The western equatorial Pacific is characterized by mean sea surface temperatures (SSTs) in excess of 28°C, weak trade winds, and deep atmospheric convection [Bjerknes, 1969; Graham and Barnett, 1987]. While the western equatorial Pacific "warm pool" SST is relatively homogenous in comparison to other regions of the world's oceans, small O(1°C) variations in the magnitude of the SST can result in dramatic shifts in global weather patterns [Palmer and Mansfield, 1984]. Understanding the processes controlling warm pool SST variability is thus vital for understanding the global climate variability.
For timescales longer than a day, variability in the surface forcing is associated primarily with westerly wind bursts that last several days to weeks. Climatically, these westerly wind bursts tend to occur more frequently during the November-April months of a developing El Niño and result in a slackening of the trades [Barnett, 1977; Luther et al., 1983; Harrison and Giese, 1991]. Intensified wind speeds and cloudiness associated with the wind bursts contribute to surface cooling due to enhanced latent heat loss and reduced shortwave radiation [Meyers et al., 1986; McPhaden and Hayes, 1991; Zhang and McPhaden, 1995; Zhang, 1996]. Increased eastward wind stress during a westerly wind burst can also cause the surface current to accelerate eastward, and the pycnocline to deepen due to both turbulent mixing and Ekman convergence [McPhaden et al., 1988, 1992]. Additionally, heavy rainfall associated with a wind burst may produce a shallow freshwater stratified barrier layer which can inhibit vertical turbulent heat fluxes [Godfrey and Lindstrom, 1989; Lukas and Lindstrom, 1991; Sprintall and McPhaden, 1994; Anderson et al., 1996]. Thus westerly wind bursts can affect all terms in the heat balance: radiative and turbulent surface heat fluxes, turbulent mixing, and advection, with potentially significant impacts on the evolution of longer-period climate variability. Moreover, wind bursts can trigger equatorial Kelvin waves that propagate into the central and eastern equatorial Pacific, where they may lead to SST warming and possible additional climate feedbacks [Miller et al., 1988; Kessler et al., 1995].
A major goal of the Tropical Ocean Global Atmosphere-Coupled Ocean Atmosphere Response Experiment (TOGA-COARE) was to describe and understand the principal processes responsible for the coupling of the ocean and atmosphere in the western Pacific warm pool system. The intensive observation period (IOP) (November 1992 through February 1993) was chosen because wind bursts have a higher frequency of occurrence during these months. Oceanic and meteorological field work in the warm pool was predominately concentrated within the 150 km × 150 km Intensive Flux Array (IFA) centered at 2°S, 156°E. The IFA was imbedded within a larger array of Tropical Atmosphere and Ocean (TAO) moored buoys [McPhaden, 1993b] enhanced with additional moorings and instrumentation specifically for COARE [Webster and Lukas, 1992].
In the analysis presented here, the upper ocean heat balance is evaluated using enhanced instrumentation on a TAO buoy at 0°, 156°E. The buoy's deployment period (September 15 through December 20, 1992) spans the October 1992 westerly wind burst which peaked just as the COARE IOP was commencing. As discussed in section 4, the ocean response to this moderate westerly wind burst was dramatic and included a cooling of the SST by nearly 1°C, a deepening of the pycnocline by 50 m, and a reversal of the surface currents.
The data used in this analysis are described in the following section. In section 3, the details of the heat balance analysis are described. In section 4, the surface conditions, subsurface conditions, and the resulting variability in the heat balance are described. An important finding of the analysis is that heat advection can at times be the dominant mechanism controlling warm pool SST variability. In section 5, the heat flux due to entrainment mixing is estimated from the turbulent energy equation following Niiler and Kraus [1977] and compared to the residual of the observed heat balance. Section 6 investigates the importance of three-dimensional processes by comparing the observations to the variability simulated with the Price et al. [1986] one-dimensional mixed layer model. The final two sections discuss and summarize the key results of the paper.
Return to Abstract or go to next section