The data include different observation data of Sunan, Gansu Province: 1) The soil properties of grassland under different management measures, soil compactness, water permeability and soil moisture content of 4-5 grazing intensity grassland; 2) The observation data of soil compactness, permeability and water content of different grazing management measures; 3) Correlation analysis data of grassland community characteristic productivity and soil moisture; 4) Correlation analysis data of height, coverage, biomass, flower shape, tiller and leaf characters of main plants with soil water content;

1 High resolution gridded West Antarctic surface mass balance dataset, its project is Polar Stereographic Projection 2. The kriging like interpolation method is used to reconstruct the high‐spatial resolution surface mass balance (SMB) over the West Antarctic Ice Sheet (WAIS) from 1800 to 2010, based on ice core records, the outputs of the European Centre for Medium‐Range Weather Forecasts “Interim” reanalysis (ERA‐Interim) as well as the latest polar version of the Regional Atmospheric Climate Model (RACMO2.3p2). 3. Its accuracy is higher than reanalysis data. 4. Temporal resolution: 1800-2010; Temporal resolution: 1 year; Spatial coverage : the whole West Antarctic Ice Sheet, Spatial resolution: 25km х 25km

The dataset is a 30-minute eddy covariance flux observation data from nine flux stations in the Three Poles, including the data of ecosystem Net Carbon Exchange (NEE), Gross Primary Productivity(GPP), and Ecosystem Respiration (ER) . The time coverage of the data is from 2000 to 2016. The main steps of data pre-processing include outlier removal (±3σ), coordinate axis rotation(three-dimensional wind rotation), Webb-Pearman-Leuning correction, outlier elimination, carbon flux interpolation and decomposition. And missing data is interpolated by the nonlinear empirical formula between CO2 flux value(Fc) and environmental factors.

At present, based on the proposed SAR ice sheet freeze-thaw detection algorithm using change detection and decision tree algorithm, the monthly average ice sheet freeze-thaw is detected using sentinel-1 EW SAR data. At the same time, using the developed production module of freeze-thaw products based on big data platform, the international first production of Antarctic ice sheet and Greenland ice sheet freeze-thaw products. Through the development of automatic weather station temperature data, the ice sheet freeze-thaw detection accuracy reaches 90%. At present, the acquisition time of data products is mainly the summer of the north and south poles, among which the Antarctic ice sheet products are January, February, March, October, November, December and Greenland products are may, June, July, August, September and October.

The coverage time of microwave radiometer ice sheet freeze-thaw data set is updated to 2016-2019, with a spatial resolution of 25 km; the remote sensing inversion method based on microwave radiometer adopts the improved wavelet based ice sheet freeze-thaw detection algorithm, which takes into account the change of ice sheet freeze-thaw brightness temperature characteristics in time. First, the long-time brightness temperature data of all ice sheet areas in Greenland is small by using wavelet transform. The multi-scale decomposition of wave is used to analyze the edge information at different scales. Thirdly, the edge information of ice sheet melting and refreezing is separated from the noise by ANOVA. Based on the extracted edge information of long-term brightness and temperature change of ice sheet, the optimal edge threshold of dry snow and wet snow classification is determined by using the generalized Gaussian model, so as to detect the melting area of Greenland ice sheet. Finally, based on the principle of space automatic error correction, the error results caused by noise are detected by using the space neighborhood error correction operator, and the error is corrected manually. The brightness and temperature data of passive microwave in long time series come from SMMR, SSM / I and SSMI / s sensors. To ensure simultaneous interpreting of the brightness temperature of different sensors, simultaneous interpreting of different sensor brightness temperatures is made before freezing and thawing. Through the verification of the actual measurement site, it shows that the detection accuracy of Greenland ice sheet freeze-thaw is more than 70%.

The microwave radiometer data set comprises brightness temperature data from SMMR (1978-1987), SSM/I (1987-2009) and SSMIS (2009-2015), with temporal coverage from 1978 to 2015 and a spatial resolution of 25 km. Each Antarctic data file consists of 316*332 grids, and each Arctic freeze-thaw data file consists of 304*448 grids. The microwave scatterometer data set comprises backscattering data from QScat (2000-2009) and ASCAT (2009-2015), with a temporal coverage from 2000 to 2015 and a spatial resolution of 4.45 km. Each Antarctic data file consists of 1940*1940 grids, and each Arctic data file consists of 810*680 grids. The temporal resolution of the data set is one day, and the data cover both Antarctica and Arctic ice sheets.

