108
Chapter 7: Provinces: The Secondary Compartments
current concepts of surface circulation, together with the distribution of oceanic frontal
zones. Original data compilations for several properties were obtained and examined:
Global climatology of mixed-layer depth: Obtained from NOAA-NODC (updated 1994
archive) on-line data, based on temperature (05
C) and density (0125 t ) criteria as
appropriate for each province. These data were arranged as monthly mean depths
on a 1
latitude and longitude grid.
Brunt-Väisälä frequency: Profiles were computed from OceanAtlas data (Osborne
et al., 1992) along many archived zonal and meridional oceanographic sections, and
also along synthetic sections derived from archived and WOCE density profiles at
standard intervals of 30
latitude and longitude.
Rossby internal radius of deformation: Obtained along meridional sections for the main
oceans from tables and maps of Emery et al. (1984) and Houry et al. (1987).
Photic depth (m): Calculated from the light field and the chlorophyll profile thought to
be typical of the season and region derived from the profile archive. The estimate is
therefore valid only for Jerlov oceanic type I-III waters (Jerlov, 1964), where phytoplankton comprise the principal source of turbidity. The photic depth is expressed
as the monthly mean depth of the 1% isolume (i.e., 1% of integrated surface light,
itself based on sun angle and regional cloudiness archives) at each 1
grid point
(Sathyendranath et al., 1995). An independent measure of water clarity was afforded
by an archive of Secchi disc depths obtained from the NODC, enabling a comparison to be made between regional, seasonal cycles of Secchi depth and of surface
chlorophyll (Fig. 7.1).
Surface nutrient fields: Obtained from seasonal multidepth climatologies (NOAANODC 1994 archives) for all regions of the ocean, issued originally in atlas format
but available also as data files for manipulation (Conkright et al., 1994).
Even these data sets failed to describe all parts of the ocean uniformly and adequately:
in particular, the South Atlantic and the South Pacific continue relatively data-poor and
are also relatively poorly understood or described. This is undoubtedly reflected in how
well the northern and southern hemisphere subtropical regions have been partitioned in
this review and will be borne in mind by the cautious reader. As we noted in Chapter 4,
in relation to the concept of “ecotones,” important regions of chlorophyll enhancement
cannot be accommodated within provinces that are defined either by circulation features
or by water mass types because the bloom is itself induced by boundary conditions
between components of the near-surface circulation. The extreme case of this situation is the linear zone of high chlorophyll that is a prominent feature in chlorophyll
images of the southern hemisphere, coincident with the austral Subtropical Convergence.
Logically, if circulation features are allowed to dominate the setting of boundaries between provinces, this zonal band of high chlorophyll values must either be split between
two adjacent provinces meeting at the convergence or assigned arbitrarily to one of them.
Either solution would be unsatisfactory, so in this case it is better to allow the chlorophyll
field to determine the boundary. The Convergence should be a province, following this
view, rather than a boundary: and so it is here.
These data and a detailed consideration of many previous proposals for partitioning
the oceans into a global set of provinces led, by a process of trial and error, to a proposal
for a series of 51 provinces. The notional boundaries of these provinces were forced
onto a rectangular grid at a scale chosen to facilitate the task of assigning data (such
as SeaWiFS surface chlorophyll values) to individual provinces: it is this grid that has
become the familiar map of provinces shown in Color plate 22.
But a static grid is clearly unreal in the sense that we have already some knowledge of
the extent to which the field of features chosen to characterize, or to bound, each province
is variable seasonally and between years. Techniques are emerging to assign dynamic
Chapter 7: Provinces: The Secondary Compartments
current concepts of surface circulation, together with the distribution of oceanic frontal
zones. Original data compilations for several properties were obtained and examined:
Global climatology of mixed-layer depth: Obtained from NOAA-NODC (updated 1994
archive) on-line data, based on temperature (05
C) and density (0125 t ) criteria as
appropriate for each province. These data were arranged as monthly mean depths
on a 1
latitude and longitude grid.
Brunt-Väisälä frequency: Profiles were computed from OceanAtlas data (Osborne
et al., 1992) along many archived zonal and meridional oceanographic sections, and
also along synthetic sections derived from archived and WOCE density profiles at
standard intervals of 30
latitude and longitude.
Rossby internal radius of deformation: Obtained along meridional sections for the main
oceans from tables and maps of Emery et al. (1984) and Houry et al. (1987).
Photic depth (m): Calculated from the light field and the chlorophyll profile thought to
be typical of the season and region derived from the profile archive. The estimate is
therefore valid only for Jerlov oceanic type I-III waters (Jerlov, 1964), where phytoplankton comprise the principal source of turbidity. The photic depth is expressed
as the monthly mean depth of the 1% isolume (i.e., 1% of integrated surface light,
itself based on sun angle and regional cloudiness archives) at each 1
grid point
(Sathyendranath et al., 1995). An independent measure of water clarity was afforded
by an archive of Secchi disc depths obtained from the NODC, enabling a comparison to be made between regional, seasonal cycles of Secchi depth and of surface
chlorophyll (Fig. 7.1).
Surface nutrient fields: Obtained from seasonal multidepth climatologies (NOAANODC 1994 archives) for all regions of the ocean, issued originally in atlas format
but available also as data files for manipulation (Conkright et al., 1994).
Even these data sets failed to describe all parts of the ocean uniformly and adequately:
in particular, the South Atlantic and the South Pacific continue relatively data-poor and
are also relatively poorly understood or described. This is undoubtedly reflected in how
well the northern and southern hemisphere subtropical regions have been partitioned in
this review and will be borne in mind by the cautious reader. As we noted in Chapter 4,
in relation to the concept of “ecotones,” important regions of chlorophyll enhancement
cannot be accommodated within provinces that are defined either by circulation features
or by water mass types because the bloom is itself induced by boundary conditions
between components of the near-surface circulation. The extreme case of this situation is the linear zone of high chlorophyll that is a prominent feature in chlorophyll
images of the southern hemisphere, coincident with the austral Subtropical Convergence.
Logically, if circulation features are allowed to dominate the setting of boundaries between provinces, this zonal band of high chlorophyll values must either be split between
two adjacent provinces meeting at the convergence or assigned arbitrarily to one of them.
Either solution would be unsatisfactory, so in this case it is better to allow the chlorophyll
field to determine the boundary. The Convergence should be a province, following this
view, rather than a boundary: and so it is here.
These data and a detailed consideration of many previous proposals for partitioning
the oceans into a global set of provinces led, by a process of trial and error, to a proposal
for a series of 51 provinces. The notional boundaries of these provinces were forced
onto a rectangular grid at a scale chosen to facilitate the task of assigning data (such
as SeaWiFS surface chlorophyll values) to individual provinces: it is this grid that has
become the familiar map of provinces shown in Color plate 22.
But a static grid is clearly unreal in the sense that we have already some knowledge of
the extent to which the field of features chosen to characterize, or to bound, each province
is variable seasonally and between years. Techniques are emerging to assign dynamic