Wang et al. (2026) Soil spectral simulation and soil parameter retrieval with an optimized four-flux MLG radiative transfer model

Identification

• Journal: Remote Sensing of Environment
• Year: 2026
• Date: 2026-01-07
• Authors: W. Wang, Xiaohe Zhang, Kun Shang, Yanli Sun, Youxin Sun, Wenliang Chen
• DOI: 10.1016/j.rse.2025.115210

Research Groups

• Aerospace Information Research Institute, Chinese Academy of Sciences
• University of Chinese Academy of Sciences
• Land Satellite Remote Sensing Applications Center, MNR
• Aerospace Information Technology University

Short Summary

This study proposes a novel method based on an optimized four-flux Maheu-Letoulouzan-Gouesbet (MLG) radiative transfer model and a genetic algorithm to achieve high-precision soil spectral simulation and efficiently remove soil moisture effects, significantly enhancing the prediction accuracy of other soil parameters.

Objective

• To develop a novel method based on an optimized four-flux Maheu-Letoulouzan-Gouesbet (MLG) radiative transfer model to achieve high-precision soil spectral simulation and efficiently remove soil moisture effects.

Study Configuration

• Spatial Scale: In-situ field spectra measurements, with potential for large-scale soil monitoring via satellite remote sensing.
• Temporal Scale: Not explicitly defined for the study's data collection, but the application is for monitoring.

Methodology and Data

• Models used: Optimized four-flux Maheu-Letoulouzan-Gouesbet (MLG) radiative transfer model, genetic algorithm (with spectral continuity constraint).
• Data sources: ICRAF dry soil dataset, Nong'an wet soil dataset, in-situ field spectra.

Main Results

• The optimized MLG model achieved the highest accuracy in soil spectral simulation under both dry and wet conditions.
• The coefficient of determination (R²) for spectral simulation improved from 0.87 to 0.94 in the ICRAF dry soil dataset and from 0.87 to 0.94 in the Nong'an wet soil dataset, compared to the Kubelka-Munk (KM) model.
• The model maintained high accuracy (R² = 0.81) even when soil moisture content exceeded 30%, particularly in water absorption bands around 1400 nm and 1900 nm.
• After removing soil moisture effects from in-situ field spectra, corrected spectra exhibited stronger correlations with organic matter, iron oxides, and texture fractions.
• Prediction accuracy for soil parameters significantly enhanced (R² > 0.90), with the relative percent difference (RPD) increasing by 0.88–3.00.

Contributions

• Introduction of a novel method based on an optimized four-flux MLG radiative transfer model for high-precision soil spectral simulation and efficient soil moisture effects removal.
• Incorporation of key soil properties (soil moisture, organic matter, iron oxides, texture fractions) into the MLG model for better characterization of spectral formation mechanisms.
• Development of a genetic algorithm with a spectral continuity constraint for global parameter optimization, simplifying calculations and improving simulation reliability.
• Demonstrated superior performance over the Kubelka-Munk (KM) model in spectral simulation and significantly enhanced the prediction accuracy of other soil parameters after soil moisture removal.

Funding

• Not specified in the provided text.

Citation

@article{Wang2026Soil,
  author = {Wang, W. and Zhang, Xiaohe and Shang, Kun and Sun, Yanli and Sun, Youxin and Chen, Wenliang},
  title = {Soil spectral simulation and soil parameter retrieval with an optimized four-flux MLG radiative transfer model},
  journal = {Remote Sensing of Environment},
  year = {2026},
  doi = {10.1016/j.rse.2025.115210},
  url = {https://doi.org/10.1016/j.rse.2025.115210}
}