Li et al. (2026) Complex network-constrained regional estimation of crop evapotranspiration using ERA5-Land data at meteorological station locations

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-10-05
• Authors: Yingxue Li, Jun Niu, Bellie Sivakumar, 郑荣, Xiaotao Zhang
• DOI: 10.1016/j.agwat.2026.110837

Research Groups

• State Key Laboratory of Efficient Utilization of Agricultural Water Resources, China Agricultural University
• National Field Observation and Research Station (Gansu Wuwei) for Efficient Water Utilization in Oasis Agriculture
• Center for Agricultural Water Research in China, China Agricultural University
• Department of Civil Engineering, Indian Institute of Technology Bombay
• Jiuquan Academy of Agricultural Sciences
• State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin, China Institute of Water Resources and Hydropower Research

Short Summary

This study developed a regional crop evapotranspiration (ETc) estimation framework that incorporates meteorological station network topology as a data-derived topological constraint on the trend component of the ETc field. The framework was applied to seed maize in the Hexi Corridor, China, and showed improved accuracy compared to conventional ordinary kriging.

Objective

• Develop a regional ETc estimation framework constrained by meteorological station network topology for application in arid regions with complex terrain and fragmented underlying surfaces.

Study Configuration

• Spatial Scale: Regional scale (Hexi Corridor, China)
• Temporal Scale: Daily to dekadal timescales (1994–2024)

Methodology and Data

• Models used: Penman-based ordinary kriging (P-OK), genetic-algorithm-optimized ordinary kriging (P-GA-OK), complex-network-constrained genetic-algorithm ordinary kriging (P-CN-GA-OK)
• Data sources: ERA5-Land climate reanalysis data, meteorological station observations

Main Results

• The P-CN-GA-OK model showed improved accuracy compared to P-OK and P-GA-OK models in estimating ETc at unsampled locations.
• The model reduced the pooled RMSE by 10.2% relative to P-OK and 5.4% relative to P-GA-OK.
• The dekad-averaged R² was raised from 0.880 to 0.905.

Contributions

• This study provides a practical option for regional water-consumption assessment in irrigated arid regions with complex terrain and fragmented underlying surfaces.
• The framework can be applied to other regions with similar characteristics.

Funding

• This research was funded by the National Natural Science Foundation of China (Grant No. 51909241) and the China Agricultural University Research Fund (Grant No. 2020XKQG02).

Citation

@article{Li2026Complex,
  author = {Li, Yingxue and Niu, Jun and Sivakumar, Bellie and 郑荣 and Zhang, Xiaotao},
  title = {Complex network-constrained regional estimation of crop evapotranspiration using ERA5-Land data at meteorological station locations},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110837},
  url = {https://doi.org/10.1016/j.agwat.2026.110837}
}