Study of Ocean Circulation since 1935
167
FORMATION OF BOTTOM WATER
That the bottom water of the northwestern North Atlantic is formed by deep convection
in the Irminger Sea had long been generally accepted. However, Lee and Ellett (1965)
showed that the bottom waters of the Irminger Sea were not formed by overturn
there. Their demonstration was by examination of /S diagrams and did not have the
immediate and broad impact that it should have. It was only when Worthington and
Wright (1970) mapped the salinity on surfaces of constant potential temperature that
the spills from the Norwegian and Greenland seas across the Greenland–Scotland sill,
and their deep penetration, were generally recognized.
Why was it not discovered before? Data were adequate by the time of the Meteor
expedition, or perhaps before to use the method of Worthington and Wright, as the displays of those data on maps at constant depths did not reveal the flow. Even Dietrich’s
(1969) maps of temperature and salinity at standard depths in the subarctic Atlantic
do not themselves suggest any significant flow across the Greenland–Scotland Ridge
except a shallow extension of the East Greenland Current, and no effect of it south
of there at any depth. It is curious that this had not been settled much earlier. If Parr
(1937) or Montgomery (1938) had carried out their studies on density surfaces farther
north (near 50
◦ N) instead of in the subtropics, the overflows from the Norwegian–
Greenland Sea would have been obvious. Or if Montgomery and Pollak (1942) had
not only traced the depth of some sigma-t surfaces from the Meteor Atlas, but also
traced the salinity on those surfaces, the overflow would have been seen.
THE ANTICYCLONIC GYRES
Within the subtropical zone of each ocean there is a large anticyclonic gyre. Near the
sea surface the westward limb of the gyre lies along 10
◦ to 15
◦ latitude, but below
500 m it lies poleward of 20
◦ latitude. The great anticyclonic gyres contract poleward
beneath the surface. This had first appeared in the Meteor Atlas. W¨ ust’s (1935) study
of the Atlantic found the tongue of high salinity from the Mediterranean outflow to
be flowing westward from the Straits of Gibraltar all across the Atlantic. It joins the
southern limb of the anticyclonic gyre, which flows westward directly beneath the
southward-flowing surface water, and becomes part of the deep Gulf Stream. There
is a hint of it in Iselin (1936). Defant (1941b) had mapped this feature in the North
and South Atlantic, and Montgomery and Pollak (1942) had found it from the maps
in the Meteor Atlas.
In the North Atlantic the contraction was also recognized clearly in Sverdrup
et al. (1942). However, the maps of the Sverdrup transport prepared by Munk (1950)
for the Pacific and Welander (1959) for the World Ocean showed only large anticyclonic gyres that resemble the near-surface circulation. The pattern of the deeper flow
was obscured.
167
FORMATION OF BOTTOM WATER
That the bottom water of the northwestern North Atlantic is formed by deep convection
in the Irminger Sea had long been generally accepted. However, Lee and Ellett (1965)
showed that the bottom waters of the Irminger Sea were not formed by overturn
there. Their demonstration was by examination of /S diagrams and did not have the
immediate and broad impact that it should have. It was only when Worthington and
Wright (1970) mapped the salinity on surfaces of constant potential temperature that
the spills from the Norwegian and Greenland seas across the Greenland–Scotland sill,
and their deep penetration, were generally recognized.
Why was it not discovered before? Data were adequate by the time of the Meteor
expedition, or perhaps before to use the method of Worthington and Wright, as the displays of those data on maps at constant depths did not reveal the flow. Even Dietrich’s
(1969) maps of temperature and salinity at standard depths in the subarctic Atlantic
do not themselves suggest any significant flow across the Greenland–Scotland Ridge
except a shallow extension of the East Greenland Current, and no effect of it south
of there at any depth. It is curious that this had not been settled much earlier. If Parr
(1937) or Montgomery (1938) had carried out their studies on density surfaces farther
north (near 50
◦ N) instead of in the subtropics, the overflows from the Norwegian–
Greenland Sea would have been obvious. Or if Montgomery and Pollak (1942) had
not only traced the depth of some sigma-t surfaces from the Meteor Atlas, but also
traced the salinity on those surfaces, the overflow would have been seen.
THE ANTICYCLONIC GYRES
Within the subtropical zone of each ocean there is a large anticyclonic gyre. Near the
sea surface the westward limb of the gyre lies along 10
◦ to 15
◦ latitude, but below
500 m it lies poleward of 20
◦ latitude. The great anticyclonic gyres contract poleward
beneath the surface. This had first appeared in the Meteor Atlas. W¨ ust’s (1935) study
of the Atlantic found the tongue of high salinity from the Mediterranean outflow to
be flowing westward from the Straits of Gibraltar all across the Atlantic. It joins the
southern limb of the anticyclonic gyre, which flows westward directly beneath the
southward-flowing surface water, and becomes part of the deep Gulf Stream. There
is a hint of it in Iselin (1936). Defant (1941b) had mapped this feature in the North
and South Atlantic, and Montgomery and Pollak (1942) had found it from the maps
in the Meteor Atlas.
In the North Atlantic the contraction was also recognized clearly in Sverdrup
et al. (1942). However, the maps of the Sverdrup transport prepared by Munk (1950)
for the Pacific and Welander (1959) for the World Ocean showed only large anticyclonic gyres that resemble the near-surface circulation. The pattern of the deeper flow
was obscured.