Exchanges between hemispheres and gyres : a direct approach to the mean circulation of the equatorial Pacific

Abstract

An extensive set of new high-quality hydrographic data is assembled in order to determine the mean circulation in the equatorial Pacific, and thus the pathways for cross-equatorial and cross-gyre exchange. Making up the core of the data set are two onetime transpacific zonal sections nominally at 10°N and 14°S. Supplementing these are repeat surveys of the equatorial currents along the 165°E meridian with direct shear measurements, and repeat surveys of the western boundary current at 8°N including direct velocity measurements. The repeat survey data are crucial for obtaining a good estimate of the mean conditions in the face of strong annual and interannual variability of the near-equatorial flow field. A comparison with historical XBT and hydrographic data shows that the interior thermocline transports in the one-time sections are fortuitously representative of the mean conditions. A detailed study of the water mass distribution along the sections is the basis for choosing reference levels for the thermal wind shear in an initial guess circulation field. Using an inverse model, the initial guess circulation is adjusted such that volume, heat and salt arc conserved in a set of subthermocline layers (δΘ > 26.7). Cross-isopycnal diffusion and advection are explicitly accounted for in the inverse model, and the diapycnal diffusivity is constrained to be positive, though its value is allowed to vary with depth and location. Net mass conservation constraints are applied to the enclosed volumes of the North Pacific and eastern Pacific, and essentially require that the Ekman divergence be equal to the geostrophic convergence. The Ekman fluxes as estimated from wind-stress climatologies are an important element of the mass budget, and yet are subject to large uncertainties. The model is therefore given the freedom to determine the Ekman fluxes within the range of error of the wind-stresses. The circulation of the coldest waters (Θ < 1.2°C) is dominated by the northward flow of Lower Circumpolar Water (LCPW) in a system of narrow western boundary currents. A net transport of 12.1 Sv of LCPW flows across 14°S, 9.6 Sv of which flows into the North Pacific across 10°N. The bulk of the LCPW flux across the equator appears to occur in the denser part of the western boundary current which follows topography directly across the equator. Dissipation in the boundary layer can thus modify the potential vorticity of the fluid and allow it to cross the equator. The circulation of the upper part of the LCPW is dominated by a strong westward jet at the equator which is supplied both by upwelling from below and the recirculation of modified LCPW from the North Pacific. At mid-depth (4.0 > Θ > 1.2°C) high silica and low oxygen concentrations mark the North Pacific Deep Water (NPDW) which is present in both the North and South Pacific Oceans. Across both 10°N and 14°S, a net of 11 Sv of NPDW flows southward, returning the northward mass flux associated with the LCPW. In contrast to the LCPW, narrow western boundary currents are not present in this layer, and it is not clear how the deep water flows across the equator. Strong zonal jets on and about the equator may be important in allowing mass to cross the equator by increasing the time available for the cross-equatorial diffusion of potential vorticity to act on a fluid parcel. At intermediate depths equatorward advection is suggested by the presence of intermediate water salinity minima formed in the subpolar latitudes: Antarctic Intermediate Water dominates the 4 to 8°C classes south of the equator, while North Pacific Intermediate Water occupies this range north of the equator. Determination of the mean circulation of the intermediate waters is, however, confounded by the large eddies that dominate the geostrophic transport stream function along the onetime zonal sections. The equatorial thermocline is occupied by waters of subtropical origin: the shallow salinity minimum waters and saline Central Water from both the North and South Pacific Ocean. The equator marks the location of a front between northern and southern subtropical gyre waters, except in the lower thermocline where water from the South Pacific subtropical gyre penetrates to about 4°N to feed the Northern Subsurface Countercurrent at 165°E. All of the equatorward flowing thermocline waters are entrained in the eastward equatorial currents which in turn feed the upwelling system in the eastern Pacific. The upwelled waters largely supply the South Equatorial Current in the eastern Pacific, accounting for its large transport compared to that predicted by Sverdrup dynamics. Northward flow across the equator of the upwelled waters in the thermocline or surface layer in the western Pacific is necessary to supply the Ekman flux into the North Pacific. The analysis indicates that the Pacific Ocean does not convert a large amount of abyssal water to thermocline water, as required by several theories of the global thermohaline circulation. In contrast to the Atlantic Ocean, the thermocline circulation in the Pacific appears decoupled from the abyssal overturning, with little upwelling of abyssal waters occurring in either the North Pacific or the equatorial Pacific. The leakage of Pacific water into the Indian Ocean is deduced to be essentially zero, though an error analysis allows a range of 0-8 x 106m3s-1.