**3. On the road to predicting remote sensing reflectance: Fundamental basis between pigments and in vivo absorption**

Algorithm development for remote sensing is focused on estimating the quantities of various optically-active constituents in the ocean. Various pigments present in phytoplankton play a role in determining the total absorption coefficient of phytoplankton, which is a key determinant of the spectral variability in remote-sensing reflectance. We utilize a reconstruction model to demonstrate the spectral variability of photoprotective and photosynthetic pigments (Sathyendranath et al. 1987, Bidigare 1990). There are differences in the weight-specific absorption spectra of various *in vitro* (extracted) phytoplankton pigments (Figure 5). The comparison of maximal peaks of mass-specific absorption spectra

Ocean Color Remote Sensing of Phytoplankton Functional Types 109

The multi-spectral radiance measurements from the satellite sensor can be corrected for the influence of the atmosphere to yield remote-sensing reflectance. It has been demonstrated using various types of models (e.g. Morel & Prieur 1977; Sathyendranath & Platt 1988) that remote sensing reflectance, Rrs (λ), at a particular wavelength (λ) increases with the backscattering coefficient of light at a wavelength, and decreases with the absorption coefficient:

Rrs (λ) is proportional to bb (λ) / [a (λ) + bb(λ)] where bb(λ) is backscattering coefficient and a(λ) is the total absorption coefficient. The total

() () () () () *w ph g d aa a a a*

where the subscripts *w*, *ph*, *g*, and *d* relate to seawater, phytoplankton, gelbstoff and detritus, respectively. A similar equation can be wiritten for backscatter. The absorption coefficient of phytoplankton can be further broken down into contributions from various pigments to

> 1 () \*() *n*

where *ci* is the concentration of the individual pigments derived from HPLC analysis and

As the pigment composition of phytoplankton changes with changes in community structure, the absorption spectra will be modified accordingly. Furthermore, it has been well demonstrated that the absorption spectra of phytoplankton are also modified by changes in the size structure of the community (Duysens 1956; Morel and Bricaud 1981; Sathyendranath et al. 1987; Moisan and Mitchell 1999). Therefore, pigment packaging

dependent changes may also be related to changes in functional types. Remote sensing of PFTs must rely on such deviations in the spectral signatures of phytoplankton associated with changes in the phytoplankton community structure. Note that such approaches can be adapted to account for the contribution of mycosporine-like amino acids to absorption in the

UV regions, which have some but limited taxonomic value (Moisan et al. 2011a, b).

derived data that describe phytoplankton taxonomic composition.

**4. Algorithm development of ocean color phytoplankton functional types** 

PFT algorithms have been developed to map the distribution of numerically- and ecologically-important organisms that demonstrate an anomalous relationship to a satellitederived reflectance product, such as chlorophyll-a. Generally, algorithm development has capitalized on the backscattering or absorption characteristics of a particular phytoplankton group, or utilized sophisticated modeling efforts. In earlier years, a particular species/group was targeted, whereas more recently mathematically sophisticated algorithms have produced multiple products relating to phytoplankton community structure. Both approaches provide valuable data products, which enhance our understanding of the spatial and temporal variability in ecological structure and function. We present a generalized description of present algorithms for *in situ* data and satellite-

*i a Ca* 

is the absorption coefficient for the phytoplankton pigment.

 

*i ci*

ph (Figure 6). To the extent that size and function are related, size-

absorption coefficient can be partitioned into its components:

the total phytoplankton absorption.

\* ( ) *<sup>i</sup> a* 

significantly affect a\*

and center wavelengths on ocean color satellite platforms proves that coupling taxonomic and optical properties is quite challenging (Table 1). Future sensors with increased temporal and spectral resolution will provide more spectral information for development of algorithms. Future development of hyperspectral sensors would cover pigment-specific peaks at relatively high spatial and temporal resolution, which would allow retrieval of photosynthetic and photo-protective pigments.

