Although notable attempts had been made to model circulation in stratified oceans in the 1950's and 1960's, the breakthroughs in determining the vertical distribution of Sverdrup transport and effects of combining wind and buoyancy forcing came in the 1980's. Luyten et al. [1983] drew on Stommel's [1979] and Iselin's [1939] ideas of how surface properties are communicated to the interior along streamlines which connect from Ekman-pumping outcropping regions into the interior of those layers. The new word which they brought to oceanography was ``subduction,'' which describes this process of communication with the interior, thus ``ventilating'' the upper ocean. The subpolar portion of that theory, in which Ekman upwelling occurs, was extended in a series of papers by Huang (see his review in [ Huang, 1991]), with an extension of their theory and the earlier thermocline theories to a continuously stratified ocean [ Huang, 1986]. The Luyten et al. [1983] model has been compared with upper ocean observations of the shallow salinity minima in the Pacific [ Talley, 1985], and with the upper ocean circulation in the South Pacific [ deSzoeke, 1987], with extensions to domains with non-zonal windstress curl (Talley) and islands (deSzoeke). Qiu and Huang [1994] have defined an ``obduction'' regime, in which streamlines irreversibly enter the seasonal mixed layer, and show both the subduction, obduction, and intermediate regimes for the North Pacific and North Atlantic.
Rhines and Young [1982] and Young and Rhines [1982] theorized that vigorous circulation in layers which do not outcrop at the sea surface can occur only in regions where the contours of potential vorticity are closed. Within these regions the potential vorticity should be homogenized by weak eddy diffusivity.
Direct surface ventilation of the upper layers of the North Pacific is quite shallow, owing to the relatively low density of the surface waters in winter, and the presence of saline deep waters which are ultimately of North Atlantic origin. Talley [1985] and Yuan and Talley [1992] show how the shallow salinity minima of the North Pacific [ Reid, 1973; Tsuchiya, 1982] can be related to the subduction of surface waters. Potential vorticity distributions on isopycnals [ Keffer, 1985; Talley, 1988] clearly show the progression from the ventilated regions of the Luyten et al. [1983] model to the interior circulation regions of the Rhines and Young model, with tongues of high potential vorticity extending around the gyre in the surface ventilated regions, and uniform potential vorticity in the unventilated but circulating regions.
North Pacific Intermediate Water
(NPIW) is a salinity minimum at a
potential density of 26.7 to
26.8
in the North
Pacific's subtropical gyre.
(
is the density in
kg/m
of a parcel of water when it
is adiabatically referenced to surface
pressure, minus 1000 kg/m
.) Its
low salinity signature is taken as
evidence of ventilation of the
subtropical gyre by waters of subpolar
origin. However, the NPIW density is
clearly greater than that of the surface
outcrops through the expanse of the
North Pacific [ Talley, 1985, 1993] and
is also somewhat greater than the
surface outcrop density in the mixed
water region between the Oyashio and
Kuroshio where the NPIW appears to
originate as a salinity minimum [ Talley
et al., 1994].
Ventilation at even greater densities also clearly occurs in the North Pacific. Signatures are: high oxygen in the northwestern Pacific [ Reid, 1965, 1973], high tritium in the western and eastern subpolar gyres [ Van Scoy et al., 1991], and chlorofluorocarbons which show penetration of surface signals to higher densities than possible through subduction of the sort defined in the Luyten et al. [1983] model [ Warner, 1988]. The mechanisms for ventilation denser than the broad-scale outcropping in the North Pacific's subpolar gyre are buoyancy forcing driven by sea ice formation in the Okhotsk Sea [ Kitani, 1973; Talley, 1991], and vertical mixing of surface properties downwards [ Reid, 1965], which may be localized to boundary regions such as the Kuril Island chain [ Talley, 1991]. Deep microstructure measurements in the North Pacific are new evidence that vertical mixing might be most important at boundaries [ Toole et al., 1994].
In the South Pacific it is more difficult to test the layered ventilation ideas since the southern boundary of the subtropical gyre is the Antarctic Circumpolar Current (ACC), across which there is a large gradient in surface density; hence outcrops extend to quite high densities within the South Pacific; lateral or isopycnic exchange or mixing across the ACC could ventilate the subtropical gyre. deSzoeke [1987] showed the congruence between the Luyten et al. [1983] model and circulation in the outcropped layers, with the densest being that of the Antarctic Intermediate Water (AAIW). McCartney [1977, 1982] suggested that the AAIW of the southern hemisphere originates as a thick layer of surface water in the southeastern Pacific, modified from similarly thick layers which advected in from the west around the ACC, and showed its signature as a thick layer around the subtropical gyre of the South Pacific.