Li et al. (2026) Classification of paddy field water conditions with unmanned aerial vehicle: A novel method incorporating the short-wave infrared band

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-07-16
• Authors: Qilin Li, Yulong Dai, Meng Zhang, Fangping Liu, Yuanlai Cui, Yufeng Luo
• DOI: 10.1016/j.agwat.2026.110626

Research Groups

• State Key Laboratory of Water Resources Engineering and Management, Wuhan University
• Ganfu Plain Water Conservancy Engineering Administration Bureau
• Jiangxi Provincial Irrigation Experimental Center Station

Short Summary

This study presents a novel method for classifying paddy field water conditions (PFWC) using unmanned aerial vehicle (UAV) remote sensing data. The proposed approach incorporates the short-wave infrared (SWIR) band and achieves satisfactory classification performance, extending the operational monitoring window compared to traditional thermal infrared (TIR) methods.

Objective

• Identify an effective information source for classifying PFWC into four classes: flooded, partially flooded, non-flooded but not yet drought-stressed, and already drought-stressed.
• Develop a UAV-based PFWC classification method that can be applied under different management practices and crop physiological conditions.

Study Configuration

• Spatial Scale: Local scale (paddy fields in the Ganfu Plain Irrigation Area)
• Temporal Scale: Diurnal scale (9:00–15:00 time window)

Methodology and Data

• Models used: Support Vector Machine (SVM), Random Forest (RF), Backpropagation Neural Network (BPNN), and Extreme Learning Machine (ELM)
• Data sources: Multispectral (MS) including SWIR, thermal infrared (TIR) including SWIR

Main Results

• The proposed method achieved an average Accuracy of 0.918 and Macro-F1 of 0.899 for the optimal model (SVM) across three growth stages.
• Using MS including SWIR as input to SVM is an effective UAV-based PFWC classification method that extends the operational monitoring window compared with TIR.

Contributions

• This study addresses the limitations of previous UAV-based field water condition monitoring methods, which relied on multi-source information including TIR and had low data acquisition and processing efficiency.
• The proposed approach provides a useful tool for intelligent irrigation forecasting in paddy irrigation areas.

Funding

• This research was funded by the National Natural Science Foundation of China (Grant No. 52109263)

Citation

@article{Li2026Classification,
  author = {Li, Qilin and Dai, Yulong and Zhang, Meng and Liu, Fangping and Cui, Yuanlai and Luo, Yufeng and Hao, Jia},
  title = {Classification of paddy field water conditions with unmanned aerial vehicle: A novel method incorporating the short-wave infrared band},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110626},
  url = {https://doi.org/10.1016/j.agwat.2026.110626}
}