CIWW1km: China High-Resolution Monthly Irrigation Water Withdrawal Dataset (2000–2020)
收藏资源简介：
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: zhanglingky@lzb.ac.cn or chetao@lzb.ac.cn· 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
CIWW1km数据集基于创新的物理约束估计框架开发。该框架整合多源遥感观测（包括蒸散发 (evapotranspiration) 与土壤湿度 (soil moisture) 产品）与再分析气象数据 (reanalysis meteorological data)，并以土壤水量平衡原理 (soil water balance principle) 为分析基础，实现了灌溉取水量组分的物理一致性估计（Zhang, Che*, et al., 2026）。在此基础上，研究引入可解释机器学习 (interpretable machine learning) 方法优化物理模型输出的时空表征能力，进而提升整体估计精度（Zhang* et al., 2025）。 CIWW1km数据集系统刻画了近二十年间中国灌溉取水量的时空演变特征，为水资源管理、农业规划以及人类影响下水文循环相关研究提供了关键的数据基础。 该数据集以像元尺度 (pixel scale) 提供两类核心变量：灌溉取水深（单位：毫米，即单位面积用水量）与灌溉取水总量（单位：立方千米，即单个栅格单元 (grid cell) 的总用水量）。数据以.tif格式提供。文件命名规范示例如下： • CIWW_1km_depth_mm_2000_v2.tif：代表2000年的灌溉取水深，"v2"表示版本号。 • CIWW_1km_depth_mm_2000_01_v2.tif：代表2000年1月的灌溉取水深。 • CIWW_1km_depth_km3_2000_v2.tif：代表2000年的灌溉取水总量。 • CIWW_1km_depth_km3_2000_01_v2.tif：代表2000年1月的灌溉取水总量。 如有任何疑问，请联系：zhanglingky@lzb.ac.cn 或 chetao@lzb.ac.cn • Zhang Ling*, Ma Hui, Hu Yingyi, Wang Yixiao, Ma Qimin, Zhao Yanbo. 知识引导机器学习与多源数据在空间显式灌溉取水量估计中的应用. 《地球的未来》. 2025, 13(12): e2025E-e6704E. • Zhang Ling, Che Tao*, Zhang Kun, Zheng Donghai, Li Xin. 融合遥感与再分析数据的像元级灌溉用水量估计新框架. 《农业水管理》. 2026, 323: 110077
