Quantifying contaminant concentration in mixed water systems using spectro-polarimetric measurement

Area of Science:

• Environmental Science
• Optical Physics
• Remote Sensing

Background:

• Water pollution poses significant threats to water quality and availability.
• Traditional in-situ water quality measurements are limited in capturing ecosystem variability.
• Remote sensing offers a promising alternative but requires accurate spectral decomposition due to optical complexities.

Purpose of the Study:

• To develop an inverse model integrating spectral and polarization measurements for improved water quality assessment.
• To enhance the estimation of contaminant concentrations under dynamic water surface conditions.
• To evaluate the performance of the new model against existing indices and identify limitations.

Main Methods:

• Integration of spectral measurements with polarization data.
• Development of an inverse model for contaminant concentration estimation.
• Validation using chlorophyll (Chl) and suspended sediment (SS) as target contaminants.

Main Results:

• The proposed model achieved R-squared values of 0.86 for Chl and SS estimation with RMSE <10%.
• Two-band indices showed superior performance compared to three-band indices.
• Limitations were observed in mixed waters and under varying viewing angles and wave conditions.

Conclusions:

• Integrating polarization with spectral measurements significantly improves water quality prediction accuracy, especially in dynamic environments.
• Existing indices are insufficient for optimal contaminant estimation.
• Further research is needed to address challenges in mixed contaminant scenarios and optical complexities.