Search

In a numerical model of the equatorial Pacific Ocean, the ∼20-day period tropical instability waves, excited in the eastern half of the domain, are found to damp the strong zonal mean currents. The waves generate large, nonlinear, advection terms in the momentum balance, change the vorticity balance, and thus modulate the low-frequency state. The authors explore whether the effect of tropical instability waves on the background flow can instead be adequately parameterized by a constant-coefficient Laplacian friction scheme. On annual mean, a Laplacian friction coefficient that varies in space is required, for the coefficient is twice as large along the equator and a few degrees more to the north than elsewhere. In addition, wave activity varies in time. During active phases, such as the second half of the year and during La Niñas, the activity increases, which would require the Laplacian coefficient of friction to be at least twice as strong as during the inactive phases. Thus, a more sophisticated damping parameterization than simple Laplacian friction is required in ocean models that do not explicitly resolve tropical instability waves. [ABSTRACT FROM AUTHOR]

Brown et al. analyzed the kinematics of flow in the equatorial Pacific Ocean, along time-varying isopycnals in a three-dimensional eddy-permitting model. Here the dynamics of these flows is explored in the same model via the zonal momentum equation (ZME). Previous work has shown that the dominant terms of the ZME, on and near the equator, are the pressure gradient, wind stress, and Coriolis term. In one model study, the nonlinear and friction terms were significant but negated each other. In this study, with a higher-resolution model and more realistic friction scheme it is shown that the nonlinear term is important along and north of the equator, while the explicit friction term is negligible. The part of the nonlinear term derived from high-frequency eddy flows acts like a friction on the Equatorial Undercurrent, while the remaining part of the nonlinear term from smooth flows enhances it. In density coordinates, meridional tropical cells lie on either side of the equator in the first half of the year (January–June) as expected. In July–December, a continuous southward surface flow appears from 4°N into the Southern Hemisphere and arises from variations in the geostrophic flow and the nonlinear term. Variations in the geostrophic flow are due to both seasonal variability in the thermocline and a surface bolus effect arising from baroclinic instability. The nonlinear term increases in the surface layers at the same time assisting the southward flow, most likely because of tropical instability waves. [ABSTRACT FROM AUTHOR]

An eddy-permitting global ocean model is used to interpret kinematics within the central and eastern equatorial Pacific Ocean, from 160°E to the coast of America. Because of high levels of variability in this region, observational studies of meridional flow are contradictory, in particular as to whether the net flow is northward or southward. Unlike most oceanographic datasets, model output can be analyzed at high temporal and spatial resolution, providing clues as to real ocean behavior. In the model, a net southward flow occurs across the equator east of 160°W, at most density layers throughout the year. In the central Pacific, from 160°E to 160°W, the net flow is northward but varies with season and occurs primarily in the mixed layer. This is a key region for the flow of Equatorial Undercurrent water into the Northern Hemisphere. The three-dimensional flow is very complex and seasonally dependent. It is vital that these flows are analyzed in an isopycnal framework, or else the pathways are very misleading. In the first half of the year, evidence is found of meridional tropical cells on either side of the equator out to ±5°. These cells appear to exist without any need for diapycnal downwelling. In the second half of the year, when tropical instability waves are active, the cells are overlaid by a strong surface southward flow that appears to be a bolus-type transport. This transport is not apparent unless the flow is calculated in the aforementioned manner. [ABSTRACT FROM AUTHOR]