Thesis Title: Analyses of Upwelling Events in the Gulf of Guinea using Satellite Observational Data and Model Outputs

Abstract: This thesis presents the analysis of upwelling events in the Gulf of Guinea and the Equatorial Atlantic region. The upwelling events in the West African subregion are usually experienced in the Gulf of Guinea, impacting the West African coastlines. During upwelling events, the wind blows across the water surface; a vacuum is created and filled with cold water and nutrient from the deep part of the ocean’s surface. This process leads to heat distribution in the ocean, preserving the ocean’s temperature. Classically, the upwelling events in the Gulf of Guinea were assumed to be generally caused by wind, Coriolis force, and Ekman transport. However, this research demonstrates that the wind that blows along the coast is usually in the north-northeastern direction. This means the correlation between the local wind stress and sea surface temperature (SST) anomaly in the Gulf of Guinea is considerably smaller. 

Early research on the upwelling events in the eastern and tropical Atlantic region used trends and patterns from observations and mathematical numerical models to examine the cause of upwelling in the Atlantic region. [13] and [3] hypothesized that upwelling in the Gulf of Guinea is associated with Kelvin waves propagating eastward from the Brazillian coast along the equator. The Kelvin waves are trapped along the coast once they reach Equatorial Guinea as coastally-trapped Kelvin waves. A major goal of this study is to test this hypothesis using satellite observational data and output from the state-of-the-art Estimating the Circulation and Climate of the
Ocean (ECCO) model.

We first investigate classical mathematical theory to understand the dynamics of upwelling. Exploring the classical theory of Ekman transport, Ekman pumping and suction will aid in explaining the reasons for the upwelling in the Gulf of Guinea. Based on [3], some mathematical derivations of the Kelvin waves theory will be used to find the features of the eastward propagating Kelvin waves on the ocean’s surface.

Data collected from the Prediction and Research Moored Array in the Atlantic (PIRATA) will aid the understanding of the interactions between the ocean and atmosphere in the tropical Atlantic region. Temperature as a function of depth and time is collected from four different locations to understand how the ocean’s temperature varies with depth. The rise of the thermal structure along the Atlantic region makes the thermocline shallower, leading to coastal upwelling, as discussed by [2] and [3]. June, July, August, and September usually are associated with low sea surface temperature (SST). However, the low SST is not enough evidence of what causes upwelling in the Gulf of Guinea, even though SST is important.

Sea surface height (SSH) signals from observations (satellite) and model data (ECCO) are composed of various waves whose characteristics and structures are different in terms of their period, wavelength, frequency, amplitude, and phase speeds. The Kelvin waves extracted from satellite data are examined and compared with the
Kelvin waves extracted from model data. The phase speed of the Kelvin waves from observations and model data is about 1.8 m/s which is consistent with the result of [7]. The lag correlation between Kelvin waves from observation and ECCO for some specific years is good. The question is, how does the observational data compare with the model data, and can the model data be useful for future predictions? It is observed from analyzing the results that the error between Kelvin waves from observations and ECCO was getting smaller in recent years. This result shows a marginal improvement in the model.

Some selected parameters in a few areas of relevance in the Atlantic region, such as wind stress from the Brazilian region and SST from the Guinea-west region, were considered to examine the upwelling analysis used by [6], to explain the remote influence of upwelling in the Gulf of Guinea region. The results obtained by [6] were
reproduced using recent observational data. In addition to the work by [6], model data from ECCO is used to produce similar results, supporting the hypothesis of [13].

PhD Candidate: Patrick Dwomfuor

Viva Link: Click here to join