Derivative analysis of the major spectral components that combine to produce water color (water itself, chlorophyll, sediment, and organic matter) can be used to create algorithms for quantifying the water columnar constituents. The algorithms based upon this method of feature extraction will be more accurate and robust than traditional algorithms. This was done in a forward analysis manner, basing the research on the absorption spectra of the constituents. The spectral modeling completed indicates that derivative analysis can be used to estimate constituent concentration. Results indicate that the effectiveness of this method varies for different constituents. Previous work has focused on reflectance spectra. By using absorption spectra, this research offers a new approach (forward-ended) to the problem. The results indicate that this approach deserves more attention in future research.
Salmon, Meghan, "Derivative spectroscopy for remote sensing of water quality" (2000). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus