158
S. George Philander
In the Pacific, interest focused on El Ni˜ no. Klaus Wyrtki’s analyses of tidegauge records
15 revealed that the latter phenomenon is associated with a horizontal
redistribution of warm surface waters across the tropical Pacific Ocean, induced by
a relaxation of the trade winds. This presumably caused the equatorial currents to
change radically. The French oceanographers in New Caledonia provided valuable
data concerning the variability of currents in the western Pacific, but no information
was available for the east. David Halpern
16 therefore developed instrumented moorings capable of measuring current and temperature fluctuations over prolonged periods
and deployed those near the Galapagos Islands. The currents of the Atlantic Ocean are
very similar to those of the Pacific, but the latter basin is significantly larger. One consequence is a dominance of the seasonal cycle in the Atlantic, of El Ni˜ no in the Pacific.
To interpret the empirical information concerning tropical variability, theoreticians developed a hierarchy of models. They range from relatively simple, highly
idealized models, described by equations that can be solved analytically, to intermediate models, of the type developed by Cane
17 for simulation of the generation of
the Equatorial Undercurrent, to the complex General Circulation Models (GCM) that
require a supercomputer. The latter models are, in principle, capable of realistic simulations but, to interpret their results, and to analyze their deficiencies, require concepts
and tools provided by more idealized models. The simple models can always be criticized for neglecting processes that in reality are important. Those models nonetheless
are of great importance. For example, the concept of a Kelvin wave emerges, not from
a GCM, but from a very simple analytical model.
For discussions of the rapid stream of new measurements and theories, Dennis
Moore provided stimulating forums—informal, semiannual meetings where a small
group of scientists engaged in lively debates. In the days before email, David Halpern
edited a hugely successful newsletter that kept everyone abreast of new results concerning similarities and differences between the three tropical oceans. Progress was
remarkably rapid. In the early 1970s very little was known about variability in tropical oceans; by 1982, oceanographers were able to document, explain, and simulate
realistically the development and decay of the most intense El Ni˜ no in more than a
century. The realism with which a GCM could reproduce the amplitude and timing of
changes in the complex of equatorial currents, undercurrents, and countercurrents
18
persuaded several skeptics that the GCM is a powerful tool for studying the oceans.
That this tool is widely used today is a tribute to Kirk Bryan and Michael Cox who
pioneered its development.
El Ni˜ no of 1982 and 1983 was an opportunity for oceanographers to show how
much they had advanced during the previous decade. However, the achievement was
flawed in an important respect. Oceanographers failed to alert the public of the devastating El Ni˜ no of 1982 in a timely manner. They were enjoying such remarkable
freedom to pursue science for the sake of science that they neglected the complementary goal of science—producing useful results. After 1982 they rapidly corrected this
imbalance. Has the pendulum now swung as far in the opposite direction? Are we
paying too little attention to science for the sake of science?
S. George Philander
In the Pacific, interest focused on El Ni˜ no. Klaus Wyrtki’s analyses of tidegauge records
15 revealed that the latter phenomenon is associated with a horizontal
redistribution of warm surface waters across the tropical Pacific Ocean, induced by
a relaxation of the trade winds. This presumably caused the equatorial currents to
change radically. The French oceanographers in New Caledonia provided valuable
data concerning the variability of currents in the western Pacific, but no information
was available for the east. David Halpern
16 therefore developed instrumented moorings capable of measuring current and temperature fluctuations over prolonged periods
and deployed those near the Galapagos Islands. The currents of the Atlantic Ocean are
very similar to those of the Pacific, but the latter basin is significantly larger. One consequence is a dominance of the seasonal cycle in the Atlantic, of El Ni˜ no in the Pacific.
To interpret the empirical information concerning tropical variability, theoreticians developed a hierarchy of models. They range from relatively simple, highly
idealized models, described by equations that can be solved analytically, to intermediate models, of the type developed by Cane
17 for simulation of the generation of
the Equatorial Undercurrent, to the complex General Circulation Models (GCM) that
require a supercomputer. The latter models are, in principle, capable of realistic simulations but, to interpret their results, and to analyze their deficiencies, require concepts
and tools provided by more idealized models. The simple models can always be criticized for neglecting processes that in reality are important. Those models nonetheless
are of great importance. For example, the concept of a Kelvin wave emerges, not from
a GCM, but from a very simple analytical model.
For discussions of the rapid stream of new measurements and theories, Dennis
Moore provided stimulating forums—informal, semiannual meetings where a small
group of scientists engaged in lively debates. In the days before email, David Halpern
edited a hugely successful newsletter that kept everyone abreast of new results concerning similarities and differences between the three tropical oceans. Progress was
remarkably rapid. In the early 1970s very little was known about variability in tropical oceans; by 1982, oceanographers were able to document, explain, and simulate
realistically the development and decay of the most intense El Ni˜ no in more than a
century. The realism with which a GCM could reproduce the amplitude and timing of
changes in the complex of equatorial currents, undercurrents, and countercurrents
18
persuaded several skeptics that the GCM is a powerful tool for studying the oceans.
That this tool is widely used today is a tribute to Kirk Bryan and Michael Cox who
pioneered its development.
El Ni˜ no of 1982 and 1983 was an opportunity for oceanographers to show how
much they had advanced during the previous decade. However, the achievement was
flawed in an important respect. Oceanographers failed to alert the public of the devastating El Ni˜ no of 1982 in a timely manner. They were enjoying such remarkable
freedom to pursue science for the sake of science that they neglected the complementary goal of science—producing useful results. After 1982 they rapidly corrected this
imbalance. Has the pendulum now swung as far in the opposite direction? Are we
paying too little attention to science for the sake of science?