Because of its unique natural conditions and geographical location, the Arctic region plays a very important role in global change. Polar sea ice, as an important influencing factor of climate change, is a sensitive instrument of global climate change. The Yellow River Station, one of China's research stations in the Arctic, focuses on supporting the three scientific fields of global change and its regional response, the polar space environment and space climate, and the life characteristics and processes in the polar environment, providing an important platform for China's in-depth scientific research activities in the Arctic. Therefore, the product data set of data validation for key areas of Arctic sea ice in recent years is constructed to monitor the key areas of Arctic sea ice.

First of all, the data of ice cover elevation change is obtained by using the data of glas12 in 2004 and 2008. In ideal case, each track is strictly repeated. However, due to the track deviation, it can not be guaranteed that the track is strictly repeated according to the design. The deviation varies from several meters to several hundred meters. The grid of 500m * 500m is taken, and the point falling in the same grid is considered as the weight of the repeated track. The elevation change in 2004-2008 is obtained by subtraction of complex points, and the annual elevation change is obtained. Ice sheet elevation change data

Digital Elevation Model (DEM) is a kind of solid ground Model that represents the ground Elevation in the form of a set of ordered numerical arrays. The third pole region of40°1′52″N~23°11′59″N、105°43′45″E~61°28′45″E of the roof of the world ecological geographic area,These include the qinghai-tibet plateau, the hengduan mountains, the Himalayas, the Hindu kush mountains and the pamirs plateau.Classified according to：At 4000 m altitude as a benchmark, the fusion of slope, reference mountain integrity and ecological system integrity, the spatial resolution of 0.008 ° x 0.008 °

The three pole aerosol type data product is an aerosol type result obtained by integrating the data assimilation of Meera 2 and the active satellite CALIPSO product through a series of data preprocessing, quality control, statistical analysis and comparative analysis. The key of this algorithm is to judge the type of CALIPSO aerosol. In the process of aerosol type data fusion, according to the type and quality control of CALIPSO aerosol, and referring to the type of merra 2 aerosol, the final aerosol type data (12 kinds in total) and quality control results in the three pole area are obtained. The data product fully considers the vertical distribution and spatial distribution of aerosols, with high spatial resolution (0.625 ° × 0.5 °) and time resolution (month).

The daily cloudless MODIS Snow area ratio data set (2000-2015) of the Qinghai Tibet Plateau is based on MODIS daily snow product - mod10a1, which is obtained by using a cloud removal algorithm based on cubic spline interpolation. The data set is projected by UTM with spatial resolution of 500m, providing daily snow cover FSC results in the Tibetan Plateau. The data set is a day-to-day document, from 24 February 2000 to 31 December 2015. Each file is the result of snow area proportion on that day, the value is 0-100%, which is envi standard file, the naming rule is: yyyddd_fsc_0.5km.img, where yyyy represents the year, DDD represents Julian day (001-365 / 366). Files can be opened and viewed directly with envi or ArcMap. The original MODIS Snow data product for cloud removal comes from the mod10a1 product processed by the National Snow and Ice Data Center (NSIDC). This data set is in the format of HDF and uses the sinusional projection. The attributes of the daily cloudless MODIS Snow area ratio data set (2000-2015) on the Qinghai Tibet Plateau consist of the spatial-temporal resolution, projection information and data format of the data set. Temporal and spatial resolution: the temporal resolution is day by day, the spatial resolution is 500m, the longitude range is 72.8 ° ~ 106.3 ° e, and the latitude is 25.0 ° ~ 40.9 ° n. Projection information: UTM projection. Data format: envi standard format. File naming rules: "yyyyddd" + ". Img", where yyyy stands for year, DDD stands for Julian day (001-365 / 366), and ". Img" is the file suffix added for easy viewing in ArcMap and other software. For example, 2000055 ﹐ FSC ﹐ 0.5km.img represents the result on the 55th day of 2000. The envi file of this data set is composed of header file and body content. The header file includes row number, column number, band number, file type, data type, data record format, projection information, etc.; take 2000055 ﹣ FSC ﹣ 0.5km.img file as an example, the header file information is as follows: ENVI Description = {envi file, created [sat APR 27 18:40:03 2013]} Samples = 5760 Lines = 3300 Bands = 1 Header offset = 0 File type = envi standard Data type = 1: represents byte type Interleave = BSQ: data record format is BSQ Sensor type = unknown Byte order = 0 Map Info = {UTM, 1.500, 1.500, - 711320.359, 4526650.881, 5.0000000000e + 002, 5.0000000000e + 002, 45, north, WGS-84, units = meters} Coordinate system string = {projcs ["UTM [u zone [45N], geocs [" GCS [WGS [1984], data ["d [WGS [1984", organization ID ["WGS [1984", 6378137.0298.257223563]], prime ["Greenwich", 0.0], unit ["degree", 0.01745532925199433]]] project ["transfer [Mercator"]] parameter ["false [easting", 500000.0], parameter ["false [easting", 500000.0], parameter [500000.0], parameter [500000.0], parameter [false [false [easting ", 500000.0], parameter], parameter [500000.0], parameter [500000.0], parameter [500000.0], parameter [false [easting", 500000.0], parameter [500000.0], parameter [500000.0], parameter [500000.0], parameter ["false_northing", 0.0], parameter ["central_meridian", 87.0], parameter ["scale" _Factor ", 0.9996], parameter [" latitude ﹣ of ﹣ origin ", 0.0], unit [" meter ", 1.0]]} Wavelength units = unknown, band names = {2000055}