Fig. 5. Weight-specific in vitro absorption spectra of various pigments, \* ( ) *<sup>i</sup> a* , derived from measuring the absorption spectra of individual pigments in solvent and shifting the maxima of the spectra according to Bidigare et al. (1990). Data obtained courtesy of A. Bricaud (Bricaud et al. 2004).


Table 1. Past and present Ocean Color Sensors and their respective center wavelengths (nm).

and center wavelengths on ocean color satellite platforms proves that coupling taxonomic and optical properties is quite challenging (Table 1). Future sensors with increased temporal and spectral resolution will provide more spectral information for development of algorithms. Future development of hyperspectral sensors would cover pigment-specific peaks at relatively high spatial and temporal resolution, which would allow retrieval of

photosynthetic and photo-protective pigments.

Fig. 5. Weight-specific in vitro absorption spectra of various pigments, \*

(Bricaud et al. 2004).

measuring the absorption spectra of individual pigments in solvent and shifting the maxima of the spectra according to Bidigare et al. (1990). Data obtained courtesy of A. Bricaud

**SeaWiFS MODIS MERIS OCM2 OCTS CZCS**  412 412 413 415 412 443 443 443 443 442 443 520 490 469 490 491 490 550 510 488 510 512 516 670

Table 1. Past and present Ocean Color Sensors and their respective center wavelengths (nm).

555 531 560 557 565 670 547 620 620 667

555 665

667 709

645 681

678

( ) *<sup>i</sup> a* 

, derived from

The multi-spectral radiance measurements from the satellite sensor can be corrected for the influence of the atmosphere to yield remote-sensing reflectance. It has been demonstrated using various types of models (e.g. Morel & Prieur 1977; Sathyendranath & Platt 1988) that remote sensing reflectance, Rrs (λ), at a particular wavelength (λ) increases with the backscattering coefficient of light at a wavelength, and decreases with the absorption coefficient:

$$\mathcal{R}\_{\mathbf{rs}}\left(\lambda\right) \text{ is proportional to } \mathbf{b}\_{\flat}\left(\lambda\right) / \left[\mathbf{a}\left(\lambda\right) + \mathbf{b}\_{\flat}(\lambda)\right]$$

where bb(λ) is backscattering coefficient and a(λ) is the total absorption coefficient. The total absorption coefficient can be partitioned into its components:

$$a(\mathcal{A}) = a\_w(\mathcal{A}) + a\_{ph}(\mathcal{A}) + a\_{\mathcal{g}}(\mathcal{A}) + a\_d(\mathcal{A})$$

where the subscripts *w*, *ph*, *g*, and *d* relate to seawater, phytoplankton, gelbstoff and detritus, respectively. A similar equation can be wiritten for backscatter. The absorption coefficient of phytoplankton can be further broken down into contributions from various pigments to the total phytoplankton absorption.

$$a(\mathcal{A}) = \sum\_{i=1}^{n} C\_i a \, \*\_{ci} (\mathcal{A})$$

where *ci* is the concentration of the individual pigments derived from HPLC analysis and \* ( ) *<sup>i</sup> a* is the absorption coefficient for the phytoplankton pigment.

As the pigment composition of phytoplankton changes with changes in community structure, the absorption spectra will be modified accordingly. Furthermore, it has been well demonstrated that the absorption spectra of phytoplankton are also modified by changes in the size structure of the community (Duysens 1956; Morel and Bricaud 1981; Sathyendranath et al. 1987; Moisan and Mitchell 1999). Therefore, pigment packaging significantly affect a\* ph (Figure 6). To the extent that size and function are related, sizedependent changes may also be related to changes in functional types. Remote sensing of PFTs must rely on such deviations in the spectral signatures of phytoplankton associated with changes in the phytoplankton community structure. Note that such approaches can be adapted to account for the contribution of mycosporine-like amino acids to absorption in the UV regions, which have some but limited taxonomic value (Moisan et al. 2011a, b).
