Study of Ocean Circulation since 1935
175
is an imponderable, something that one not only does not challenge, but can’t even
think about challenging. It has delayed some findings.
Why, for example, did the concept of overturn to the bottom in the Irminger
Sea endure so long? There were data available that, used properly, could easily refute
it, as was finally done by Lee and Ellett (1965).
Likewise, the spillovers from the Norwegian–Greenland Sea to the bottom of
the North Atlantic would have been obvious if people had looked at tracers, not just
on surfaces of constant depth, as in the atlases, which do not reveal directly the deep
spills, but along other surfaces (Worthington and Wright on potential temperature, or
even sigma-t, as Montgomery and Pollak did, but not far enough down).
Did W¨ ust’s proposed southward flow fail to convince Rossby and Iselin simply
because he did not show or refer to a deep Gulf Stream? His core method (the
intermediate oxygen maximum) did not apply there, and could give no sense of
northward flow. As both Rossby and Iselin knew the vertical shear there, they could
have filled in the deep Gulf Stream by choosing a reference speed in the right place,
and given us something like our present view. Why not? Defant did.
The apparent contradiction between a westward flow of the Mediterranean
water, which had to cross the southward limb of the anticyclonic gyre, need not have
arisen at all. Some people had been overwhelmed by the great anticyclonic gyre that
Munk (1950) had found in the North Pacific and Welander (1959) found in the other
oceans, and made the mistake of assuming that the gyres had the same shape all the
way down, instead of contracting and leaving room for the westward flow. Sverdrup
et al. (1942) had shown this shift much earlier.
The traditional view of the deep flow in the Atlantic, one deep layer flowing north along the bottom and another southward above it, led W¨ ust (1935) (and
how many others?) to ignore the middepth northward flow that takes place east of
his southward-flowing Middle Deep Water, and returns some of the water from the
Antarctic Circumpolar Current northward toward the sources of the southward flow
(Reid, 2005). Edmond and Anderson (1971) took data that showed it, though it was
ignored at the time. The low oxygen and high nutrients from the south can be seen
on meridional vertical sections only as separate blobs, because the northward flow
meanders as it crosses the sections.
As computers continue to become faster and the information and the ideas we
have about the circulation improve, better models will follow. But up to the present,
most of what we have learned about the circulation of the ocean has come from the
measurements that have been made. Are there any currents or large-scale patterns of
flow in the ocean that were indicated by theory or models before they were revealed
by observations?
With the recent emphasis on numerical modeling, some editors and referees
have chosen to discount and discourage descriptive studies. This is Phariseeism. With
the uncertainties of the constraints and other necessarily subjective inputs the models
cannot yet claim eternal verity, but only numerical experiments and useful conjectures
and tests.
175
is an imponderable, something that one not only does not challenge, but can’t even
think about challenging. It has delayed some findings.
Why, for example, did the concept of overturn to the bottom in the Irminger
Sea endure so long? There were data available that, used properly, could easily refute
it, as was finally done by Lee and Ellett (1965).
Likewise, the spillovers from the Norwegian–Greenland Sea to the bottom of
the North Atlantic would have been obvious if people had looked at tracers, not just
on surfaces of constant depth, as in the atlases, which do not reveal directly the deep
spills, but along other surfaces (Worthington and Wright on potential temperature, or
even sigma-t, as Montgomery and Pollak did, but not far enough down).
Did W¨ ust’s proposed southward flow fail to convince Rossby and Iselin simply
because he did not show or refer to a deep Gulf Stream? His core method (the
intermediate oxygen maximum) did not apply there, and could give no sense of
northward flow. As both Rossby and Iselin knew the vertical shear there, they could
have filled in the deep Gulf Stream by choosing a reference speed in the right place,
and given us something like our present view. Why not? Defant did.
The apparent contradiction between a westward flow of the Mediterranean
water, which had to cross the southward limb of the anticyclonic gyre, need not have
arisen at all. Some people had been overwhelmed by the great anticyclonic gyre that
Munk (1950) had found in the North Pacific and Welander (1959) found in the other
oceans, and made the mistake of assuming that the gyres had the same shape all the
way down, instead of contracting and leaving room for the westward flow. Sverdrup
et al. (1942) had shown this shift much earlier.
The traditional view of the deep flow in the Atlantic, one deep layer flowing north along the bottom and another southward above it, led W¨ ust (1935) (and
how many others?) to ignore the middepth northward flow that takes place east of
his southward-flowing Middle Deep Water, and returns some of the water from the
Antarctic Circumpolar Current northward toward the sources of the southward flow
(Reid, 2005). Edmond and Anderson (1971) took data that showed it, though it was
ignored at the time. The low oxygen and high nutrients from the south can be seen
on meridional vertical sections only as separate blobs, because the northward flow
meanders as it crosses the sections.
As computers continue to become faster and the information and the ideas we
have about the circulation improve, better models will follow. But up to the present,
most of what we have learned about the circulation of the ocean has come from the
measurements that have been made. Are there any currents or large-scale patterns of
flow in the ocean that were indicated by theory or models before they were revealed
by observations?
With the recent emphasis on numerical modeling, some editors and referees
have chosen to discount and discourage descriptive studies. This is Phariseeism. With
the uncertainties of the constraints and other necessarily subjective inputs the models
cannot yet claim eternal verity, but only numerical experiments and useful conjectures
and tests.