JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Rapid characterization of heavy metals in soil using a novel integrated strategy for near-infrared spectroscopy models

  • 0School of Food and Liquor Engineering, Sichuan University of Science and Engineering, Yibin 644000, PR China; Biogas Institute of Ministry of Agriculture and Rural Affairs, Chengdu, Sichuan Province 610041, PR China.
Spectrochimica Acta. Part a, Molecular and Biomolecular Spectroscopy +

|

January 1, 2026

|

January 1, 2026

View abstract on PubMed

Summary

This summary is machine-generated.

This study developed a weighted fusion strategy to improve the accuracy of predicting soil heavy metals (Pb, Cd, Cr) using near-infrared spectroscopy. This approach overcomes limitations of single models, enhancing soil quality assessment for sustainable agriculture.

Area Of Science

  • Environmental Science
  • Soil Science
  • Analytical Chemistry

Background

  • Soil quality is crucial for ecological security and agricultural sustainability.
  • Heavy metal (HM) contamination poses a significant threat to soil health.
  • Existing methods for soil HM assessment face challenges in standardization, efficiency, and precision.

Purpose Of The Study

  • To develop a standardized, efficient, and high-precision technology for acquiring soil HM information.
  • To address limitations of single predictive models in universality and reproducibility.
  • To improve the inversion accuracy of soil HM contents using near-infrared spectral data.

Main Methods

  • Employed linear and nonlinear modeling approaches with multi-step spectral data processing.
  • Constructed near-infrared spectral prediction models for lead (Pb), cadmium (Cd), and chromium (Cr).
  • Integrated model averaging and weighted fusion strategies, utilizing six sub-models to overcome single-model limitations.

Main Results

  • Nonlinear algorithms showed potential for inverting soil biochemical properties.
  • Weighted fusion strategy significantly reduced average prediction errors for Pb (2.12%), Cd (5.35%), and Cr (3.51%).
  • Residual prediction deviations on the test set improved by 0.43 for Pb, 0.15 for Cd, and 0.20 for Cr compared to single models.

Conclusions

  • The weighted fusion strategy effectively addresses limitations and instability of single-model estimations for soil HM content.
  • This integrated approach enhances the practical applicability of soil regression models for HM assessment.
  • Provides new insights for optimizing soil regression models in environmental monitoring and agricultural management.

Keywords: