Looking back

We have examined, in some detail, the observations relevant to both the tropical Pacific and to ENSO, and the processes in the atmosphere and the ocean needed to explain ENSO. In the atmosphere, these processes are: the processes that anchor the regions of persistent precipitation to warm SST anomalies; the processes that determine the convergence of moisture over warm SST anomalies to maintain the regions of persistent precipitation; and the processes that determine the anomalous surface winds in terms of anomalous SST and its associated anomalous precipitation. In the ocean, the processes of interest are: the processes that change SST to produce SST anomalies; the processes that determine the depth of the mixed layer; and the processes that determine the time-dependent anomalous position of the thermocline in response to forcing by anomalous wind stresses.

Because the present complex coupled numerical models of climate do not yet simulate the tropical climatology or the phenomenon of ENSO with a sufficient degree of realism, we have concentrated analysis on the simpler (“intermediate”) coupled models. These models are simple in that they have relatively few degrees of freedom, can be run for large numbers of cases and parameter changes, and can be analyzed relatively exhaustively. Rather than repeat what we have already discussed in Chapters 7 and 8, let us summarize what is known and not known about ENSO and its predictability.