TY  - Data
T1  - 1km high-resolution monthly-scale GWSA dataset
A1  - Yifei Fan
A1  - Chen Zhang
A1  - han wen xia
A1  - CHE Tao 
A1  - Gai yingchun
DO  - 10.12072/ncdc.hydrology.db7323.2026
PY  - 2026
DA  - 2026-05-15
PB  - National Cryosphere Desert Data Center
AB  - This dataset provides globally downscaled monthly groundwater storage anomaly (Groundwater Storage Anomaly, GWSA) data for global land areas. The dataset has a spatial resolution of 0.008333° (~1 km at the equator), covering the period from April 2002 to December 2020 with a monthly temporal resolution.Data format: The dataset is released in multi-band GeoTIFF format with the coordinate reference system of WGS 1984 (EPSG:4326). The data can be visualized and analyzed using geographic information processing software such as ArcGIS and QGIS.File naming convention: Each multi-band raster is named in the format of “GWSA_YYYY” (e.g., “GWSA_2002”), representing the downscaled results for the corresponding year. Monthly information is stored as raster bands arranged in chronological order.Data production: Monthly terrestrial water storage anomalies (Terrestrial Water Storage Anomaly, TWSA) were derived from the GRACE/GRACE-FO RL06 products released by the Center for Space Research (CSR), University of Texas, and missing values in the time series were reconstructed using Singular Spectrum Analysis (SSA). Soil moisture storage, snow water equivalent, and canopy water storage were extracted from the NASA Global Land Data Assimilation System Version 2.1 (GLDAS Noah v2.1) and converted into anomaly values relative to the 2004–2009 baseline period. Monthly GWSA at the coarse spatial resolution (0.25°) was then calculated based on the vertical water balance equation. Subsequently, 19 variables including topography, soil properties, climate, hydrology, and vegetation dynamics were used as high-resolution (0.008333°) predictors. XGBoost-based downscaling models were independently developed for different aquifer regions to generate the final 0.008333° GWSA dataset.Data accuracy: Cross-validation against the original coarse-resolution GWSA data yielded an overall accuracy of R² = 0.97 and RMSE = 2.10 cm. Validation against 1,518 independent groundwater observation wells worldwide ach
DB  - NCDC
UR  - http://www.ncdc.ac.cn/portal/metadata/6e5c2ad4-4abb-4c21-9a2f-1fbc658eb54b
ER  -