Band-ratio or spectral-curvature algorithms for satellite remote sensing?

For the retrieval of chlorophyll concentrations or the total absorption coefficients of oceanic waters based on water color, there are algorithms that use either band ratios or spectral curvatures of remote-sensing reflectance or water leaving radiance. We show that band-ratio algorithms have the potential to be applied to a wider dynamic range of oceanic waters, whereas spectral-curvature algorithms show stable performance as long as the data set falls within the appropriate range.