The Predictability of El Nino the 1997/98 El Nino

The Southern Oscillation, a natural mode of oscillation of the coupled ocean-atmosphere system, at times appears to be self-sustaining and hence relatively easy to predict. The 1980s seem to have been such a period; El Nino of 1982183 was the start of two complete cycles of the Southern Oscillation. At other times the oscillation seems to be a damped mode that is present for a cycle at most, after being excited by random disturbances. Disturbances that very effectively can excite El Niiio, because their surface winds have a spatial structure that coincides with those of the Southern Oscillation, are certain brief convective activities that are associated with two-week bursts of westerly winds over the western equatorial Pacific. (These convective activities do not involve ocean-atmosphere interactions that characterise the Southern Oscillation.) This type of disturbance was influential in initiating El Nifio of 1997198. That event followed El Nifio of 1992 which persisted for a surprisingly long time, and which petered out, without being followed by a significant La Nina. It appears that the Southern Oscillation was damped during the 1990s, in contrast to the 1980s when it seems to have been more self-sustaining. Consistent with this view is the performance of the coupled ocean-atmosphere models that simulate and predict El Nino; those models had far more success with the events of the 1980s than those of the 1990s. Evidence that the Southern Oscillation is subject to long-term modulations, so that it is prominent and energetic during some decades, less so during others, is available from coral records that cover a century or more. One of the factors responsible for this modulation is the time-averaged depth of the equatorial thermocline, which depends on exchanges between the tropical and extratropical oceans. Current research on the Southern Oscillation is therefore concerned with ocean-atmosphere interactions, not only in the tropics, but also in higher latitudes.