Hou et al. (2025) Characterization of single-cropping rice net irrigation water requirements in China's major rice-producing regions using time-frequency domain methods

Identification

• Journal: Agricultural Water Management
• Year: 2025
• Date: 2025-12-10
• Authors: Miao Hou, Xing Yang, Wenye Zhang, Yugeng Guo, Fei Qi, Linpeng Zhai, Chaozhen He
• DOI: 10.1016/j.agwat.2025.110071

Research Groups

• Jiangsu Hydraulic Research Institute, Nanjing, China
• College of Agricultural Science and Engineering, Hohai University, Nanjing, China
• Jiangsu Rural Water Conservancy and Science Technology Development Center, Nanjing, China
• Jiangsu Surveying and Design Institute of Water Resources Co., Ltd., Yangzhou, China

Short Summary

This study characterizes single-cropping rice net irrigation water requirements (RIWR) in China's major rice-producing regions from 1951 to 2023 using time-frequency domain methods and projects future changes. It reveals substantial spatiotemporal variability in RIWR, with national annual averages ranging from 524.6 to 791.5 mm and dominant 15–25 year cycles, primarily driven by effective precipitation, sunshine duration, maximum air temperature, and relative humidity.

Objective

• To calculate rice net irrigation water requirements (RIWR) using a water balance approach that incorporates the FAO Penman-Monteith formula.
• To identify the main climate factors affecting these requirements and investigate their time-frequency characteristics at multiple scales using comprehensive time-domain methods.
• To analyze trends in RIWR during both the historical period (1951–2023) and the projected period (2015–2100).

Study Configuration

• Spatial Scale: Five major rice-producing regions in China (Northeast, Southwest, South, Central, and East China), covering 18 provinces. Analyses were conducted at daily, provincial, regional, and national scales.
• Temporal Scale: Historical period from 1951 to 2023 (73 years) and future projections from 2015 to 2100 under SSP2–4.5 and SSP5–8.5 climate scenarios.

Methodology and Data

• Models used:
  • Water balance method (for RIWR calculation)
  • FAO Penman-Monteith formula (for reference crop evapotranspiration, ETo)
  • Crop coefficient method (for actual crop evapotranspiration, ETc)
  • Wavelet analysis (Complex Morlet Wavelet for time-frequency characteristics, wavelet variance, cross-spectrum, coherence spectrum)
  • Mann-Kendall (MK) test (for trend analysis)
  • Pearson correlation analysis
  • Association-degree weight analysis
  • Sensitivity analysis (scenario simulations)
• Data sources:
  • Meteorological data: Daily maximum, minimum, and mean air temperatures, effective precipitation, wind speed, sunshine duration, and relative humidity from 54 meteorological stations (3 per province) across China (1951–2023), sourced from China’s Meteorological Administration.
  • Circulation indices: Multivariate ENSO Index (MEI), North Atlantic Oscillation Index (NAOI), Arctic Oscillation Index (AOI) from National Oceanic and Atmospheric Administration (NOAA); East Asian Summer Monsoon Index (EASMI) from Huang and Zhao (2022).
  • Future climate data: Downscaled CMIP6 models (SRCPCN10 dataset) based on Shared Socioeconomic Pathway (SSP) framework (SSP2–4.5 and SSP5–8.5 scenarios) from the National Tibetan Plateau Data Center.
  • Soil texture spatial distribution data (www.gis5g.com) for percolation estimation.
  • Published literature, field experience, provincial rice irrigation quotas, and agricultural water use bulletins for crop phenological periods and validation of percolation values.

Main Results

• Temporal Variability (1951–2023): National annual average RIWR fluctuated between 524.6 mm and 791.5 mm, showing an overall decreasing trend of 0.93 mm/yr. RIWR was above the long-term mean from 1951–1980, below from 1981–2000, and alternating from 2001–2023.
• Spatial Variability: Province-level RIWR ranged from 315.0 mm/yr (Anhui) to 1250.1 mm/yr (Jilin). The regional variation range (largest to smallest) was Northeast > Central China > East China > South China > Southwest. Northeast China's average annual RIWR was approximately 32% higher than the national average.
• Periodic Variability: RIWR variability is dominated by cycles with periods of 15–25 years (e.g., 23-year pseudo-period). Interregional phase differences of up to 180° contribute to spatial heterogeneity, with high RIWR periods in Northeast China often corresponding to low periods in South China, and vice versa.
• Main Drivers and Sensitivity: Effective precipitation (Pre), sunshine duration (SD), maximum air temperature (Tmax), and relative humidity (RH) were identified as primary meteorological drivers. The sensitivity of RIWR to these factors varies regionally: Tmax dominates in Northeast China, SD in South China, Pre in Southwest and East China, and RH in Central China.
• Future Projections (2015–2100): Under SSP2–4.5 and SSP5–8.5 scenarios, the historical decline in RIWR may slow or reverse in some regions (Southwest, Central, and East China), while Northeast and South China are expected to maintain decreasing trends. Projected annual RIWR by 2100 generally remains below historical maximum values across all regions, with less than 1% of years slightly exceeding historical peaks.
• Percolation Losses: Ranged from 215.6 mm to 577.2 mm per crop season, with a national average of 372.2 mm. Southwest China exhibited the largest losses (averaging 56.19% of RIWR), followed by East (48.26%) and Central China (40.68%).

Contributions

• Provides the first long-term (1951–2023) and national-scale characterization of single-cropping rice net irrigation water requirements (RIWR) in China's major rice-producing regions.
• Introduces and applies a novel time-frequency domain method, integrating wavelet analysis, Pearson correlation, association-degree weight analysis, and variance shape comparison, to comprehensively identify key meteorological drivers and their multi-scale periodic variability, addressing a research gap.
• Quantifies future changes in irrigation demand under SSP2–4.5 and SSP5–8.5 scenarios, offering insights into potential shifts in water stress.
• Highlights the critical importance of accounting for spatial, temporal, and frequency-domain variations for developing adaptive and region-specific irrigation strategies.
• Validates the robustness of the identified associated indices by demonstrating their exact correspondence with the most sensitive factors influencing RIWR in each region.

Funding

• Water Resources Science and Technology Project of Jiangsu Province (Grant No. 2022006).

Citation

@article{Hou2025Characterization,
  author = {Hou, Miao and Yang, Xing and Zhang, Wenye and Guo, Yugeng and Qi, Fei and Zhai, Linpeng and He, Chaozhen},
  title = {Characterization of single-cropping rice net irrigation water requirements in China's major rice-producing regions using time-frequency domain methods},
  journal = {Agricultural Water Management},
  year = {2025},
  doi = {10.1016/j.agwat.2025.110071},
  url = {https://doi.org/10.1016/j.agwat.2025.110071}
}