China Grid-level Natural Streamflow Estimates (CHASE v1.0): Daily Version

This record contains daily flow records of CHASE v1.0. For monthly and yearly averaged records, please visit https://doi.org/10.5281/zenodo.17384262 Overview CHinA grid-level natural Streamflow Estimates (CHASE v1.0) provides 1-km gridded estimates of natural streamflow across China, reconstructed using a coupled land-surface-hydrologic-hydrodynamic modeling system. It spans the period from January 1, 1962, to December 31, 2024, at a daily temporal resolution and 1-km spatial resolution. Validation against 1,225 flow gauges shows median daily Nash-Sutcliffe Efficiency (NSE) of 0.57, monthly NSE of 0.77, daily Kling-Gupta Efficiency (KGE) of 0.65, and monthly KGE of 0.73. Lake water level simulations are also satisfactory in major freshwater systems like Poyang, Dongting, Tai, Chao, and Hongze Lakes. Request Access By requesting access, you agree to the following conditions: You are a researcher requesting access from a research-affiliated institution. You will use the dataset only for research purposes and will not redistribute it. In your request, please provide your full name, affiliation, and relevant funding information, following the example below: Name: Ningpeng Dong Affiliation: China Institute of Water Resources and Hydropower Research Funding Information: National Natural Science Foundation of China (42401053). Data Files The dataset consists of NetCDF files. Spatial Coverage: China, bounded approximately by latitudes 18°N to 53°N and longitudes 73°E to 135°E. Grid dimensions: 4087 rows (Y) × 4780 columns (X). Temporal Coverage: Daily timesteps, with 365 or 366 days per file depending on leap years. Resolution: 1 km × 1 km (nominal grid cell size). Variables time (double, dimension: time): Time. XLONG (float, dimensions: Y, X): Longitude of each grid cell. XLAT (float, dimensions: Y, X): Latitude of each grid cell. streamflow (int, dimensions: time, Y, X): Natural streamflow rate. Units: m³/s; Scale factor: 0.1 (multiply stored values by 0.1 to get actual m³/s); Add offset: 0; Fill/missing value: -9999. Data are packed as integers for efficiency. Usage Notes Processing: Files can be read using standard NetCDF libraries. To unpack streamflow values: actual_value = (stored_value * scale_factor) + add_offset. Negative or fill values indicate no data (e.g., outside the model domain). Limitations: Data are model-derived and assume natural conditions; they do not include human interventions. Users should consult the associated papers for detailed validation metrics. Reference Ningpeng Dong, Mingxiang Yang, Jianhui Wei, et al. Reconstructing China's Natural Streamflow at 1 km resolution. Authorea. October 24, 2025. DOI: 10.22541/au.176132833.37694469/v1 For questions, contact: Ningpeng Dong ([e-mail]).