Que et al. (2026) Retrieving Soil Water Content in Winter Wheat Fields Using UAV-Based Multi-Source Remote Sensing and Machine Learning

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Agronomy
• Year: 2026
• Date: 2026-03-30
• Authors: Yanhong Que, Dan Wu, Mingliang Jiang, Jianyi Deng, Cong Liu, Su Wu, Fengbo Li, Yanpeng Li
• DOI: 10.3390/agronomy16070717

Research Groups

The provided text does not explicitly list the research groups, labs, or departments involved in the study.

Short Summary

This study proposes a novel hybrid framework integrating an improved water cloud model (IWCM) with machine learning to retrieve farmland soil water content (SWC) in winter wheat with high accuracy and physical interpretability. The framework, using multi-modal UAV data, significantly enhances SWC retrieval performance, achieving an R² of 0.865, MAE of 0.0152, and RMSE of 0.0197 with a Random Forest model driven by spectral reflectance.

Objective

• To develop a novel hybrid framework that integrates an improved water cloud model (IWCM) with machine learning algorithms to retrieve farmland soil water content (SWC) in winter wheat, aiming for both high accuracy and physical interpretability.

Study Configuration

• Spatial Scale: Farmland (winter wheat fields), likely at a field or plot scale given the use of Unmanned Aerial Vehicles (UAVs).
• Temporal Scale: Two consecutive years (2024–2025) during the heading stage of winter wheat.

Methodology and Data

• Models used: Improved Water Cloud Model (IWCM), Back Propagation Neural Network (BPNN), Random Forest (RF).
• Data sources: Multi-modal Unmanned Aerial Vehicle (UAV) experiments using a synchronized system equipped with:
  • Miniature Synthetic Aperture Radar (MiniSAR)
  • Multi-spectral sensor
  • Derived features: Spectral reflectance, vegetation indices, MiniSAR polarimetric parameters, and fractional vegetation cover (used in IWCM).

Main Results

• The proposed hybrid framework significantly enhances soil water content retrieval performance in winter wheat fields.
• The Random Forest (RF) model, when driven by spectral band reflectance within the physically constrained IWCM architecture, achieved optimal accuracy.
• Optimal retrieval accuracy metrics: Coefficient of determination (R²) = 0.865, Mean Absolute Error (MAE) = 0.0152, and Root Mean Square Error (RMSE) = 0.0197.
• The IWCM explicitly decouples vegetation and soil scattering contributions by incorporating fractional vegetation cover, leading to the derivation of physically meaningful soil backscatter coefficients.
• Compared to purely empirical approaches, the IWCM significantly improved the physical interpretability of microwave polarimetric characteristics.
• The integration of mechanistic models with multi-source UAV remote sensing data, guided by physics-derived features, better represents the complex interactions among vegetation, soil, and microwave scattering.

Contributions

• Proposes a novel hybrid framework that effectively integrates a mechanistic model (IWCM) with machine learning for soil water content retrieval, bridging the gap between purely empirical and complex physical models.
• Introduces an improved Water Cloud Model (IWCM) that enhances physical interpretability by explicitly decoupling vegetation and soil scattering contributions through the incorporation of fractional vegetation cover.
• Demonstrates the superior performance and physical interpretability achieved by using physics-derived features from multi-modal UAV remote sensing data (MiniSAR and multi-spectral) to guide data-driven modeling.
• Achieves high accuracy in retrieving soil water content for winter wheat, providing a robust solution for a challenging agricultural monitoring task.
• Offers a valuable reference for developing operationally applicable and physically interpretable farmland soil water content monitoring systems.

Funding

The provided text does not contain information regarding the funding sources for this research.

Citation

@article{Que2026Retrieving,
  author = {Que, Yanhong and Wu, Dan and Jiang, Mingliang and Deng, Jianyi and Liu, Cong and Wu, Su and Li, Fengbo and Li, Yanpeng},
  title = {Retrieving Soil Water Content in Winter Wheat Fields Using UAV-Based Multi-Source Remote Sensing and Machine Learning},
  journal = {Agronomy},
  year = {2026},
  doi = {10.3390/agronomy16070717},
  url = {https://doi.org/10.3390/agronomy16070717}
}