328
Equatorial Dynamics of the Thermocline: The Equatorial Undercurrent
presence of substantial dissipation throughout the water column. This, as we
see below, is neither necessary theoretically nor appropriate observationally
and is related to the isolation of the model current from its surroundings and
its treatment as a fundamentally two-dimensional phenomenon.
The diagnostic observational study of Bryden and Brady (1985) is particularly illuminating in emphasizing the link between the equatorial and higher
latitude ocean regimes. Bryden and Brady present a diagnostic model of the
tropical Pacific based on a series of observations in the eastern Pacific at ll0°W
and 150°W. Noting that the equatorial zone is well known for being a region of
strong upwelling, they stressed the three-dimensional character of the motion.
In the past the upwelling at the equator had been pictured largely as a twodimensional motion in the meridional plane in which poleward motion in each
hemisphere in the upper mixed layer, driven by the westward Trade Winds, is
balanced by equatorward geostrophic motion beneath the mixed layer. The
circuit in this two-dimensional picture is closed at the equator by strong upwelling across the isothermal surfaces. This upwelling across the isothermal
surfaces was considered responsible for the bowing of the isotherms observed
in the depth-latitude cross sections. Bryden and Brady's analysis of the motion
led them to suggest instead that the vertical motion at the equator is a more
complex three-dimensional cell in which deep motion approaches the equator
and then rises along isopycnal surfaces as it moves eastward in the EUC until, in
the eastern part of the ocean, the cold water follows the isopycnals to enter the
mixed layer, producing the eastern pools of cold water observed in each ocean
basin at the equator. The departure of the streamlines from the sloping isopycnals is slight and they estimated the cross-isopycnal velocity near the core of
the EUC to be only about one tenth of the vertical velocity itself. To a first
approximation the fluid glides up the isopycnal surfaces, conserving density.
The motion is fundamentally three-dimensional, and the zonal velocity of the
EUC forms a limb of a strongly slanted upwelling cell at the equator.
The conservative nature of the flow has two important and related consequences. First, if density and temperature are nearly conserved, it implies
that the explanation for the density structure of the equatorial thermocline
must be determined, in analogy with the midlatitude thermocline, as an advective process in which the density at the equator is determined by the density
of the fluid carried to the equator from the surrounding region. The current is
then not an isolated phenomenon. Second, it implies that the dynamics is
conservative, and in particular that a conservative theory of the EUC circulation is apt. Conservative dynamics means the fluid has a strong memory of its
dynamical properties (such as potential vorticity), and the lasting memory
produces a link between regions which are spatially separated.
Of course, the fact that density is conserved is suggestive but not sufficient
to mean that the dynamics is completely conservative, for now the vorticity
dynamics is not likely to simply be the inviscid Sverdrup vorticity equation.
Vertical mixing of momentum and vorticity as well as momentum and vorticity
advection might a priori be important at the equator. In this chapter we
Equatorial Dynamics of the Thermocline: The Equatorial Undercurrent
presence of substantial dissipation throughout the water column. This, as we
see below, is neither necessary theoretically nor appropriate observationally
and is related to the isolation of the model current from its surroundings and
its treatment as a fundamentally two-dimensional phenomenon.
The diagnostic observational study of Bryden and Brady (1985) is particularly illuminating in emphasizing the link between the equatorial and higher
latitude ocean regimes. Bryden and Brady present a diagnostic model of the
tropical Pacific based on a series of observations in the eastern Pacific at ll0°W
and 150°W. Noting that the equatorial zone is well known for being a region of
strong upwelling, they stressed the three-dimensional character of the motion.
In the past the upwelling at the equator had been pictured largely as a twodimensional motion in the meridional plane in which poleward motion in each
hemisphere in the upper mixed layer, driven by the westward Trade Winds, is
balanced by equatorward geostrophic motion beneath the mixed layer. The
circuit in this two-dimensional picture is closed at the equator by strong upwelling across the isothermal surfaces. This upwelling across the isothermal
surfaces was considered responsible for the bowing of the isotherms observed
in the depth-latitude cross sections. Bryden and Brady's analysis of the motion
led them to suggest instead that the vertical motion at the equator is a more
complex three-dimensional cell in which deep motion approaches the equator
and then rises along isopycnal surfaces as it moves eastward in the EUC until, in
the eastern part of the ocean, the cold water follows the isopycnals to enter the
mixed layer, producing the eastern pools of cold water observed in each ocean
basin at the equator. The departure of the streamlines from the sloping isopycnals is slight and they estimated the cross-isopycnal velocity near the core of
the EUC to be only about one tenth of the vertical velocity itself. To a first
approximation the fluid glides up the isopycnal surfaces, conserving density.
The motion is fundamentally three-dimensional, and the zonal velocity of the
EUC forms a limb of a strongly slanted upwelling cell at the equator.
The conservative nature of the flow has two important and related consequences. First, if density and temperature are nearly conserved, it implies
that the explanation for the density structure of the equatorial thermocline
must be determined, in analogy with the midlatitude thermocline, as an advective process in which the density at the equator is determined by the density
of the fluid carried to the equator from the surrounding region. The current is
then not an isolated phenomenon. Second, it implies that the dynamics is
conservative, and in particular that a conservative theory of the EUC circulation is apt. Conservative dynamics means the fluid has a strong memory of its
dynamical properties (such as potential vorticity), and the lasting memory
produces a link between regions which are spatially separated.
Of course, the fact that density is conserved is suggestive but not sufficient
to mean that the dynamics is completely conservative, for now the vorticity
dynamics is not likely to simply be the inviscid Sverdrup vorticity equation.
Vertical mixing of momentum and vorticity as well as momentum and vorticity
advection might a priori be important at the equator. In this chapter we