To better understand what drives El Nino, an energy balance model of the coupled ocean-atmosphere system over the tropical Pacific is constructed. The atmosphere is approximated as a linear feedback system whose surface winds are driven by SST gradients and whose thermal effect is to restore the entire equatorial SST to its maximum value--the SST of the warm-pool. The upper ocean is represented by a shallow water model capped by a mixed layer with a constant depth. The zonal mean stratification of the thermocline is maintained by upwelling from the deep ocean.
The model captures the oscillatory behavior of the present tropical Pacific climate-the El Nino Southern Oscillation. The main features of the oscillation in the model agree well with the observed El Nino including the period of the oscillation and the phase relationship between the variations of SST and the variations in the depth of the
thermocline. Moreover, the model predicts that the tropical Pacific climate has two regimes: one is warm and steady, and the other is cold and oscillating, consistent with the inference from geoarcheological data that El Nino not exist during the early mid Holocene when the global and regional climate was warmer than today. The transition from
the steady climate to the oscillating climate takes place when the temperature contrast between the surface warm-pool and the deep ocean exceeds a critical value. A stability analysis reveals that the zonal SST contrast and the accompanying wind-driven currents have to be sufficiently strong to become oscillatory and that requires a ufficiently large difference between the temperature of the warm-pool and the temperature of the deep ocean.
On the time-scale of millenia, a sufficiently cold equatorial deep ocean implies a sufficiently cold high latitudes, a condition which is met by the present climate, but probably not by the climate of the early-to-mid Holocene.
In the oscillating regime, the magnitude of El Nino is found to increase monotonically with increases in the difference between the temperature of the warm-pool and the temperature of the deep ocean. The implication of these results for the response of El Nino to an increase in the greenhouse effect is discussed.