This data set is a 21 year (1988-2008) surface soil moisture data set in the Qinghai Tibet Plateau. The temporal resolution is yearly, the spatial resolution is 25km, and the data unit is m3 / m3. Based on the retrieval of plateau soil moisture by van der Velde et al. (2014), the data set was generated by using three-dimensional discrete transformation method to make up for the lack of value. The data has been verified by the site, and compared with the reanalysis data, it is found that the data quality is better. The data can be used to study the spatiotemporal variation of soil moisture in spring.

China's daily snow depth simulation and prediction data set is the estimated daily snow depth data of China in the future based on the nex-gdpp model data set. The artificial neural network model of snow depth simulation takes the maximum temperature, minimum temperature, precipitation data and snow depth data of the day as the input layer of the model, The snow depth data of the next day is used as the target layer of the model to build the model, and then the snow depth simulation model is trained and verified by using the data of the national meteorological station. The model verification results show that the iterative space-time simulation ability of the model is good; The spatial correlations of the simulated and verified values of cumulative snow cover duration and cumulative snow depth are 0.97 and 0.87, and the temporal and spatial correlations of cumulative snow depth are 0.92 and 0.91, respectively. Based on the optimal model, this model is used to iteratively simulate the daily snow depth data in China in the future. The data set can provide data support for future snow disaster risk assessment, snow cover change research and climate change research in China. The basic information of the data is as follows: historical reference period (1986-2005) and future (2016-2065), as well as rcp4.5 and rcp8.5 scenarios and 20 climate models. Its spatial resolution is 0.25 ° * 0.25 °. The projection mode of the data is ease GR, and the data storage format is NC format. The following is the data file information in NC Time: duration (unit: day) Lon = 320 matrix, 320 columns in total Lat = 160 matrix, 160 rows in total X Dimension: Xmin = 60.125; // Coordinates of the corner points of the lower left corner grid in the X direction of the matrix Y Dimension: Ymin = 15.125; // Coordinates of the corner points of the grid at the lower left corner of the Y-axis of the matrix

As an important parameter of snowpack, the snow depth can adjust changes in sea ice and plays a vital role in the climate system. This dataset provides the daily snow depth on Arctic sea ice during cold-season from 2012 to 2020 (October-April). The snow depth retrieval algorithm combines linear regression and deep learning algorithm (LSTM). Firstly, according to AMSR2 brightness temperature data and the buoy snow depth data, the optimal gradient ratio for estimating the snow depth on first-year ice (FYI) is determined, and the snow depth retrieval model on FYI is established. The input of LSTM is different gradient rates, polarization ratio (PR) of the brightness temperature at 37 GHz, and the output is snow depth on sea ice. Then, the snow depth retrieval model on multi-year ice (MYI) is determined. Finally, the snow depth on Arctic sea ice with a 25-km spatial resolution is obtained. This dataset performs well in reflecting snow depths over thick and thin sea ice, which agrees well with the NASA Operation IceBridge (OIB) data, with the root mean square error of 7.35 cm. Hence, this dataset provides accurate snow depth estimates for retrieving sea ice thickness and is conducive to analyses of Arctic mass balance and energy balance.

We release three data products for monthly groundwater storage (GWS) changes in North America. In the first one, we provide an independent estimate for monthly GWS changes within North America in 1-degree-grids and their trends. In the second one, we give the monthly GWS changes and the trends averaged for the 5 major GWS trend anomalies in around Saskatchewan, Nevada, California, Arizona and Texas, respectively. The third data product includes the monthly GWS changes and the trends averaged for the 14 states or provinces in the US and Canada, affected by the above GWS trend anomalies, i.e., for Saskatchewan, Montana, Nevada, California, Arizona, New Mexico, Texas, Oklahoma, Kansas, Alberta, North Dakota, Minnesota, Colorado and Chihuahuas, respectively. Our estimates of monthly GWS changes and their trends can serve as alternative and beneficial input for sustainable management of groundwater resources in North America. Our data products are derived from the release-6 version of GRACE monthly level-2 data, GNSS data, two land surface models of GLDAS 2.1 for soil moisture and snow water equivalent, and satellite altimetric lake level data. Unlike previous studies, glacial isostatic adjustment (GIA) effects are eliminated by employing an independent separation approach with the aid of GNSS vertical velocity data (Wang et al., 2013). The monthly changes of those GWS anomalies are validated by well level data. The monthly GWS changes for the 14 states or provinces are basically to show compatible variations with precipitation drought intensity level variations. We find a GWS increasing trend anomaly in Saskatchewan and 4 GWS declining trend anomalies with peaks in Nevada, California, Arizona and Texas, respectively. As they are not estimated using GIA models in the correction and their comparison with available well level and drought data confirms their reliability, we suggest our data products as alternative input to groundwater resource management in the discussed areas.

