A 0.1° terrestrial water storage anomaly dataset over China

This repository provides a 0.1° monthly terrestrial water storage anomaly (TWSA) dataset over China together with the the corresponding training datasets. The downscaled TWSA is generated through a deep learning framework, enforcing agreement with GRACE TWSA at the patch scale and with physically informed patterns at the pixel scale. Two versions of the dataset are included: (1) Standard version (2002–2019): This version is trained and validated by a comprehensive set of observations. Satellite-based glacier mass balance is included to constrain the results. (2) Extended version (2020–2023): This version extends the dataset to support more recent hydrologic studies. Due to the lack of satellite-based glacier mass balance products with sufficient spatial coverage during this period, the extended dataset is recommended for analysis in non-glacierized regions. All datasets are provided in NetCDF (.nc) format. Compared with Version 1, Version 2 additionally includes the complete training datasets used for model development and reproducibility. Please refer to README.md for detailed descriptions of the datasets, variables, metadata, and file organization.