increase in chlorophyll is not paralleled by an increase in carbon fixing capacity
and thus results in a lower apparent assimilation efficiency (i.e., NPP/Chl
decreases). In nature, regions of low incident light and/or deep mixing also tend to
have lower SST. Thus, lower SST is broadly associated with low growth irradiance
(I g ) and, thus, lower assimilation efficiencies. This tendency is illustrated in
Fig. 8.1a where mean mixed layer light levels (I g ) exhibit roughly an exponential
positive relationship with SST when evaluated over the global open ocean. The
increase in I g with SST drives a physiological acclimation that yields increasing
assimilation efficiency with increasing SST. As a result, the implied change in
cellular chlorophyll content viewed as a function of SST exhibits a pattern
remarkably similar to expected changes as a function of I g (Fig. 8.1b).
8.3 Methods
Satellite sensors provide a range of geophysical products relevant to NPP calculations, with three central properties being chlorophyll concentration, cloudinesscorrected PAR, and sea surface temperature (SST). These 3 variables are sufficient
to initiate many NPP models (e.g., Behrenfeld and Falkowski 1997b), while other
models require additional inputs. For example, some models require information
on mixed layer depths (Howard and Yoder 1997; Westberry et al. 2008) or employ
precalculated lookup tables (Antoine et al. 1996; others). Some recent NPP models
have been developed that are based on inherent optical properties derived from
ocean color inversion algorithms, including phytoplankton absorption coefficients
(Lee et al. 1996) and/or particulate backscattering coefficients (b bp ) (Westberry
et al. 2008). In many cases, NPP models can be viewed as modular in construct, in
the sense that alternative formulations can be readily substituted. For example, the
Vertically Generalized Production Model (VGPM) of Behrenfeld and Falkowski
(1997b) is often executed with different temperature functions for P
b
opt , most often
with exponential Eppley-type dependence.
While most global-scale NPP algorithms are based on chlorophyll as the index
of standing stock, a notable exception is the model of Westberry et al. (2008). In
their approach, phytoplankton carbon concentration is inferred from satellite b bp
data and used as the core biomass index. Simultaneous satellite retrievals of
chlorophyll and carbon concentrations are then used to directly infer information
on physiological status of phytoplankton within the surface mixed layer using
understanding of Chl:C variability from laboratory studies. The Westberry et al.
model is both depth and wavelength resolved, but is also unique in that it operates
iteratively through the water column. Specifically, the model assumes biomass and
physiological uniformity within the mixed layer, and then iteratively calculates
spectral irradiance at each subsequent depth horizon as a function of integrated
changes in biomass, pigment, and attenuation directly above. The resulting light
field, in turn, defines the photoacclimation state (i.e., cellular pigmentation) of the
210
T. K. Westberry and M. J. Behrenfeld
and thus results in a lower apparent assimilation efficiency (i.e., NPP/Chl
decreases). In nature, regions of low incident light and/or deep mixing also tend to
have lower SST. Thus, lower SST is broadly associated with low growth irradiance
(I g ) and, thus, lower assimilation efficiencies. This tendency is illustrated in
Fig. 8.1a where mean mixed layer light levels (I g ) exhibit roughly an exponential
positive relationship with SST when evaluated over the global open ocean. The
increase in I g with SST drives a physiological acclimation that yields increasing
assimilation efficiency with increasing SST. As a result, the implied change in
cellular chlorophyll content viewed as a function of SST exhibits a pattern
remarkably similar to expected changes as a function of I g (Fig. 8.1b).
8.3 Methods
Satellite sensors provide a range of geophysical products relevant to NPP calculations, with three central properties being chlorophyll concentration, cloudinesscorrected PAR, and sea surface temperature (SST). These 3 variables are sufficient
to initiate many NPP models (e.g., Behrenfeld and Falkowski 1997b), while other
models require additional inputs. For example, some models require information
on mixed layer depths (Howard and Yoder 1997; Westberry et al. 2008) or employ
precalculated lookup tables (Antoine et al. 1996; others). Some recent NPP models
have been developed that are based on inherent optical properties derived from
ocean color inversion algorithms, including phytoplankton absorption coefficients
(Lee et al. 1996) and/or particulate backscattering coefficients (b bp ) (Westberry
et al. 2008). In many cases, NPP models can be viewed as modular in construct, in
the sense that alternative formulations can be readily substituted. For example, the
Vertically Generalized Production Model (VGPM) of Behrenfeld and Falkowski
(1997b) is often executed with different temperature functions for P
b
opt , most often
with exponential Eppley-type dependence.
While most global-scale NPP algorithms are based on chlorophyll as the index
of standing stock, a notable exception is the model of Westberry et al. (2008). In
their approach, phytoplankton carbon concentration is inferred from satellite b bp
data and used as the core biomass index. Simultaneous satellite retrievals of
chlorophyll and carbon concentrations are then used to directly infer information
on physiological status of phytoplankton within the surface mixed layer using
understanding of Chl:C variability from laboratory studies. The Westberry et al.
model is both depth and wavelength resolved, but is also unique in that it operates
iteratively through the water column. Specifically, the model assumes biomass and
physiological uniformity within the mixed layer, and then iteratively calculates
spectral irradiance at each subsequent depth horizon as a function of integrated
changes in biomass, pigment, and attenuation directly above. The resulting light
field, in turn, defines the photoacclimation state (i.e., cellular pigmentation) of the
210
T. K. Westberry and M. J. Behrenfeld