Ocean Wave Dynamics and El Niño

The response of an ocean general circulation model to specified wind stress is used to understand the role of ocean wave propagation in the evolution of El Niño events in sea surface temperature (SST) in the equatorial Pacific Ocean. In a control experiment the ocean model reproduces observed equatorial Pacific interannual variability in response to forcing by the observed wind stress. The ocean model is then forced with the same wind stress but with the time evolution of the wind stress forcing reversed. An analysis of the anomalies from the annual cycle in these two experiments delineates the parts of the response that are in equilibrium with and out of equilibrium with the wind stress forcing. In the western equatorial Pacific the SST anomalies are in equilibrium with the wind stress anomalies. In the eastern equatorial Pacific the SST anomalies are not in equilibrium with the wind stress anomalies but rather tend toward equilibrium with the upper-ocean heat content anomalies. The experiment demonstrates that the heat content is not in equilibrium with the wind stress forcing either on or near the equator. Very close to the equator the slope of the thermocline is in equilibrium with the wind stress, but the mean heat content is far from equilibrium. Slightly off of the equator in the western Pacific westward propagating heat content anomalies appear to originate in regions of strong wind stress forcing and then propagate to the western boundary. These westward propagating anomalies also depart significantly from equilibrium with the wind stress forcing. Additional experiments allow these westward propagating anomalies to be identified as freely propagating Rossby waves. The Rossby waves are shown to determine the equatorial heat content response to the wind stress forcing when they arrive at the western boundary and to be responsible for the nonequilibrium behavior of the equatorial mean heat content. A simplified coupled model is derived by fitting the results and estimating parameter values from the numerical experiments. In this model the time delay introduced by the Rossby waves is responsible for coupled instability, and dependence of the Rossby wave amplitude on the frequency of the forcing is responsible for the long period of the coupled oscillations.