The response of the coupled tropical ocean–atmosphere to westerly wind bursts

Abstract

Two different perspectives on El Niño are dominant in the literature: it is viewed either as one phase of a continual southern oscillation (SO), or alternatively as the transient response to the sudden onset of westerly wind bursts (WWBs). Occasionally those bursts do indeed have a substantial effect on the SO—the unusual strength of El Niño of 1997/98 appears to be related to a sequence of bursts—but frequently the bursts have little or no impact. What processes cause some bursts to be important, while others remain insignificant? The question is addressed by using a simple coupled tropical ocean–atmosphere model that simulates a continual, possibly attenuating, oscillation to study the response to WWBs. The results show that the impact of WWBs depends crucially on two factors: (i) the background state of the system as described by the mean depth of the thermocline and intensity of the mean winds, and (ii) the timing of the bursts with respect to the phase of the SO. Changes in the background conditions alter the sensitivity of the system, so that the impact of the bursts on El Niño may be larger during some decades than others. Changes in the timing of WWBs affect the magnitude and other characteristics of the SO by modifying the energetics of the ocean–atmosphere interactions. A reasonable analogy is a swinging pendulum subject to modest blows at random times—those blows can either magnify or diminish the amplitude, depending on their timing. It is demonstrated that a WWB can increase the strength of El Niño significantly, if it occurs 6 to 10 months before the peak of warming, or can reduce the intensity of the subsequent El Niño, if it occurs during the cold phase of the continual SO. Copyright © 2002 Royal Meteorological Society.