The data set includes 1. permaice (map of frozen soil types), 2. subsea (subsea boundary vectorgraph), 3. treeline (timberline vectorgraph), 4. nhipa (grid map) and 5. llipa (grid map). Permaice includes the following attribute fields: Num_code (frozen soil attribute code), Combo (frozen soil attribute), extent (frozen soil coverage) and content (ice content). The attribute comparison is as follows. (1) Frozen soil attribute comparison table: 0 （No information） 1 - chf （Continuous permafrost extent with high ground ice content and thick overburden） 2 - dhf （Discontinuous permafrost extent with high ground ice content and thick overburden） 3 - shf （Sporadic permafrost extent with high ground ice content and thick overburden） 4 - ihf （Isolated patches of permafrost extent with high ground ice content and thick overburden） 5 - cmf （Continuous permafrost extent with medium ground ice content and thick overburden） 6 - dmf （Discontinuous permafrost extent with medium ground ice content and thick overburden） 7 - smf （Sporadic permafrost extent with medium ground ice content and thick overburden） 8 - imf （Isolated patches of permafrost extent with medium ground ice content and thick overburden） 9 - clf （Continuous permafrost extent with low ground ice content and thick overburden） 10 - dlf （Discontinuous permafrost extent with low ground ice content and thick overburden） 11 - slf （Sporadic permafrost extent with low ground ice content and thick overburden） 12 - ilf （Isolated patches of permafrost extent with low ground ice content and thick overburden） 13 - chr （Continuous permafrost extent with high ground ice content and thin overburden and exposed bedrock） 14 - dhr （Discontinuous permafrost extent with high ground ice content and thin overburden and exposed bedrock） 15 - shr （Sporadic permafrost extent with high ground ice content and thin overburden and exposed bedrock） 16 - ihr （Isolated patches of permafrost extent with high ground ice content and thin overburden and exposed bedrock） 17 - clr （Continuous permafrost extent with low ground ice content and thin overburden and exposed bedrock） 18 - dlr （Discontinuous permafrost extent with low ground ice content and thin overburden and exposed bedrock） 19 - slr （Sporadic permafrost extent with low ground ice content and thin overburden and exposed bedrock） 20 - ilr （Isolated patches of permafrost extent with low ground ice content and thin overburden and exposed bedrock） 21 - g （Glaciers） 22 - r （Relict permafrost） 23 - l （Inland lakes） 24 - o （Ocean/inland seas） 25 - ld （Land） （2）The frozen soil coverage attribute comparison table c = continuous (90-100%) d = discontinuous (50-90%) s = sporadic (10-50%) i = isolated patches (0-10%) （3）The ice content comparison table h = high (>20% for "f" landform codes) (>10% for "r" landform codes) m = medium (10-20%) l = low (0-10%) ------------------------------------------------------------ Projection of the shapefiles is: PROJCS["Sphere_ARC_INFO_Lambert_Azimuthal_Equal_Area", GEOGCS["GCS_Sphere_ARC_INFO", DATUM["Sphere_ARC_INFO", SPHEROID["Sphere_ARC_INFO",6370997.0,0.0]], PRIMEM["Greenwich",0.0], UNIT["Degree",0.0174532925199433]], PROJECTION["Lambert_Azimuthal_Equal_Area"], PARAMETER["False_Easting",0.0], PARAMETER["False_Northing",0.0], PARAMETER["longitude_of_center",180.0], PARAMETER["latitude_of_center",90.0], UNIT["Meter",1.0]] Projection for the raster (*.byte) files is: Projection: Lambert Azimuthal Units: meters Spheroid: defined Major Axis: 6371228.00000 Minor Axis: 6371228.000 Parameters: radius of the sphere of reference: 6371228.00000 longitude of center of projection: 0 latitude of center of projection: 90 false easting (meters): 0.00000 false northing (meters): 0.00000