Land surface temperature (LST) is a key parameter in the study of surface energy balance. It is widely used in the fields of meteorology, climate, hydrology, agriculture and ecology. As an important means to obtain global and regional scale LST information, satellite (thermal infrared) remote sensing is vulnerable to the influence of cloud cover and other atmospheric conditions, resulting in temporal and spatial discontinuity of LST remote sensing products, which greatly limits the application of LST remote sensing products in related research fields. The preparation of this data set is based on the empirical orthogonal function interpolation method, using Terra / Aqua MODIS surface temperature products to reconstruct the lst under ideal clear sky conditions, and then using the cumulative distribution function matching method to fuse era5 land reanalysis data to obtain the lst under all-weather conditions. This method makes full use of the spatio-temporal information of the original MODIS remote sensing products and the cloud impact information in the reanalysis data, alleviates the impact of cloud cover on LST estimation, and finally reconstructs the high-quality global 0.05 ° spatio-temporal continuous ideal clear sky and all-weather LST data set. This data set not only realizes the seamless coverage of space-time, but also has good verification accuracy. The reconstructed ideal clear sky LST data in the experimental areas of 17 land cover types in the world, the average correlation coefficient (R) is 0.971, the bias (bias) is -0.001 K to 0.049 K, and the root mean square error (RMSE) is 1.436 K to 2.688 K. The verification results of the reconstructed all-weather LST data and the measured data of ground stations: the average R is 0.895, the bias is 0.025 K to 2.599 K, and the RMSE is 4.503 K to 7.299 K. The time resolution of this data set is 4 times a day, the spatial resolution is 0.05 °, the time span is 2002-2020, and the spatial range covers the world.

The data include oxygen isotope data and core age data of Jiang Co Lake in Qinghai Tibet Plateau. The first column is age, and the second column is oxygen isotope value; This data records the changes of oxygen isotopes in the past 84-2015.

This data set is the global high accuracy global elevation control point dataset, including the geographic positioning, elevation, acquisition time and other information of each elevation control point. The accuracy of laser footprint elevation extracted from satellite laser altimetry data is affected by many factors, such as atmosphere, payload instrument noise, terrain fluctuation in laser footprint and so on. The dataset extracted from the altimetry observation data of ICESat satellite from 2003 to 2009 through the screening criteria constructed by the evaluation label and ranging error model, in order to provide global high accuracy elevation control points for topographic map or other scientific fields relying on good elevation information. It has been verified that the elevation accuracy of flat (slope<2°), hilly (2°≤slope<6°), and mountain (6°≤slope<25°) areas meet the accuracy requirements of 0.5m, 1.5m, and 3m respectively.

Central Asia (referred to as CA) is among the most vulnerable regions to climate change due to the fragile ecosystems, frequent natural hazards, strained water resources, and accelerated glacier melting, which underscores the need of high-resolution climate projection datasets for application to vulnerability, impacts, and adaption assessments. We applied three bias-corrected global climate models (GCMs) to conduct 9-km resolution dynamical downscaling in CA. A high-resolution climate projection dataset over CA (the HCPD-CA dataset) is derived from the downscaled results, which contains four static variables and ten meteorological elements that are widely used to drive ecological and hydrological models. The static variables are terrain height (HGT, m), land use category (LU_INDEX, 21 categories), land mask (LANDMASK, 1 for land and 0 for water), and soil category (ISLTYP, 16 categories). The meteorological elements are daily precipitation (PREC, mm/day), daily mean/maximum/minimum temperature at 2m (T2MEAN/T2MAX/T2MIN, K), daily mean relative humidity at 2m (RH2MEAN, %), daily mean eastward and northward wind at 10m (U10MEAN/V10MEAN, m/s), daily mean downward shortwave/longwave flux at surface (SWD/LWD, W/m2), and daily mean surface pressure (PSFC, Pa). The reference and future periods are 1986-2005 and 2031-2050, respectively. The carbon emission scenario is RCP4.5. The results show the data product has good quality in describing the climatology of all the elements in CA, which ensures the suitability of the dataset for future research. The main feature of projected climate changes in CA in the near-term future is strong warming (annual mean temperature increasing by 1.62-2.02℃) and significant increase in downward shortwave and longwave flux at surface, with minor changes in other elements. The HCPD-CA dataset presented here serves as a scientific basis for assessing the impacts of climate change over CA on many sectors, especially on ecological and hydrological systems.

