How many independent quantities can be extracted from ocean color?

Abstract

Products derived from remote sensing reflectances ((Figure presented.)), for example, chlorophyll, phytoplankton carbon, euphotic depth, or particle size, are widely used in oceanography. Problematically, (Figure presented.) may have fewer degrees of freedom (DoF) than measured wavebands or derived products. Here, we show that a global sea surface hyperspectral (Figure presented.) dataset has DoF = 4. MODIS-like multispectral equivalent in situ data also have DoF = 4, while their SeaWiFS equivalent has DoF = 3. Both multispectral-equivalent datasets predict individual hyperspectral wavelengths' (Figure presented.) within nominal uncertainties. Remotely sensed climatological multispectral (Figure presented.) have DoF = 2, as information is lost by atmospheric correction, shifting to larger spatiotemporal scales, and/or more open-ocean measurements, but suites of (Figure presented.) -derived products have DoF = 1. These results suggest that remote sensing products based on existing satellites' (Figure presented.) are not independent and should not be treated as such, that existing (Figure presented.) measurements hold unutilized information, and that future multi- or especially hyper-spectral algorithms must rigorously consider correlations between (Figure presented.) wavebands.