[Algorithm of inland water retrieval based on spectral matching]

Area of Science:

• Remote Sensing
• Ocean Optics
• Water Quality Monitoring

Context:

• Inland water bodies exhibit complex optical properties due to interacting components like chlorophyll, suspended matter, and yellow substances.
• Traditional statistical and bio-optical models face challenges in accurately retrieving water quality parameters due to spectral interference and seasonal variations.

Purpose:

• To develop and validate a new inversion method for inland water quality retrieval using spectral matching.
• To create a comprehensive look-up table of spectral reflectance (Rrs) for various combinations of water constituents.
• To adapt hyperspectral data to MODIS spectral bands for broader applicability.

Summary:

• The method utilizes hyperspectral data and bio-optical models to derive the backscattering coefficient of suspended matter.
• A look-up table correlating inherent optical properties (chlorophyll, suspended matter, yellow substance) with Rrs is generated.
• This table is converted to MODIS spectral data and applied using a minimum distance principle to determine constituent concentrations.
• Average relative errors for chlorophyll and suspended matter were 38.6% and 28%, respectively.

Impact:

• This approach enhances the accuracy of inland water quality retrieval by mitigating the instability issues associated with inherent optical properties.
• It offers a robust alternative to existing methods, improving the reliability of remote sensing for inland water monitoring.
• The method provides a valuable tool for understanding and managing inland water resources.