Models of Regional Nutrient Flux and Limitation
85
but is applied only to the tropical Pacific domain and uses Tegen and Fung (1994) for
Fe input.
What is remarkable is that each of this wide range of models tracks rather well
the CZCS or SeaWIFS chlorophyll fields, with believable interyear variability. Gregg
and Conright (2001) remark that the model lacking Fe input “was able to represent
the seasonal distribution of chlorophyll during the SeaWIFS era and was capable of
distinguishing the widely different processes” that occur globally in Niño and Niña years.
The model presented global monthly maps of sea surface chlorophyll that adequately
matched SeaWIFS maps for both pattern and values except near coasts, and SeaWIFS
monthly means in 12 domains were tracked quite closely, excepting only the equatorial
domains where the lack of river input (Amazon) and coastal influences (upwelling
coasts) produced values that were overall too low. A similar model, with Fe input and
a more complex representation of phytoplankton functional groups (Gregg et al., 2003),
performed very similarly, with very similar deviation from observations in the equatorial
domains.
The model of Moore et al. (2002) likewise presents monthly global patterns of surface chlorophyll that closely match SeaWIFS observations, and global patterns of derived
properties. Standard runs were compared with runs having no Fe input or having saturating Fe input globally; both produce global maps of primary production that are
realistic and represent all the major features seen in maps of production rates computed
directly from SeaWIFS surface chlorophyll. Yentsch’s effect of geostrophy is remarkably
well preserved. The Fe-saturated model notably retains the high chlorophyll in the frontal
zones of the Southern Ocean, strongly isolated from the more oligotrophic conditions
between; the run lacking Fe input produced unrealistically low chlorophyll accumulation,
unlike the model of Gregg (2002) that also lacks an Fe input but simulates reasonable
chlorophyll biomass fields. Depending on the solubility assumption, the standard runs of
the Moore et al. model indicated annual global primary production of 42–47 Gt C y
−1 .
This range was extended to 35–70 Gt C y
−1 by including the runs with no Fe and with
saturating Fe input; direct computations from satellite data all fall within this range.
Now, what would happen if dry deposition patterns of Fe were significantly to change
in some way, so that deposition rates over the eastern Pacific resembled those over
the eastern Atlantic? If indeed skies everywhere became as dusty as under West African
harmattan winds then, after a brief period of adjustment, each regional ecosystem would
adopt a pattern of primary production and cell accumulation rather closely resembling
what we see today. Further, there would be very little change in the specific composition
of the characteristic phytoplankton communities, although the greatest uncertainty might
be in the distribution of N-fixing organisms. The eastern Pacific would continue to
accumulate chlorophyll in a manner typical of equatorial regions but quite different from
what we expect in higher latitudes, or in low-latitude oligotrophic regions. I doubt if
the sea-surface chlorophyll field of a dusty Southern Ocean would differ greatly, once
equilibrium was established, from that of the present day. The new field would continue
to reflect the simple fact that where Sverdrup does not permit growth of cell populations,
none will occur, no matter what the ratio and concentration of available nutrients.
Obviously, where Sverdrup does permit cell population growth, but this does not
occur because of lack of an essential element, then its addition—as in the iron-enrichment
experiments—must induce a bloom. That such events occur naturally is not in doubt
and there are now many observations that confirm this. For example, a response to dry
deposition of Fe-rich terrestrial dust particles was observed in the ADIOS project in the
oligotrophic, high-S North Pacific at 26
N (Young et al., 1991). Here, in an oligotrophic
ocean, when mixed-layer NO 3 was in very low concentration, each deposition event
created a brief, descending pulse (as the aerosol particles sank) of increased production
rates of about 50% over ambient. Both iron and nitrogen were delivered in the dust,
85
but is applied only to the tropical Pacific domain and uses Tegen and Fung (1994) for
Fe input.
What is remarkable is that each of this wide range of models tracks rather well
the CZCS or SeaWIFS chlorophyll fields, with believable interyear variability. Gregg
and Conright (2001) remark that the model lacking Fe input “was able to represent
the seasonal distribution of chlorophyll during the SeaWIFS era and was capable of
distinguishing the widely different processes” that occur globally in Niño and Niña years.
The model presented global monthly maps of sea surface chlorophyll that adequately
matched SeaWIFS maps for both pattern and values except near coasts, and SeaWIFS
monthly means in 12 domains were tracked quite closely, excepting only the equatorial
domains where the lack of river input (Amazon) and coastal influences (upwelling
coasts) produced values that were overall too low. A similar model, with Fe input and
a more complex representation of phytoplankton functional groups (Gregg et al., 2003),
performed very similarly, with very similar deviation from observations in the equatorial
domains.
The model of Moore et al. (2002) likewise presents monthly global patterns of surface chlorophyll that closely match SeaWIFS observations, and global patterns of derived
properties. Standard runs were compared with runs having no Fe input or having saturating Fe input globally; both produce global maps of primary production that are
realistic and represent all the major features seen in maps of production rates computed
directly from SeaWIFS surface chlorophyll. Yentsch’s effect of geostrophy is remarkably
well preserved. The Fe-saturated model notably retains the high chlorophyll in the frontal
zones of the Southern Ocean, strongly isolated from the more oligotrophic conditions
between; the run lacking Fe input produced unrealistically low chlorophyll accumulation,
unlike the model of Gregg (2002) that also lacks an Fe input but simulates reasonable
chlorophyll biomass fields. Depending on the solubility assumption, the standard runs of
the Moore et al. model indicated annual global primary production of 42–47 Gt C y
−1 .
This range was extended to 35–70 Gt C y
−1 by including the runs with no Fe and with
saturating Fe input; direct computations from satellite data all fall within this range.
Now, what would happen if dry deposition patterns of Fe were significantly to change
in some way, so that deposition rates over the eastern Pacific resembled those over
the eastern Atlantic? If indeed skies everywhere became as dusty as under West African
harmattan winds then, after a brief period of adjustment, each regional ecosystem would
adopt a pattern of primary production and cell accumulation rather closely resembling
what we see today. Further, there would be very little change in the specific composition
of the characteristic phytoplankton communities, although the greatest uncertainty might
be in the distribution of N-fixing organisms. The eastern Pacific would continue to
accumulate chlorophyll in a manner typical of equatorial regions but quite different from
what we expect in higher latitudes, or in low-latitude oligotrophic regions. I doubt if
the sea-surface chlorophyll field of a dusty Southern Ocean would differ greatly, once
equilibrium was established, from that of the present day. The new field would continue
to reflect the simple fact that where Sverdrup does not permit growth of cell populations,
none will occur, no matter what the ratio and concentration of available nutrients.
Obviously, where Sverdrup does permit cell population growth, but this does not
occur because of lack of an essential element, then its addition—as in the iron-enrichment
experiments—must induce a bloom. That such events occur naturally is not in doubt
and there are now many observations that confirm this. For example, a response to dry
deposition of Fe-rich terrestrial dust particles was observed in the ADIOS project in the
oligotrophic, high-S North Pacific at 26
N (Young et al., 1991). Here, in an oligotrophic
ocean, when mixed-layer NO 3 was in very low concentration, each deposition event
created a brief, descending pulse (as the aerosol particles sank) of increased production
rates of about 50% over ambient. Both iron and nitrogen were delivered in the dust,