106
Chapter 7: Provinces: The Secondary Compartments
that had been postulated subjectively could not be supported objectively. Available data
archives, for both biological rate processes and the distribution of biota, are inadequate
for this task and even the most elementary physical observations, such as mixed-layer
depths, are adequate only for some parts of the ocean. Elsewhere, as in the South Atlantic
and South Pacific, there are too few data even to contour seasonal mean mixed-layer
depth fields (e.g., Levitus, 1982).
However, some attempts have been made to devise an objective methodology for
deriving boundaries between meaningful provinces in oceanographic data. Using sub- and
near-surface temperature and salinity from a repeated meridional transect of the Atlantic,
Hooker et al. (2000) used the pattern of local extrema in the first spatial derivative of
density t to locate boundaries between zonal hydrographic provinces for the Atlantic
Ocean. The result was intuitive, and most of the indicated boundaries were located at
the edges of flow patterns. And, according to the authors, “if judicious choices are made
in grouping the 14 indicated provinces” the basic arrangement of the provinces used in
this book can be reproduced.
But the latter were not derived in this objective manner. Because, to be frank, it
was rather the result than the method that was paramount, I took a more classical
approach—following, for example, Colborn (1975), who derived operational partitions of
the stratification of the Indian Ocean, from the Bay of Bengal to Antarctica. He assembled
all available hydrographic data and subjectively partitioned this region into 34 provinces
by what he called their “Distinct Thermal Structure.” This was derived from the annual
cycle of stability in the upper half kilometer of the water column, and the provinces
were grouped into “Types,” distinguished by the following criteria. Is the permanent
thermocline shallow or deep? Is a summer thermocline formed in the mixed layer or
not? Does winter mixing—if it occurs—reach shallow or deep? If one were judiciously to
group those adjacent provinces having the same type of seasonal cycle, one would arrive
at a total of about eight or nine “mega-provinces” for the noncoastal parts of this ocean.
These would be the Red Sea and Gulf, the Arabian Sea, the Bay of Bengal, the equatorial
regions from 5
N to 15
S, the southern gyre down to the subtropical convergence, and
some subdivision of the Southern Ocean. As we shall see, such an arrangement would
conform quite closely to the system used in the later chapters of this book.
Colborn’s provinces, Sverdrup tells us, must also have seasonal cycles of phytoplankton
production and of everything that stems therefrom, appropriate to the observed seasonal
water column stability, transparency, and local irradiance. This is precisely what Banse
(1987) found when he merged Colborn’s compilation with CZCS surface chlorophyll to
obtain three types of “Area” in the northern Arabian Sea, each having similar phytoplankton calendars. Had Colborn performed his critical analysis for the whole globe, we should
probably have what we want. Unfortunately, he did not look beyond the Indian Ocean.
Also unfortunately for an enterprise such as this, the planetary wind systems, and
the driving forces they exert on the ocean circulation change not only from season to
season, but from year to year and also at longer time scales. Consequently, the global
circulation pattern is always changing, sometimes more and sometimes less, from the
long-term mean seasonal pattern, a problem that shall be discussed in the following
chapter. Although we know that the features we choose to delineate the provinces are in
a constant state of change, it may be necessary to draw the boundaries as if fixed in space
and time as a matter of convenience when describing the distinctive characteristics of
each, as discussed below. As Platt and Sathyendranath (1999) remarked, “When we name
a province, we give it a nominal boundary only for convenience on the other hand,
the instantaneous boundary of the province has practical value in real applications.” It
should not be necessary to emphasize that dynamic province boundaries must be the ideal
and that static boundaries can never perfectly resemble reality, which must be the ideal.
Chapter 7: Provinces: The Secondary Compartments
that had been postulated subjectively could not be supported objectively. Available data
archives, for both biological rate processes and the distribution of biota, are inadequate
for this task and even the most elementary physical observations, such as mixed-layer
depths, are adequate only for some parts of the ocean. Elsewhere, as in the South Atlantic
and South Pacific, there are too few data even to contour seasonal mean mixed-layer
depth fields (e.g., Levitus, 1982).
However, some attempts have been made to devise an objective methodology for
deriving boundaries between meaningful provinces in oceanographic data. Using sub- and
near-surface temperature and salinity from a repeated meridional transect of the Atlantic,
Hooker et al. (2000) used the pattern of local extrema in the first spatial derivative of
density t to locate boundaries between zonal hydrographic provinces for the Atlantic
Ocean. The result was intuitive, and most of the indicated boundaries were located at
the edges of flow patterns. And, according to the authors, “if judicious choices are made
in grouping the 14 indicated provinces” the basic arrangement of the provinces used in
this book can be reproduced.
But the latter were not derived in this objective manner. Because, to be frank, it
was rather the result than the method that was paramount, I took a more classical
approach—following, for example, Colborn (1975), who derived operational partitions of
the stratification of the Indian Ocean, from the Bay of Bengal to Antarctica. He assembled
all available hydrographic data and subjectively partitioned this region into 34 provinces
by what he called their “Distinct Thermal Structure.” This was derived from the annual
cycle of stability in the upper half kilometer of the water column, and the provinces
were grouped into “Types,” distinguished by the following criteria. Is the permanent
thermocline shallow or deep? Is a summer thermocline formed in the mixed layer or
not? Does winter mixing—if it occurs—reach shallow or deep? If one were judiciously to
group those adjacent provinces having the same type of seasonal cycle, one would arrive
at a total of about eight or nine “mega-provinces” for the noncoastal parts of this ocean.
These would be the Red Sea and Gulf, the Arabian Sea, the Bay of Bengal, the equatorial
regions from 5
N to 15
S, the southern gyre down to the subtropical convergence, and
some subdivision of the Southern Ocean. As we shall see, such an arrangement would
conform quite closely to the system used in the later chapters of this book.
Colborn’s provinces, Sverdrup tells us, must also have seasonal cycles of phytoplankton
production and of everything that stems therefrom, appropriate to the observed seasonal
water column stability, transparency, and local irradiance. This is precisely what Banse
(1987) found when he merged Colborn’s compilation with CZCS surface chlorophyll to
obtain three types of “Area” in the northern Arabian Sea, each having similar phytoplankton calendars. Had Colborn performed his critical analysis for the whole globe, we should
probably have what we want. Unfortunately, he did not look beyond the Indian Ocean.
Also unfortunately for an enterprise such as this, the planetary wind systems, and
the driving forces they exert on the ocean circulation change not only from season to
season, but from year to year and also at longer time scales. Consequently, the global
circulation pattern is always changing, sometimes more and sometimes less, from the
long-term mean seasonal pattern, a problem that shall be discussed in the following
chapter. Although we know that the features we choose to delineate the provinces are in
a constant state of change, it may be necessary to draw the boundaries as if fixed in space
and time as a matter of convenience when describing the distinctive characteristics of
each, as discussed below. As Platt and Sathyendranath (1999) remarked, “When we name
a province, we give it a nominal boundary only for convenience on the other hand,
the instantaneous boundary of the province has practical value in real applications.” It
should not be necessary to emphasize that dynamic province boundaries must be the ideal
and that static boundaries can never perfectly resemble reality, which must be the ideal.