Simulating the effects of inelastic scattering on upwelling radiance in coastal and inland waters: implications for hyperspectral remote sensing

• Published in: Current science (Bangalore) 2015-03, Vol.108 (5), p.903-914
• Main Authors: Vigneshwaran, J., Shanmugam, P., Gokul, E. A.
• Format: Article
• Language: English
• Subjects: Chlorophylls; Coastal water; Fluorescence; Marine; Modeling; Radiance; Raman scattering; Remote sensing; SPECIAL SECTION: HYPERSPECTRAL REMOTE SENSING; Upwelling water; Water depth; Wavelengths
• ISSN: 0011-3891

Successful interpretation of spectral images from hyperspectral sensors is particularly important to achieve some far-reaching goals, including the detection and assessment of harmful algal blooms, water quality characteristics, trophic status of water bodies and marine/freshwater primary production. The present work is motivated by the desire to study the effect of inelastic scattering (chlorophyll and gelbstoff fluorescence and water Raman scattering) on upwelling radiance through a direct method. The contribution of inelastic scattering to the emergent (upwelling) radiance from the water column is of significant importance and has relevance for the hyperspectral remote sensing of water colour. The method introduced to simulate inelastic scattering is a straightforward approach and differs from previous work in the way of derivation. The model allows for simulation of upwelling radiance for clear, turbid and productive (phytoplankton-dominated) waters. The results of this model are assessed by comparison with depth-dependant upwelling radiance data and results from the Hydrolight numerical model for these three water types. The comparisons show good agreement between measured and simulated values from the present model. The errors of this model are significantly small (when compared with those of the Hydrolight model) in the red wavelength region, where the chlorophyll fluorescence emission tends to peak around 685-720 nm when the concentration of chlorophyll is high in productive eutrophic waters. The magnitude and peak position at these wavelengths as produced by the present model could be used as precise indicators and predictors for the phytoplankton concentration in marine and inland water bodies. Thus, this work will have important implications in refining hyperspectral bio-optical algorithms for fluorescence and water Raman scattering as well as for the high concentrations and variabilities of water constituents in coastal and inland regions, and in evaluating remote sensing reflectance measurements of these water bodies.