On the evolution of sea surface temperature in the tropical Pacific

Abstract

The aim of this thesis is to improve our understanding of El Niño, by using recently available observational datasets to analyze and verify mechanisms that drive El Niño. A secondary goal is to improve model simulations of El Niño, which can lead to improved forecasts.

Chapter 2 gives an analysis of how changes in SST are related to thermocline depth variability. There is a time delay between a local thermocline depth anomaly and the resulting SST anomaly at the surface. It is shown that the delay varies with longitude. Two important pathways are distinguished that cause the relation between thermocline depth and SST. The upwelling pathway (involving kelvin waves, upwelling and mixing) is found to be dominant in the eastern Pacific, approximately east of 140°W. The wind coupling pathway (involving convection, mixing and evaporation) is dominant west of 140°W.

Chapter 3 analyzes the mechanisms that are important for the development of SST anomalies in the western equatorial Pacific, the warm pool region. In a budget study it is shown how SST anomalies propagate zonally from the western Pacific to the central/ eastern Pacific and back during the development and decay of an El Niño event. An analysis of the mixed layer heat budget in an ocean model simulation shows that both zonal wind anomalies (anomalous upwelling and zonal advection) and wind speed anomalies (anomalous latent heat flux and changes in mixed layer depth) are important. Eastward propagation of SST anomalies during the growth phase of El Niño is caused partially by a reduction of the mixed layer depth east of the SST anomaly, and partially by zonal advection. Westward propagation during the decay phase is caused by warming in the western Pacific through mean zonal advection across an anomalous temperature gradient, and radiative cooling east of the SST anomaly.

The theme of chapter 4 is the question "How does El Niño change under the influence of human induced global warming?". An analysis of 62 coupled ocean-atmosphere model simulations over the period 1940-2080 shows no significant changes in ENSO characteristics, despite a simulated global average warming of ∼ 1.2K. A detailed investigation is performed to find out if flaws in the model used for the simulation are responsible for the insensitivity of El Niño to global warming. The model behavior is found to be qualitatively similar to that of a stable system driven by stochastic noise. The zonal wind response to SST anomalies in the equatorial Pacific is shown to be insensitive to changes in background SST. The zonal wind response pattern is too narrow around the equator, which leads to a more stable system, insensitive to changes in background temperature. Concluding, model deficiencies make the model insensitive to global warming. It is necessary to analyze and improve coupled general circulation models with respect to these deficiencies before they can be used answer the questions whether and how ENSO will change due to global warming.

Chapter 5 discusses the results found in chapters 2,3 and 4 and provides a summary of the conclusions.