changes in the Northern Hemisphere, the study used CRU (Climatic Research Unit) grid data to calculate the spatial distribution of the average annual temperature of 30 years (1971-2000). The annual average temperature decreases with the latitude increasing, and varies from greater than 30 °C to less than -25 °C. In the regions of the same latitudes, the annual average temperature in high altitude areas (such as the Tibetan Plateau, the Mongolian Plateau, and the Western Siberian Mountains) presented the trend of low temperature. At the same time, the annual average temperature trend distribution map of the Northern Hemisphere with a resolution of 0.5 ° × 0.5 ° from 1901 to 2016 was completed.

The dataset of ground truth measurement synchronizing with Envisat ASAR was obtained in the arid region hydrological experimental area on Sep. 19, 2007 during the pre-observation period. One scene of Envisat ASAR image was captured on Sep. 19. The data were in AP mode and VV/VH polarization combinations, and the overpass time was approximately at 11:29 BJT. Those provide reliable ground data for remote sensing retrieval and validation of soil moisture from Envisat ASAR image. Observation items included: (1) soil moisture measured by the cutting ring method in Linze reed land, Zhangye farmland, Zhangye gobi, Linze maize land, Linze alfalfa land, Zhangye weather station, and Linze wetland. (2) GPS measured by GARMIN GPS 76 (3) vegetation measurements including the vegetation height, the green weight, the dry weight, the sampling method, and descriptions on the land type, uniformity and dry and wet conditions (4) atmospheric parameters at Daman Water Management office measured by CE318 (produced by CIMEL in France). The total optical depth, aerosol optical depth, Rayleigh scattering coefficient, column water vapor in 936 nm, particle size spectrum and phase function were then retrieved from these observations. The optical depth in 1020nm, 936nm, 870nm, 670nm and 440nm were all acquired by CE318. Those data include the raw data in .k7 and can be opened by ASTPWin. ReadMetext files (.txt) is attached for detail. Processed data (after retrieval of the raw data) archived as Excel files are on optical depth, rayleigh scattering, aerosol optical depth, the horizontal visibility, the near surface air temperature, the solar azimuth, zenith, solar distance correlation factors, and air column mass number. (5) roughness measured by the roughness plate together with the digital camera. The coordinates of the sample would be got with the help of ArcView; and after geometric correction, surface height standard deviation (cm) and correlation length (cm) could be acquired based on the formula listed on pages 234-236, Microwave Remote Sensing (Vol. II). The roughness data were initialized by the sample name, which was followed by the serial number, the name of the file, standard deviation and correlation length. Each text files (.txt) file is matched with one sample photo and standard deviation and correlation length represent the roughness. In addition, the length of 101 radius is also included for further checking.

On 1 August 2012 (UTC+8), a Wide-angle Infrared Dual-mode line/area Array Scanner (WIDAS) carried by the Harbin Y-12 aircraft was used in a visible near Infrared thermal Dual-mode airborne remote sensing experiment, which is located in the upper reaches of the Heihe River Basin. WIDAS includes a CCD camera with a spatial resolution of 0.08 m, a visible near Infrared multispectral camera with five bands scanner (an maximum observation angle 48° and spatial resolution 0.4 m), and a thermal image camera with a spatial resolution of 2 m. The CCD camera data are recorded in DN values processed by mosaic and orthorectification. The mutispectral camera data production are recorded in reflectance processed by atmospheric and geometric correction. Thermal image camera data production are recorded in radiation brightness temperature processed by atmospheric and geometric correction.