Yan et al. (2025) Dynamically Updated Irrigation Canal Scheduling Rules Based on Risk Hedging

Identification

• Journal: Agriculture
• Year: 2025
• Date: 2025-12-05
• Authors: Ming Yan, Fengyan Wu, L. H. Chen, Yong Liu, Xiang Zeng
• DOI: 10.3390/agriculture15242527

Research Groups

• State Key Laboratory of Water Resources Engineering and Management, Wuhan University, Wuhan, China
• Hubei International Irrigation and Drainage Research and Training Center, Hubei Water Resources Research Institute, Wuhan, China
• Department of Hydraulic Engineering, Hubei Water Resources Technical College, Wuhan, China
• National Engineering Research Center of Eco-Environment in the Yangtze River Economic Belt, China Three Gorges Corporation, Wuhan, China
• School of Resource and Environmental Sciences, Wuhan University, Wuhan, China

Short Summary

This study develops a novel "Bi-level, Two-stage" (BT) model for dynamically updated irrigation canal scheduling, integrating historical data-derived Target Residual Lump-Sum Water Quota (TRLSWQ) and hydrometeorological forecasts. The BT model significantly improves irrigation efficiency and water utilization by reducing water shortage indices and increasing water quota utilization compared to conventional methods, effectively hedging against future water shortage risks.

Objective

• To explore and propose a dynamically updated canal scheduling framework that reduces current-period water allocation to increase remaining-period allocation, hedging against future water shortage risks.
• To study the balance strategy between current and remaining-period water shortage risks.
• To establish a new dynamically updated canal scheduling model integrating hydrometeorological forecast information and long-term historical canal scheduling data of irrigation districts.
• To analyze the model performance and the mechanism of canal scheduling deviations caused by forecast errors.

Study Configuration

• Spatial Scale: Yongji Irrigation District, part of the Hetao Irrigation Area, Inner Mongolia, China. The study focuses on the Yongji main canal and its six associated sub-canals (Yonglan, Yonggang, Xile, Xinhua, Zhengshao, and Datuishui). The irrigated area is over 7600 square kilometers.
• Temporal Scale:
  • Computational Period: Crop growing season, from 6 April to 12 September annually.
  • Time Step: Daily (160 computation intervals per year).
  • Dataset Period: 2001 to 2020 (20 consecutive years).
  • Calibration Period: 2001 to 2010.
  • Validation Period: 2011 to 2020.

Methodology and Data

• Models used:
  • "Bi-level, Two-stage" (BT) model: A novel dynamically updated canal scheduling model developed in this study, comprising:
    • Canal Scheduling Rule Optimization Model: A bi-level optimization structure where the upper-level minimizes the multi-year sum of the water shortage index by optimizing the Target Residual Lump-Sum Water Quota (TRLSWQ), and the lower-level minimizes canal schedule efficiency loss by optimizing the Threshold Water Volume (TWV) for each sub-canal.
    • Water Demand Calculation Model: Based on field water balance, using the Penman-Monteith method for reference crop evapotranspiration.
  • Conventional Dynamically Updated Canal Scheduling (CC) model: Used for comparison, relying solely on forecast information.
  • Optimal Static Canal Scheduling (OC) model: Used as a reference, assuming perfectly accurate hydrometeorological forecasts for the entire period.
  • All models were written and executed in Fortran using the Intel Fortran Compiler (version 2021.3.0) within Microsoft Visual Studio Community 2019.
• Data sources:
  • Hetao Irrigation Area Administration Bureau: Provided data on canal engineering parameters, irrigation area, maximum water supply, and residual lump-sum water quota.
  • Meteorological data: Used for water demand calculation and forecasts.
  • Historical canal scheduling data: Used for calibrating the BT model's TRLSWQ and TWV parameters (2001–2010).
  • Crop planting areas: Average planting areas of wheat, corn, and sunflower in the Yongji Irrigation District from 2016 to 2020.
  • Crop coefficients and suitable soil moisture content limits: Provided in tables within the study.

Main Results

• The "Bi-level, Two-stage" (BT) model significantly improved canal scheduling performance compared to the conventional dynamically updated canal scheduling (CC) model.
• The BT model reduced the multi-year average total water shortage index (SWSI) of sub-canals from 40.81 to 31.44 (a 22.9% decrease) during the validation period (2011–2020).
• The BT model increased the multi-year average utilization rate of the initial lump-sum water quota (URILSWQ) from 89.32% to 92.82% (a 3.9% increase) compared to the CC model.
• The BT model's multi-year average canal system water loss rate (TWLR) was 2.99%, which was lower than the CC model's 3.10% and closer to the optimal static model's 2.80%.
• The calibrated Target Residual Lump-Sum Water Quota (TRLSWQ) and Threshold Water Volume (TWV) for each sub-canal exhibited a clear decreasing trend from April to September, aligning with the seasonal reduction in irrigation water demand.
• The study clarified that early-stage rainfall under-forecasting leads to excessive early water allocation and subsequent severe late-season shortages, while early-stage over-forecasting results in withheld early allocation and unused water. The BT model effectively balances these risks by activating a hedging mechanism when the residual lump-sum water quota falls below the threshold water volume.

Contributions

• Proposed a novel dynamically updated canal scheduling framework that integrates both long-term historical canal scheduling data (via Target Residual Lump-Sum Water Quota, TRLSWQ) and short-term hydrometeorological forecast information, addressing limitations of models relying solely on forecasts.
• Developed a "Bi-level, Two-stage" (BT) optimization model that defines clear, transparent, and operationally interpretable canal scheduling rules (TRLSWQ and Threshold Water Volume, TWV), enhancing practicality and traceability of decisions.
• Demonstrated superior performance of the BT model over conventional dynamically updated scheduling, significantly reducing water shortages and improving water utilization efficiency in a real-world irrigation district.
• Provided insights into the mechanism of canal scheduling deviations caused by forecast errors and showed how the BT model's hedging strategy mitigates these risks, preventing both premature water depletion and unnecessary restrictions.
• Offers a robust and practical strategy for irrigation district managers, requiring minimal forecast lead time and leveraging diverse information sources for efficient canal scheduling.

Funding

• Key Water Conservancy Research Project of Hubei Province (HBSLKY202409)
• National Natural Science Foundation of China (52179022)

Citation

@article{Yan2025Dynamically,
  author = {Yan, Ming and Wu, Fengyan and Chen, L. H. and Liu, Yong and Zeng, Xiang and Hu, Tiesong},
  title = {Dynamically Updated Irrigation Canal Scheduling Rules Based on Risk Hedging},
  journal = {Agriculture},
  year = {2025},
  doi = {10.3390/agriculture15242527},
  url = {https://doi.org/10.3390/agriculture15242527}
}