Surface soil moisture (SSM) is a crucial parameter for understanding the hydrological process of our earth surface. Passive microwave (PM) technique has long been the primary choice for estimating SSM at satellite remote sensing scales, while on the other hand, the coarse resolution (usually >~10 km) of PM observations hampers its applications at finer scales. Although quantitative studies have been proposed for downscaling satellite PM-based SSM, very few products have been available to public that meet the qualification of 1-km resolution and daily revisit cycles under all-weather conditions. In this study, therefore, we have developed one such SSM product in China with all these characteristics. The product was generated through downscaling of AMSR-E and AMSR-2 based SSM at 36-km, covering all on-orbit time of the two radiometers during 2003-2019. MODIS optical reflectance data and daily thermal infrared land surface temperature (LST) that have been gap-filled for cloudy conditions were the primary data inputs of the downscaling model, in order to achieve the “all-weather” quality for the SSM downscaling outcome. Daily images from this developed SSM product have achieved quasi-complete coverage over the country during April-September. For other months, the national coverage percentage of the developed product is also greatly improved against the original daily PM observations. We evaluated the product against in situ soil moisture measurements from over 2000 professional meteorological and soil moisture observation stations, and found the accuracy of the product is stable for all weathers from clear sky to cloudy conditions, with station averages of the unbiased RMSE ranging from 0.053 vol to 0.056 vol. Moreover, the evaluation results also show that the developed product distinctly outperforms the widely known SMAP-Sentinel (Active-Passive microwave) combined SSM product at 1-km resolution. This indicates potential important benefits that can be brought by our developed product, on improvement of futural investigations related to hydrological processes, agricultural industry, water resource and environment management.

The maximum freezing depth is an important index of the thermal state of seasonal frozen ground. Due to global warming, the maximum freezing depth of seasonal frozen ground continues to decline. The maximum freezing depth data set of five provinces in Northwest China, Tibet and surrounding areas from 1961 to 2020 was released, with a spatial resolution of 1 km. The data set is a support vector regression (SVR) model based on the measured data of maximum freezing depth from 2001 to 2010 and spatial environmental variables, which simulates the maximum freezing depth in Northwest China, Tibet and surrounding areas from 1961 to 2020. The validation results show that the SVR model has good spatial generalization ability, and there is a high consistency between the predicted value and the observed value of the maximum soil freezing depth. The determination coefficients of the simulation results in the four periods of 1980s, 1990s, 2000s and 2010s are 0.77, 0.83, 0.73 and 0.71 respectively. The percentile range of the prediction results shows that the simulation results have good stability. Based on this data set, it is found that the maximum soil freezing depth in Northwest China continues to decline, among which Qinghai has the fastest decline rate, with an average decline of 0.53 cm every decade. The data set provides data support for the study of seasonal frozen soil in Northwest China, High Mountain Asia and the Third Pole.

CAMELE: Collocation-Analyzed Multi-Source Ensembled Land Evapotranspiration data provide an estimation of global land total evapotranspiration at 0.1°-8daily and 0.25°-daily resolutions. The 0.1°-8daily collection covers the period from 20010101 to 20190829, while the 0.25°-daily provides the estimation from 19810101 to 20200831. TCA-based algorithms are used to evaluate the uncertainties and the error cross-correlation value of five widely used global land evapotranspiration products, including ERA5-land total evaporation, FLUXCOM-RS, PMLV2 (Penman-Monteith-Leuning model version 2 global evaporation), GLEAM v3.3a and GLDASv2.1 Noah. By minimizing the mean square error, the optimal weights of each product for linear combination are given using the evaluation results. Multiple information including the core collection method, synthetic experiments, site-based validation and evaluation of the merging data were described in our paper.