Zhang et al. (2025) A novel framework for pixel-wise estimation of irrigation water use by integrating remote sensing and reanalysis data

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-12-13
• Authors: Ling Zhang, Tao Che, Kun Zhang, Donghai Zheng, Xin Li
• DOI: 10.1016/j.agwat.2025.110077

Research Groups

• Key Laboratory of Cryospheric Science and Frozen Soil Engineering, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences (CAS).
• Heihe Remote Sensing Experimental Research Station, Chinese Academy of Sciences (CAS).
• School of Geospatial Engineering and Science, Sun Yat-Sen University.
• National Tibetan Plateau Data Center, State Key Laboratory of Tibetan Plateau Earth System, Resource and Environment, Institute of Tibetan Plateau Research, Chinese Academy of Sciences (CAS).

Short Summary

This study develops a novel soil water balance framework that integrates satellite-derived soil moisture and evapotranspiration with reanalysis data to estimate irrigation water use at 1 km resolution across China. The resulting 20-year dataset reveals that China's irrigation water use increased from 339 to 395 km³/year between 2001 and 2020, driven primarily by the expansion of irrigated croplands.

Objective

• To develop a spatially explicit, physically based framework for pixel-wise estimation of irrigation water use (IWU) by integrating multi-source remote sensing and reanalysis data.
• To evaluate the performance of root zone soil moisture (RSM) versus surface soil moisture (SSM) balance models in capturing irrigation signals.

Study Configuration

• Spatial Scale: Mainland China at 1 km grid resolution.
• Temporal Scale: 2001–2020 (monthly and annual intervals).

Methodology and Data

• Models used: Two alternative soil water balance models (RSM-based and SSM-based) that quantify IWU as the residual between satellite-observed (irrigated) and reanalysis-modeled (natural/rainfed) soil moisture, evapotranspiration (ET), and drainage.
• Data sources:
  • Reanalysis: ERA5-Land (soil moisture and ET for natural conditions).
  • Satellite ET: MODIS (MOD16A2GF) and PML-V2.
  • Satellite SSM: AMSR-D and ESA-CCI-D (1 km downscaled products).
  • Auxiliary: ChinaMet (1 km precipitation), CIrrMap250 (annual irrigated area maps), and prefecture-level IWU statistical reports for calibration.
• Ensemble Approach: Integration of multiple product combinations using Summed Square Error (SSE) and Bayesian Three-Cornered Hat (BTCH) weighting methods.

Main Results

• Model Performance: Both models performed well during validation (2010–2020), with $R^2$ values between 0.72 and 0.90 and RMSE of 0.55–0.66 km³/year. The SSM-based model generally achieved higher $R^2$ than the RSM-based model.
• National Trends: China’s total IWU rose from 339 km³/year in 2001 to 395 km³/year in 2020, with an average of 360 km³/year.
• Drivers of Change: The increasing trend was primarily driven by an 18 million hectare expansion in irrigated area, while interannual variability was controlled by water use intensity (averaging ~425 mm/year).
• Comparison: The proposed framework consistently outperformed the Random Forest (RF) machine learning algorithm, particularly when calibration data was limited or at coarse spatial scales.

Contributions

• Methodological Integration: Unlike previous studies that rely on either ET or soil moisture alone, this framework integrates both to capture a more complete irrigation signal.
• High-Resolution Dataset: Provides the first 1 km resolution, 20-year IWU dataset for China that accounts for dynamic changes in irrigated areas.
• Improved Accuracy: Demonstrates significantly higher accuracy than existing global products (e.g., IWU-Huang, IWU-Cheng, and IWU-Zhang) when validated against independent sub-national statistical reports.

Funding

• National Key Research and Development Program of China (2023YFF0804901).
• National Natural Science Foundation of China (42271286).
• Youth Innovation Promotion Association of Chinese Academy of Sciences (2023454).
• Key Research Program of Gansu Province (23ZDKA0004).

Citation

@article{Zhang2025novel,
  author = {Zhang, Ling and Che, Tao and Zhang, Kun and Zheng, Donghai and Li, Xin},
  title = {A novel framework for pixel-wise estimation of irrigation water use by integrating remote sensing and reanalysis data},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.110077},
  url = {https://doi.org/10.1016/j.agwat.2025.110077}
}

Generated by BiblioAssistant using gemini-3-flash-preview (Google API)