ChinaWR: a high-resolution and long-term gridded water resources dataset for China based on machine learning

We develop a new water resources dataset for China, referred to as ChinaWR, providing 1 km and monthly estimates of surface, groundwater, and total water resources (SWR, GWR, and TWR) from 1980 to 2020. ChinaWR was generated using a random forest model based on high-resolution (prefecture-level) data reported in the Water Resources Bulletins of China.All the data are stored in NetCDF files. The file naming format follows the template ChinaWR_VV_SSSdeg_YYYYMM.nc (e.g., ChinaWR_SWR_010deg_198001.nc), where VV represents the variable name, SSS indicates the spatial resolution (with the decimal point omitted), YYYY denotes the four-digit year, and MM represents the two-digit month. For more detailed information, please refer to the attributes of each NetCDF file.Any questions please feel free to contact: [e-mail]