A new approach to quantify chlorophyll-a over inland water targets based on multi-source remote sensing data

Area of Science:

• Environmental remote sensing
• Water quality monitoring
• Phytoplankton dynamics

Background:

• Chlorophyll-a (Chl-a) concentration is a key indicator of phytoplankton biomass and eutrophication in inland waters.
• Long-term Chl-a monitoring is crucial for effective water resource management.
• Existing remote sensing methods struggle with multi-source data integration and accuracy for inland water Chl-a estimation.

Purpose of the Study:

• To address the challenges of long-term inland water Chl-a estimation using multi-source remote sensing data.
• To evaluate the consistency and fidelity of reflectance data from multiple satellite sensors (OLCI, MSI, OLI).
• To develop and validate a robust model for accurate, high-precision Chl-a retrieval from diverse remote sensing inputs.

Main Methods:

• Evaluation of satellite-derived reflectance products (R_rhos) against in situ measurements from Erhai Lake.
• Assessment of sensor data consistency, particularly in the green-red spectral bands.
• Development and application of an extreme gradient boosting (XGB) model for multi-source remote sensing data fusion.

Main Results:

• R_rhos products demonstrated high correlation with measured reflectance, indicating reliable atmospheric correction.
• Satellite reflectance showed higher fidelity in the green-red bands, crucial for Chl-a estimation.
• The proposed XGB model with R_rhos products significantly outperformed other methods in estimating Chl-a concentrations.

Conclusions:

• The XGB model effectively integrates multi-source remote sensing data for accurate long-term Chl-a monitoring in inland waters.
• This approach provides a valuable tool for supporting inland lake management and water quality assessments.
• The study establishes a robust framework for utilizing diverse satellite data for environmental monitoring over extended periods.