Jin-xi et al. (2025) Influence of Soil Background Noise on Accuracy of Soil Moisture Content Inversion in Alfalfa Fields Based on UAV Multispectral Data

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

Identification

• Journal: Soil Systems
• Year: 2025
• Date: 2025-09-12
• Authors: Chen Jin-xi, Yuanbo Jiang, Wenjing Yu, Guangping Qi, Yanxia Kang, Minhua Yin, Yanlin Ma, Yayu Wang, Jiajun Zhu, Yanbiao Wang, Boda Li
• DOI: 10.3390/soilsystems9030098

Research Groups

Not specified in the provided text.

Short Summary

This study develops and evaluates drone-based multispectral remote sensing models for estimating topsoil moisture (0–10 cm) in alfalfa, finding that the XG-Boost model using spectral reflectance is most effective and that removing soil background noise does not significantly improve estimation accuracy in this specific environment.

Objective

• To systematically analyze the relationship between spectral reflectance, spectral indices, and soil moisture content in the top 0–10 cm layer of alfalfa using UAV multispectral imagery.
• To construct and comprehensively evaluate K-Nearest Neighbors (KNN), Random Forest Regression (RFR), ridge regression (RR), and XG-Boost inversion models for soil moisture.
• To investigate the impact of removing soil background noise on the accuracy of soil moisture estimations.

Study Configuration

• Spatial Scale: Field-scale (alfalfa crops).
• Temporal Scale: Not specified in the provided text.

Methodology and Data

• Models used: K-Nearest Neighbors (KNN), Random Forest Regression (RFR), Ridge Regression (RR), XG-Boost.
• Data sources: Unmanned aerial vehicle (UAV) multispectral imagery.

Main Results

• The XG-Boost model, using spectral reflectance as input, achieved the highest R² value (0.812) on the validation set, outperforming other models (R² from 0.465 to 0.770).
• When spectral indices were used as input, the RFR model achieved the highest R² value (0.632), significantly better than other models (R² from 0.366 to 0.535).
• Removing soil background noise did not significantly improve the accuracy of soil moisture estimates for any model; specifically, the XG-Boost R² decreased to 0.803 (from 0.812) and the RFR R² dropped to 0.628 (from 0.632).
• Spectral reflectance consistently provided more accurate data support for soil moisture inversion than spectral indices, both before and after soil background noise removal.
• The XG-Boost model, with spectral reflectance as the input variable, was identified as the most effective model for soil moisture content inversion.

Contributions

• Provides theoretical and technical support for the retrieval of surface soil moisture content in alfalfa using drone-based multispectral remote sensing.
• Offers evidence validating large-scale soil moisture remote sensing monitoring.
• Demonstrates that, contrary to common assumptions, removing soil background noise may not necessarily enhance the precision of soil moisture estimates in specific environments (e.g., areas with strong winds).

Funding

Not specified in the provided text.

Citation

@article{Jinxi2025Influence,
  author = {Jin-xi, Chen and Jiang, Yuanbo and Yu, Wenjing and Qi, Guangping and Kang, Yanxia and Yin, Minhua and Ma, Yanlin and Wang, Yayu and Zhu, Jiajun and Wang, Yanbiao and Li, Boda},
  title = {Influence of Soil Background Noise on Accuracy of Soil Moisture Content Inversion in Alfalfa Fields Based on UAV Multispectral Data},
  journal = {Soil Systems},
  year = {2025},
  doi = {10.3390/soilsystems9030098},
  url = {https://doi.org/10.3390/soilsystems9030098}
}