The Fourier analysis has been applied to observational sea surface tem
perature (SST) data from the Southern Pacific. The spectral response o
f the daily temperature fluctuations indicates that the so-called El N
ino Southern Oscillation (ENSO) belongs to the class of dynamical phen
omena which are in a self-organized critical state. This has implicati
ons on the predictability of the significant events in the ocean-atmos
phere interaction process. A toy model is used to point out the simila
rities of this system with other large scale phenomena, as earthquakes
and volcanic eruptions, which are considered to be consistent with th
e hypothesis of self-organized criticality. The qualitative agreement
between observational data and results of numerical simulations demons
trates the validity extent of the theoretical approach.