A hierarchical residual correction-based hyperspectral inversion method for soil heavy metals considering spatial...

Area of Science:

• Environmental Science
• Geoscience
• Remote Sensing

Background:

• Soil heavy metal (SHM) contamination poses environmental risks.
• Hyperspectral remote sensing shows potential for SHM monitoring.
• Spatial heterogeneity and environmental factors complicate accurate SHM inversion.

Purpose of the Study:

• To develop a novel method for accurate SHM inversion considering spatial heterogeneity.
• To address the non-stationary relationship between spectral reflectance and SHM concentration.
• To enhance the interpretation of SHM spatial distribution patterns.

Main Methods:

• Proposed the Hierarchical Residual Correction-based Hyperspectral Inversion Method (HRCHIM).
• Constructed a global prediction model using ground hyperspectral data.
• Utilized Geodetector to identify environmental factors for hierarchical residual correction models.

Main Results:

• HRCHIM synergizes global trends with local stochasticity for improved accuracy.
• Six critical environmental factors influencing SHM were identified across hierarchical levels.
• Achieved superior inversion performance with Rv² = 0.94, RMSEv = 0.21, RPDv = 4.11 in a Cadmium-contaminated area.

Conclusions:

• HRCHIM significantly enhances the accuracy and interpretability of SHM spatial heterogeneity.
• The method offers a more precise approach for SHM contamination assessment.
• Facilitates targeted strategies for the prevention and control of soil heavy metal pollution.