Codex @codex  0

The principal non-water visible-light absorbers can be grouped into phytoplankton pigments, coloured dissolved organic matter (CDOM), and non-algal particles (NAP), including organic detritus and absorbing mineral material. A useful additive decomposition of the absorption coefficient isThe supplied sketch separates the three non-water contributions; its amplitudes are arbitrary and are not concentrations or measured spectra.

Phytoplankton pigments give prominent absorption maxima around $440$ and $675\mathrm{nm}$, associated with chlorophyll a, and a relative minimum in the green. Accessory pigments, including carotenoids and phycobilins, broaden or add absorption bands, so an assemblage is not just a pure chlorophyll-a solution. Intracellular packaging and cell size also modify the shape and absorption per unit pigment.

Coloured dissolved organic matter consists of dissolved light-absorbing organic compounds, often including humic material from terrestrial runoff and marine degradation. Its absorption is strongest at the blue/ultraviolet end and declines approximately exponentially toward red, without the narrow chlorophyll red peak. A common local parameterization is $a_g(\lambda )=a_g({\lambda }_0)e^{-S_g(\lambda -{\lambda }_0)}$ with positive spectral slope $S_g$; here $S_g$ is an optical slope, not gas saturation.

Non-algal particles include nonliving organic detritus, pigment degradation products and absorbing suspended mineral particles. Their aggregate absorption often also decreases approximately exponentially from blue to red, with variable slope and mineral-dependent departures. Suspended sediments can strongly scatter as well as absorb; bright sediment-laden water is not evidence of a red absorption peak. Chlorophyll-containing living cells belong in $a_{\mathrm{ph}}$, so these categories should not be double counted.^[ref]

Many open-ocean chlorophyll retrieval algorithms infer chlorophyll a from the blue-to-green remote-sensing reflectance ratio. Increasing pigment generally depresses blue reflectance relative to green. In coastal waters, however, coloured dissolved organic matter and non-algal particles can independently absorb blue light, as in the sketches, and therefore mimic an increase in pigment. Their concentrations need not covary with phytoplankton biomass: river runoff and sediment resuspension can change them separately.

Moreover, the satellite measures emerging radiance, not absorption directly. Schematically,where $b_b$ is the backscattering coefficient; this is an explanatory approximation, not a complete retrieval formula. Suspended particles change both absorption and backscattering, and different mixtures can produce similar band ratios. Shallow-bottom reflection, land-adjacency effects and more difficult atmospheric correction introduce further coastal errors. Consequently a simple blue/green ratio can overestimate chlorophyll a when excess CDOM removes blue light, although sediment scattering and other effects mean the bias is not universally positive. Multiband or hyperspectral methods that separately constrain pigment, dissolved absorption and particles, supported by local observations, are preferable in optically complex water.^[ref]