Wu et al. (2026) Accurate diagnosis of soil moisture and canopy water content in citrus orchards based on synergistic multi-source remote sensing

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-08-31
• Authors: Zongjun Wu, Bin Zhu, Ningbo Cui, Wei Zhao, Yanjun Wang, Liwen Xing, Quanshan Liu, Shenglin Wen, Bihang Fan
• DOI: 10.1016/j.agwat.2026.110725

Research Groups

• State Key Laboratory of Hydraulics and Mountain River Engineering & College of Water Resource and Hydropower, Sichuan University, Chengdu, China
• Institute of Mountain Hazards and Environment, Chinese Academy of Sciences, Chengdu, China

Short Summary

This study developed a synergistic multi-source remote sensing framework, integrating Landsat-8/9, Sentinel-2, and Sentinel-1A data, to enhance the Water Cloud Model with phenology-driven dynamic canopy parameters for accurate soil moisture and canopy water content diagnosis in citrus orchards. The improved model successfully retrieved soil moisture at a regional scale, with Sentinel-1A VV polarization demonstrating superior accuracy (R²: 0.785–0.854, RMSE: 6.125–8.351%) compared to VH polarization.

Objective

• To establish an effective framework for regional-scale soil moisture retrieval in citrus orchards based on an optimized Water Cloud Model (WCM).
• To calibrate internal WCM parameters using a nonlinear least squares optimization algorithm to improve model adaptability for orchard environments.
• To retrieve vegetation water content by integrating multi-temporal optical remote sensing data and a modified soil-adjusted vegetation index to effectively characterize canopy water conditions.
• To achieve reliable regional 10 cm depth soil moisture inversion for large-scale citrus orchards using the improved Water Cloud Model.
• To generate spatially continuous soil moisture distribution maps under different polarization modes to reveal the spatial variability of soil water status in orchard systems.

Study Configuration

• Spatial Scale: A late-maturing citrus industrial park in Southwest China (104°51'6″ E, 29°55'7″ N, 519.0 m elevation), covering approximately 1.87 square kilometers (2800 mu). Soil moisture was measured at 102 ground sampling points at 0–10 cm depth. All remote sensing data were resampled to a unified 10 m grid resolution.
• Temporal Scale: The study period covered 2023 and 2024, encompassing the full citrus growth cycle, including shoot emergence and flowering (March 15 – April 30), young fruit (May 1 – June 30), fruit expansion (July 1 – October 31), and fruit maturation (November 1 – December 31). Multi-temporal satellite data were acquired throughout these periods.

Methodology and Data

• Models used:
  • Improved Water Cloud Model (WCM) with phenology-driven dynamic canopy parameters.
  • Modified Soil-Adjusted Vegetation Index (MSAVI) for vegetation water content estimation.
  • Normalized Difference Vegetation Index (NDVI) for vegetation fractional coverage.
  • Nonlinear least squares optimization algorithm and ant colony optimization algorithm for WCM parameter calibration.
• Data sources:
  • Satellite:
    • Landsat-8/9 (OLI sensor): Optical, multi-spectral imagery (11 bands, 30 m spatial resolution, 12-bit radiometric resolution).
    • Sentinel-2 (MSI sensor): Optical, multi-spectral imagery (12 bands, 10-60 m spatial resolution, 12-bit radiometric resolution).
    • Sentinel-1A (C-band SAR): Microwave, dual-polarization (VV and VH), Interferometric Wide (IW) imaging mode, Ground Range Detected (GRD) product type (5 m × 20 m resolution, resampled to 10 m).
  • Observation (Ground data):
    • Surface soil moisture (0–10 cm depth) measured by the oven-drying method.
    • Leaf Area Index (LAI) measured using a Yaxin-1201 plant canopy analyzer.
    • Canopy Water Content (CWC) calculated from leaf fresh/dry weight and leaf area.
    • Daily meteorological variables (air temperature, precipitation, solar global radiation, sunshine hours, reference crop evapotranspiration, relative humidity, wind speed) for 2023 and 2024.

Main Results

• Canopy water content (CWC) retrieved from optical remote sensing exhibited significant spatiotemporal heterogeneity across the orchard.
  • CWC ranged from 1.65–5.71 kg⋅m⁻² in 2023 and 1.55–5.78 kg⋅m⁻² in 2024.
  • High CWC values (e.g., 3.28–5.71 kg⋅m⁻² in 2023 spring/summer) were observed during shoot-flowering and fruit expansion stages, while lower values (e.g., 1.65–3.28 kg⋅m⁻² in 2023 autumn/winter) occurred during autumn and winter.
• Soil moisture inversion using multi-temporal Sentinel-1A data showed that VV polarization consistently outperformed VH polarization.
  • For VV polarization, the coefficient of determination (R²) ranged from 0.785–0.854, and the root mean square error (RMSE) ranged from 6.125–8.351%. The Nash-Sutcliffe efficiency (NSE) ranged from 0.782–0.871.
  • For VH polarization, R² ranged from 0.515–0.759, and RMSE ranged from 6.942–8.412%. The NSE ranged from 0.623–0.782.
• The highest soil moisture inversion accuracy for both polarizations was achieved during the fruit maturation stage (VV: R² 0.811–0.842, RMSE 6.125–6.132%; VH: R² 0.682–0.711, RMSE 6.942–7.025%).
• Backscattering coefficients from Sentinel-1A (ranging 0.01–0.84 in 2023 and 0.01–0.95 in 2024) showed seasonal fluctuations driven by canopy density and water content, with VV polarization consistently yielding higher coefficients than VH.

Contributions

• Developed a novel synergistic multi-source remote sensing framework that integrates optical (Landsat-8/9, Sentinel-2) and microwave (Sentinel-1A) data to overcome limitations of single-source data and fixed canopy parameters in traditional models.
• Enhanced the Water Cloud Model by incorporating phenology-driven dynamic canopy parameters (canopy coverage and canopy water content) retrieved from high-resolution optical imagery, improving its adaptability and accuracy for complex orchard environments.
• Effectively mitigated the challenge of persistent cloud cover in mountainous regions by leveraging all-weather C-band SAR data, enabling continuous monitoring of soil moisture.
• Provided a robust methodology for high-resolution and precision inversion of surface soil moisture at the regional farmland scale, offering critical decision support for precision irrigation and water resource management in citrus orchards.
• Quantitatively demonstrated the superior performance of Sentinel-1A VV polarization over VH polarization for soil moisture retrieval in vegetated orchard areas, providing insights for optimal data selection in future studies.

Funding

• National Natural Science Foundation of China (52609091, U25A20769, 2021YFD1600803, 52309057)
• Sichuan Science and Technology Program (2024ZHCG0101, 2026YFHZ0094)
• Research Funds for the Central Universities, China (20822041J4119)
• Renshou Huarui Agricultural Investment Limited Liability Company

Citation

@article{Wu2026Accurate,
  author = {Wu, Zongjun and Zhu, Bin and Cui, Ningbo and Zhao, Wei and Wang, Yanjun and Xing, Liwen and Liu, Quanshan and Wen, Shenglin and Fan, Bihang},
  title = {Accurate diagnosis of soil moisture and canopy water content in citrus orchards based on synergistic multi-source remote sensing},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110725},
  url = {https://doi.org/10.1016/j.agwat.2026.110725}
}