Study of Ocean Circulation since 1935
169
necessity implies that there is a fairly strong motion upstream, below the minimum
layer” (p. 39). He extended Dietrich’s calculations below 1000 m and he showed
that using top-to-bottom data on the Chesapeake Bay section with a zero flow near
800 m would imply “that the amount flowing upstream must be in excess of 78 ×
10
6 m
3 /sec, or more than double the 31 × 10
6 m
3 /sec carried downstream by the same
calculation” (p. 40). He concludes that “W¨ ust’s studies indicate that there must be a
slow movement southward of the oxygen-minimum layer in the western half of the
North Atlantic but there is no reason to assume that this motion is concentrated below
the Gulf Stream as indicated by Dietrich” (p. 40).
Iselin (1936) suggested, on the basis of the average temperature distribution
in a section between Chesapeake Bay and Bermuda, that this slow southward drift
occurs considerably east of the Gulf Stream, in the vicinity of Bermuda. He was
aware of these studies by Dietrich, Rossby, and Defant , but he held to a single deep
Gulf Stream (his Fig. 27) whose sources had been uncertain. He concluded that the
origin of this deep water was a turn back just south of the Grand Banks, leading to
a southwestward flow offshore of the Gulf Stream, the Gulf Stream return flow. He
stopped short of making the return connect to the Stream near Hatteras, but was close.
Swallow and Worthington (1961) used floats to measure the flow and find a
reference surface near 1500 m. Transports calculated from the density field gave a
deep southward flow of 6.7 × 10
6 m
3 /sec beneath a northward flow of 56.4 × 10
6
m
3 /sec.
Richardson (1985) combined surface drifters, SOFAR floats, and current meters along a section along 55
◦ W of directly measured average zonal currents in and
adjacent to the Gulf Stream. It showed the Stream with two flanking countercurrents,
the southwestward flow along the boundary and a deep southwestward flow beyond
the Stream. This was the sort of return flow Iselin had proposed, but much enlarged,
and pictured by Sverdrup et al. (1942). Part of this southward flow turns northward
near Cape Hatteras and becomes the deep Gulf Stream.
It is curious that Rossby and Iselin did not accept W¨ ust’s deep southward flow
near the Gulf Stream, which Dietrich had supported (though with a reference depth
too shallow). Perhaps the greatest obstacle to acceptance of W¨ ust’s proposal of a
southward flow along the western boundary was that it did not show a deep Gulf
Stream. W¨ ust could not find a core—a vertical extreme in salinity or oxygen—to
define such a flow, so he could not map it, or perhaps did not believe in it.
CIRCUMPOLAR WATER AT MIDDEPTH
IN THE NORTH ATLANTIC
The traditional view of the meridional flow of the deeper waters of the Atlantic was
that all of the deeper flow was northward and the flow above it was southward. The
boundary between the two flows was proposed to lie near the two-degree isotherm
in potential temperature, which rises from near 4000 m in the North Atlantic to less
169
necessity implies that there is a fairly strong motion upstream, below the minimum
layer” (p. 39). He extended Dietrich’s calculations below 1000 m and he showed
that using top-to-bottom data on the Chesapeake Bay section with a zero flow near
800 m would imply “that the amount flowing upstream must be in excess of 78 ×
10
6 m
3 /sec, or more than double the 31 × 10
6 m
3 /sec carried downstream by the same
calculation” (p. 40). He concludes that “W¨ ust’s studies indicate that there must be a
slow movement southward of the oxygen-minimum layer in the western half of the
North Atlantic but there is no reason to assume that this motion is concentrated below
the Gulf Stream as indicated by Dietrich” (p. 40).
Iselin (1936) suggested, on the basis of the average temperature distribution
in a section between Chesapeake Bay and Bermuda, that this slow southward drift
occurs considerably east of the Gulf Stream, in the vicinity of Bermuda. He was
aware of these studies by Dietrich, Rossby, and Defant , but he held to a single deep
Gulf Stream (his Fig. 27) whose sources had been uncertain. He concluded that the
origin of this deep water was a turn back just south of the Grand Banks, leading to
a southwestward flow offshore of the Gulf Stream, the Gulf Stream return flow. He
stopped short of making the return connect to the Stream near Hatteras, but was close.
Swallow and Worthington (1961) used floats to measure the flow and find a
reference surface near 1500 m. Transports calculated from the density field gave a
deep southward flow of 6.7 × 10
6 m
3 /sec beneath a northward flow of 56.4 × 10
6
m
3 /sec.
Richardson (1985) combined surface drifters, SOFAR floats, and current meters along a section along 55
◦ W of directly measured average zonal currents in and
adjacent to the Gulf Stream. It showed the Stream with two flanking countercurrents,
the southwestward flow along the boundary and a deep southwestward flow beyond
the Stream. This was the sort of return flow Iselin had proposed, but much enlarged,
and pictured by Sverdrup et al. (1942). Part of this southward flow turns northward
near Cape Hatteras and becomes the deep Gulf Stream.
It is curious that Rossby and Iselin did not accept W¨ ust’s deep southward flow
near the Gulf Stream, which Dietrich had supported (though with a reference depth
too shallow). Perhaps the greatest obstacle to acceptance of W¨ ust’s proposal of a
southward flow along the western boundary was that it did not show a deep Gulf
Stream. W¨ ust could not find a core—a vertical extreme in salinity or oxygen—to
define such a flow, so he could not map it, or perhaps did not believe in it.
CIRCUMPOLAR WATER AT MIDDEPTH
IN THE NORTH ATLANTIC
The traditional view of the meridional flow of the deeper waters of the Atlantic was
that all of the deeper flow was northward and the flow above it was southward. The
boundary between the two flows was proposed to lie near the two-degree isotherm
in potential temperature, which rises from near 4000 m in the North Atlantic to less