U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications
The upper-ocean profiles taken near 160°W in 1982-83 (Firing et al. 1983) reveal
some of the most dramatic equatorial current changes ever observed in the central
Pacific. The equatorial undercurrent (EUC) and near-surface South Equatorial
Current each underwent periods of rapid acceleration and deceleration. Flow
at the normal depth of the EUC core reversed direction and became westerly for
a period of more than a month. The normally westward near-surface current became,
for about two months, a surface-trapped eastward-flowing jet with a maximum
speed in excess of 1 m s
. Unusually large
displacements of the thermocline and changes in stratification were also observed.
Although no time series of multiyear duration are available for this site, mooring
data from 140°W suggest that subsurface behavior during a non-ENSO period would
involve only modest variations from time-mean conditions (McPhaden
and Hayes 1991).
Firing et al. (1983) offered a qualitative description of processes that could lead to the observed behavior, based on the surface wind stress fields produced by Jim Sadler and his coworkers at the University of Hawaii. They found it plausible to attribute a significant amount of the behavior to the remote forcing of Kelvin wave pulses by zonal wind fluctuations west of the region. McCreary and Lukas (1986) offered the hypothesis that resonantly forced Kelvin waves could account for the unusual vertical structure of the near-surface zonal velocity changes during the undercurrent deceleration phase. In a different analysis, Philander and Seigel (1985) used a multilevel primitive equation ocean circulation model forced by a monthly mean wind stress field based on the National Meteorological Center (NMC) operational analysis to carry out a hindcast of the 1982-83 ENSO. They note that the results from the model reproduce some aspects of the observed behavior, and suggest that the behavior of the ocean prior to November 1982 was likely a near-equilibrium response to their slowly varying wind field, followed by a Kelvin-type response to the abrupt return of easterlies west of the date line in November-December 1982. They do not attempt to explore the behavior at 160°W in any detail. Other studies of the generalized ENSO phenomenon, rather than of 1982-83 specifically, using simple coupled models, suggest that both eastward-propagating Kelvin waves and westward-propagating Rossby waves are central parts of the ENSO cycle (e.g., Battisti and Hirst 1988; Schopf and Suarez 1989).
Harrison et al. (1989) carried out hindcasts similar to that of Philander and Seigel (1985), using five different analyses of the monthly mean surface wind stress field. The different wind stress fields were qualitatively similar but had large quantitative differences and many differences in detail. These differences were sufficient to produce a variety of oceanic responses in the various hindcasts. Harrison et al. (1990) described the SST evolution and the processes controlling SST variation in these hindcasts and found that the warming and cooling patterns of the different hindcasts often resulted from a range of different processes. The only common elements were that the equatorial midocean warmed in mid-1982 primarily due to increased zonal advection and that the equatorial midocean cooled in late spring 1983 with the return of surface easterlies that resulted in upwelling and advection of cool water.
A comparison of the five hindcasts at 0°, 160°W with the observed data of Firing shows that the hindcast that used the SADLER wind stress field (referred to as SADLER hereafter) reproduces the observed currents and temperatures during the period from July 1982 through March 1983 well. No other hindcast reproduces so much of the observed timing and amplitude of the oceanic fluctuations. With this in mind, the SADLER hindcast experiment will be analyzed in this paper to investigate possible scenarios for the variability and the dynamics underlying this variability of the equatorial Pacific at 160°W during the 1982-83 El Niño. Other related experiments will also be analyzed to investigate the importance of local versus remote forcing and zonal versus meridional forcing. In the next section, the observations from Firing et al. (1983) at 0°N, 159°W and the results from the SADLER hindcast experiment are compared. Section 3 discusses the importance of local forcing in the SADLER hindcast, and section 4 discusses the importance of remote forcing. A brief discussion of the balance of terms in the zonal momentum and heat equations for some of the various experiments is presented in section 5. Section 6 offers some summary and discussion of our findings.
Return to Abstract or go to next section