An advanced remote sensing retrieval method for urban non-optically active water quality parameters: An example from...

Area of Science:

• Environmental Science
• Remote Sensing Technology
• Water Quality Monitoring

Background:

• Optical insensitivity of non-optically active water quality parameters (NAWQPs) challenges remote sensing monitoring.
• Urban water quality assessment and management rely on accurate remote sensing tools.
• Spectral characteristics of water bodies are influenced by multiple NAWQPs.

Purpose of the Study:

• To propose a novel machine learning method for retrieving urban NAWQPs.
• To develop a multi-spectral scale morphological combined feature (MSMCF) for enhanced water quality monitoring.
• To evaluate the accuracy and stability of the proposed method across different datasets.

Main Methods:

• Development of a machine learning model utilizing MSMCF.
• Integration of local and global spectral morphological features.
• Application of a multi-scale approach for improved robustness and applicability.
• Testing retrieval accuracy and stability on measured and hyperspectral data.

Main Results:

• The MSMCF method demonstrates good retrieval performance for urban NAWQPs.
• The method shows applicability to hyperspectral data with varying spectral resolutions.
• The approach exhibits noise suppression capabilities.
• Sensitivity analysis reveals varying sensitivity of NAWQPs to spectral morphological features.

Conclusions:

• The proposed MSMCF machine learning method offers a more accurate and robust solution for remote sensing-based NAWQP retrieval.
• This research advances hyperspectral and remote sensing technologies for urban water quality management.
• Findings provide valuable references for addressing urban water quality deterioration.