CIWW1km: China High-Resolution Monthly Irrigation Water Withdrawal Dataset (2000–2020)

The CIWW1km dataset was developed based on an innovative physics-constrained estimation framework. By integrating multi-source remote sensing observations (including evapotranspiration and soil moisture products) with reanalysis meteorological data, and grounding the analysis in the soil water balance principle, this framework achieves physically consistent estimates of irrigation water withdrawal components (Zhang, Che*, et al., 2026). Building upon this, interpretable machine learning methods were introduced to optimize the spatiotemporal representation of the physical model outputs, thereby enhancing the overall estimation accuracy (Zhang* et al., 2025). CIWW1km systematically characterizes the spatiotemporal evolution of irrigation water withdrawal in China over the past two decades, providing a critical data foundation for water resources management, agricultural planning, and research on the hydrological cycle under human influence.
The dataset provides two key variables at the pixel scale: Irrigation Water Withdrawal Depth (mm, water use per unit area) and Irrigation Water Withdrawal Volume (km³, total water use per grid cell). The data is provided in .tif format. Examples of the file naming convention are as follows:
• CIWW_1km_depth_mm_2000_v2.tif (Represents the irrigation water withdrawal depth for the year 2000; "v2" indicates the version number.)
• CIWW_1km_depth_mm_2000_01_v2.tif (Represents the irrigation water withdrawal depth for January 2000.)
• CIWW_1km_depth_km3_2000_v2.tif (Represents the irrigation water withdrawal volume for the year 2000.)
• CIWW_1km_depth_km3_2000_01_v2.tif (Represents the irrigation water withdrawal volume for January 2000.)
Any questions please feel free to contact: [e-mail] or [e-mail]
· Zhang Ling*, Ma Hui, Hu Yingyi, Wang Yixiao, Ma Qimin, Zhao Yanbo. On the Use of Knowledge-Informed Machine Learning and Multisource Data for Spatially Explicit Estimation of Irrigation Water Withdrawal. Earth's Future. 2025. 13(12). e2025E-e6704E.
· Zhang Ling, Che Tao*, Zhang Kun, Zheng Donghai, Li Xin. A novel framework for pixel-wise estimation of irrigation water use by integrating remote sensing and reanalysis data. Agricultural Water Management. 2026. 323. 110077