U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications

Dynamics of seasonal and intraseasonal variability in the eastern equatorial Pacific

Michael J. McPhaden and Bruce A. Taft

NOAA/Pacific Marine Environmental Laboratory, Seattle, Washington

Journal of Physical Oceanography, 18(11), 1713-1732 (1988)
Not subject to U.S. copyright. Published in 1988 by the American Meteorological Society.

2. Background

Zonal wind stress in the equatorial Pacific is on average strongest between about 120 and 160W (Fig. 1a). This representation of the wind field, based on 23 years of ship data (Goldenberg and O'Brien 1981), indicates stress values of 25 10 and 55 10 N m at the 110 and 140W mooring locations, respectively. Sverdrup's (1947) linear theory of the mean circulation predicts that along the equator,

where P is depth integrated pressure gradient and is zonal wind stress. For the winds in Fig. 1a, this leads to a rise in depth integrated pressure of about 1.5 10 N m between 110 to 140W (Fig. 1b). The theory also predicts that zonal transport per unit width (hereafter referred to as transport) should be about 10 m s along the equator (Fig. 1c). One would not expect the Sverdrup balance to be valid at all times and locations however because of the importance of nonlinearity and planetary scale wave processes.

Figure 1. (a) Mean zonal wind stress (in 10 N m) from 1961-83 from ship wind analyses (Goldenberg and O'Brien 1981) using a drag coefficient of 1.2 10 and air density of 1.2 kg m; (b) depth integrated pressure (in 10 N m) from the Sverdrup (1947) balance. Values are relative to an arbitrary integration constant chosen as 5 10 N m; (c) Zonal transport per unit width (in m s) from the Sverdrup relation. Dashed lines in (c) indicate negative values. Positions of the current meter moorings are shown by the solid dots at 110, 124.5 and 140W.

Philander and Pacanowski (1980) have shown in nonlinear numerical simulations of the Equatorial Undercurrent (EUC) that meridional advection in the thermocline concentrates eastward momentum on the equator, where it is upwelled towards the surface. Similarly, westward momentum put in at the surface by the easterlies is carried poleward in a shallow Ekman layer. This produces a narrower, shallower EUC with larger eastward mass transports relative to linear models (e.g., McPhaden 1981; McCreary 1981). Nonlinear enhancement of eastward transport along the equator can alternately be inferred from the tendency of the flow to conserve absolute vorticity in the presence of an eastward pressure gradient force (e.g., Fofonoff and Montgomery 1955; Cane 1980; Pedlosky 1987).

Nonlinearity in the Pacific EUC has also been examined empirically from velocity section data (Knauss 1966) and several months of moored current meter data (Halpern 1980). These studies indicate that advective terms in the zonal momentum balance may be comparable in magnitude to the zonal pressure gradient. The representativeness of these results is unclear, though, because of the limited data on which they were based.

Ocean current and temperature fields near the equator adjust to wind stress variations on monthly time scales primarily through the excitation of equatorial baroclinic Kelvin and long Rossby waves. Adjustment can occur very rapidly relative to midlatitudes because these waves have O(1 m s) zonal phase and group speeds. Theory and modeling studies suggest that in a basin the width of the equatorial Atlantic, wave processes can adjust to ocean to equilibrium on time scales less than a year (Cane and Sarachik 1981; Philander and Pacanowski 1981). This probably accounts for the observational result that zonal wind stress and pressure gradient nearly balance at the annual period in the equatorial Atlantic (Katz et al. 1977, 1986).

There is less agreement on whether the equatorial Pacific should be in equilibrium with winds at the annual cycle. Cane and Sarachik (1981) argued that if the wind forcing were uniform across the basin, the eastern two-thirds of the Pacific should respond in a manner similar to the Atlantic at the annual cycle. Philander (1979) on the other hand suggested that the Pacific would be further from equilibrium than the Atlantic because of its larger zonal extent. Philander and Pacanowski (1981) showed that if significant trade wind variability were confined to a band of meridians--as it is in the equatorial Pacific (e.g., Meyers 1979; Lukas and Firing 1985)--then in the directly forced region the ocean should respond in a succession of steady states at the annual cycle.

Attempts to directly estimate the dynamical response of the equatorial Pacific to fluctuating winds have primarily relied on ship wind and hydrographic data. Mangum and Hayes (1984) and Bryden and Brady (1985) used NORPAX and EPOCS CTD data from 1979-81 to show that in the mean there is a balance between zonal wind stress and pressure gradient along the equator between 110 and 150W. However, Mangum and Hayes (1984) found near zero correlation between time varying zonal winds and pressure gradients from eight quasi-synoptic pairs of CTD casts at 110 and 150W. Tsuchiya (1979), on the other hand, used data from the EASTROPAC Expedition in 1967-68 to show that between 95 and 120W, both easterly wind stress and zonal pressure gradients of comparable magnitude were weak from March to May and strong in June and July. Meyers (1979) computed a mean seasonal cycle of 14C isotherm depths using mechanical and expendable bathythermograph data and compared variability in isotherm slopes with that seen in climatological winds (Wyrtki and Meyers 1975). In contrast to Tsuchiya (1979), he found that in the eastern and central equatorial Pacific the slope (which was used as a proxy measure of zonal pressure gradient) was weakest in May-June and strongest in October-November, whereas the wind stress was weakest in March-April and strongest in July and December.

The lack of agreement between these studies may be related to aliasing of high frequency energy in both wind and oceanic data. Monthly or quarterly hydrographic measurements along the equator, for example, are not capable of resolving tides (Weisberg et al. 1987), high frequency inertia-gravity waves (Hayes 1981), 20-30 day instability waves (Legeckis et al. 1983), O(10 day) Kelvin wave pulses generated in the western Pacific (Knox and Halpern 1982; Eriksen et al. 1983; McPhaden et al. 1988), or energetic intraseasonal variations at periods of 40-60 days (Enfield 1987). Similarly, infrequent ship wind measurements are subject to large uncertainties due to the presence of pronounced synoptic scale variability in the atmosphere (Luther and Harrison 1984; Halpern 1987d). Better definition of seasonal dynamics thus requires more finely resolved time series measurements that may be filtered to remove high frequency energy.

Go back to previous section or forward to next section

PMEL Outstanding Papers

PMEL Publications Search

PMEL Homepage