174
Joseph L. Reid
This is not to say that his zero-shear choice really meant zero flow, and was a
complete solution. Instead, the shear above and below his deep reference surface was
strong enough to show something like the major circulation nearly everywhere. His
results were not exact, of course, but much better than any that could be made from
a fixed level of no motion and would have been a great step forward if they had been
accepted.
It is surprising that this work was not taken more seriously. Swallow and
Worthington (1961), in publishing the results of their own measurements of the southward flow, note that Defant had calculated velocities of 3 to 6 cm/sec at 2000 m, but it
seems more likely that it was Stommel’s (1957) survey of current theory that suggested
their work.
After Defant (1941b) the next large-scale maps of dynamic topography were
of the Pacific Ocean, made by Fleming et al. (1945) from the Carnegie expedition
data. Although the measurements were sparse they gave a general picture of the
anticyclonic gyre of the North Pacific.
Perhaps because Defant had not explained that his reference surface was really
determined from a commonsense examination of the shear and the patterns of characteristics, not a questionable theory, dynamic topographies were not encouraged,
though transports were, by Deacon (1937) and Sverdrup et al. (1942). A very few
maps of geostrophic flow were produced for the Atlantic Ocean. Parr (1935) made
some maps of dynamic topography in the Gulf of Mexico and Montgomery (1938) in
the Atlantic. Stommel (1964) combined the maps of the dynamic topography of the
Atlantic, Indian, and Pacific oceans by Defant (1941a), Lacombe (1951), and Reid
(1961) to make a world map of the surface topography relative to 1000 decibars.
Dynamic topography has been used more extensively in the Pacific. Early examples are McEwen et al. (1930) from the United States and Canada, Uda (1935),
Koenuma (1937), and Saito (1952) from Japan, Rotschi (1958) from France, and
Wyrtki (1962) from Australia. It has been used extensively by various Soviet scientists. For example, Burkov has used it to study the circulation of the Pacific (1972)
and later in a study of the World Ocean (1980) and Maksimov (1958) used it to study
westward flow along the coast of Antarctica.
But it has been used much less extensively in the Atlantic. Is this a legacy of
doubting Defant’s results? Aside from the Equalant Atlas (Kolesnikov, 1973) almost
the only Atlantic open-ocean map of dynamic topography after Defant that I can
easily recall is that by Stommel et al. (1978). Perhaps this is the only map of steric
height that ever came out of Woods Hole. (They would have fired anyone else who
made one, but they couldn’t fire Henry Stommel.)
RUMINATIONS
Hindsight reveals that we have made some mistakes. Some come from the assumption
that something that has been generally accepted in the past must be correct, that it
Joseph L. Reid
This is not to say that his zero-shear choice really meant zero flow, and was a
complete solution. Instead, the shear above and below his deep reference surface was
strong enough to show something like the major circulation nearly everywhere. His
results were not exact, of course, but much better than any that could be made from
a fixed level of no motion and would have been a great step forward if they had been
accepted.
It is surprising that this work was not taken more seriously. Swallow and
Worthington (1961), in publishing the results of their own measurements of the southward flow, note that Defant had calculated velocities of 3 to 6 cm/sec at 2000 m, but it
seems more likely that it was Stommel’s (1957) survey of current theory that suggested
their work.
After Defant (1941b) the next large-scale maps of dynamic topography were
of the Pacific Ocean, made by Fleming et al. (1945) from the Carnegie expedition
data. Although the measurements were sparse they gave a general picture of the
anticyclonic gyre of the North Pacific.
Perhaps because Defant had not explained that his reference surface was really
determined from a commonsense examination of the shear and the patterns of characteristics, not a questionable theory, dynamic topographies were not encouraged,
though transports were, by Deacon (1937) and Sverdrup et al. (1942). A very few
maps of geostrophic flow were produced for the Atlantic Ocean. Parr (1935) made
some maps of dynamic topography in the Gulf of Mexico and Montgomery (1938) in
the Atlantic. Stommel (1964) combined the maps of the dynamic topography of the
Atlantic, Indian, and Pacific oceans by Defant (1941a), Lacombe (1951), and Reid
(1961) to make a world map of the surface topography relative to 1000 decibars.
Dynamic topography has been used more extensively in the Pacific. Early examples are McEwen et al. (1930) from the United States and Canada, Uda (1935),
Koenuma (1937), and Saito (1952) from Japan, Rotschi (1958) from France, and
Wyrtki (1962) from Australia. It has been used extensively by various Soviet scientists. For example, Burkov has used it to study the circulation of the Pacific (1972)
and later in a study of the World Ocean (1980) and Maksimov (1958) used it to study
westward flow along the coast of Antarctica.
But it has been used much less extensively in the Atlantic. Is this a legacy of
doubting Defant’s results? Aside from the Equalant Atlas (Kolesnikov, 1973) almost
the only Atlantic open-ocean map of dynamic topography after Defant that I can
easily recall is that by Stommel et al. (1978). Perhaps this is the only map of steric
height that ever came out of Woods Hole. (They would have fired anyone else who
made one, but they couldn’t fire Henry Stommel.)
RUMINATIONS
Hindsight reveals that we have made some mistakes. Some come from the assumption
that something that has been generally accepted in the past must be correct, that it