8
Bruce A. Warren
useful calculation, one would probably evaluate his comparison with some reserve
nowadays.
There were a few attempts to conjure up oceanwide zero-velocity surfaces as
references for geostrophic calculations; none was convincing in the end. In the Atlantic
Defant (1941) identified a finite, continuous layer at varying intermediate depth in
which (in effect) the horizontal density gradient was zero (or relatively small), and he
proposed that the horizontal velocity itself should be zero there. Even though his idea
was purely intuitive, with only esthetic justification, his reference-surface map was
influential for a time—perhaps for lack of anything more plausible. Hank Stommel
once asked him how he came to think it up. Defant’s reply, as Stommel understood
him, was that it was a “kosmischer Schwank” (cosmic prank). Many years later Fritz
Schott suggested to me that Defant may really have said “kosmischer Schwung”
(cosmic spring, or swing), because Defant had been known to recommend that one
contour sparse data points with a “kosmischer Schwung” in the hand.
Whatever he said, W¨ ust, in several publications (e.g., 1958), used Defant’s
reference surface to calculate velocities and volume transports on the Meteor sections
in the South Atlantic. His numerical values are generally doubtful, but he did show
indisputably the strong western-boundary currents in the middepth and bottom-water
layers, and his overall velocities and transports for them are probably not off by more
than 50%.
Clowes (1933) made a more rational choice of zero-velocity surface when he
calculated the transport of the Circumpolar Current through Drake Passage. Since the
tracer-property observations of the Discovery Investigations showed that the current
reached to great depth, he assumed zero velocity at 3500 m, which is roughly the
depth of the passage. His computed transport was 110 × 10
6 m
3 s
−1 , just 15% less
than the modern, directly measured value there.
Maps of the geopotential topography of isobaric surfaces (especially the sea
surface) relative to that of some deeper surface, by exhibiting patterns and strengths of
relative geostrophic flow, sharpened current maps obtained from ship drifts, provided
snapshots of surface currents, and helped to delineate subsurface flow fields. Reid’s
(1961) composite map of surface geostrophic flow for the whole Pacific, for example,
defined more surely than could Schott (1943) the subarctic gyre and the flow in the
central South Pacific and in the Southern Ocean; and it disclosed a narrow, weak
South Equatorial Countercurrent.
Quasi-synoptic maps of surface dynamic topography (apparently relative to
800 decibars), derived from repeated surveys off Japan that had begun in the 1930s,
first documented the famous great meander of the Kuroshio that comes and goes
south of Honshu at intervals of several years (Uda, 1951, 1964). This remarkable
phenomenon seems not even yet to have been explained satisfactorily. The most practical application of dynamic calculations was the preparation for many years by the
International Ice Patrol of springtime maps of surface dynamic topography (usually
relative to 1000 decibars) offshore of Labrador and Newfoundland to help predict
the possible drift of icebergs into shipping lanes. (These nicely hand-drawn maps,
Bruce A. Warren
useful calculation, one would probably evaluate his comparison with some reserve
nowadays.
There were a few attempts to conjure up oceanwide zero-velocity surfaces as
references for geostrophic calculations; none was convincing in the end. In the Atlantic
Defant (1941) identified a finite, continuous layer at varying intermediate depth in
which (in effect) the horizontal density gradient was zero (or relatively small), and he
proposed that the horizontal velocity itself should be zero there. Even though his idea
was purely intuitive, with only esthetic justification, his reference-surface map was
influential for a time—perhaps for lack of anything more plausible. Hank Stommel
once asked him how he came to think it up. Defant’s reply, as Stommel understood
him, was that it was a “kosmischer Schwank” (cosmic prank). Many years later Fritz
Schott suggested to me that Defant may really have said “kosmischer Schwung”
(cosmic spring, or swing), because Defant had been known to recommend that one
contour sparse data points with a “kosmischer Schwung” in the hand.
Whatever he said, W¨ ust, in several publications (e.g., 1958), used Defant’s
reference surface to calculate velocities and volume transports on the Meteor sections
in the South Atlantic. His numerical values are generally doubtful, but he did show
indisputably the strong western-boundary currents in the middepth and bottom-water
layers, and his overall velocities and transports for them are probably not off by more
than 50%.
Clowes (1933) made a more rational choice of zero-velocity surface when he
calculated the transport of the Circumpolar Current through Drake Passage. Since the
tracer-property observations of the Discovery Investigations showed that the current
reached to great depth, he assumed zero velocity at 3500 m, which is roughly the
depth of the passage. His computed transport was 110 × 10
6 m
3 s
−1 , just 15% less
than the modern, directly measured value there.
Maps of the geopotential topography of isobaric surfaces (especially the sea
surface) relative to that of some deeper surface, by exhibiting patterns and strengths of
relative geostrophic flow, sharpened current maps obtained from ship drifts, provided
snapshots of surface currents, and helped to delineate subsurface flow fields. Reid’s
(1961) composite map of surface geostrophic flow for the whole Pacific, for example,
defined more surely than could Schott (1943) the subarctic gyre and the flow in the
central South Pacific and in the Southern Ocean; and it disclosed a narrow, weak
South Equatorial Countercurrent.
Quasi-synoptic maps of surface dynamic topography (apparently relative to
800 decibars), derived from repeated surveys off Japan that had begun in the 1930s,
first documented the famous great meander of the Kuroshio that comes and goes
south of Honshu at intervals of several years (Uda, 1951, 1964). This remarkable
phenomenon seems not even yet to have been explained satisfactorily. The most practical application of dynamic calculations was the preparation for many years by the
International Ice Patrol of springtime maps of surface dynamic topography (usually
relative to 1000 decibars) offshore of Labrador and Newfoundland to help predict
the possible drift of icebergs into shipping lanes. (These nicely hand-drawn maps,