Seasonal Variability of the Buoyancy Flux along the Northern Coast of the Gulf of Guinea ()
1. Introduction
The Gulf of Guinea (GG) is located in the Atlantic Ocean from 20˚W to the western coast of Africa at cap Lopez (8˚43'E) and from 10˚S to the African coast around 5˚N (Figure 1). This region is characterized by an Atlantic cold tongue along the equator [2] [3] and a seasonal upwelling is observed from December to March and from July to August along its northern coast [4] [5] [6] [7]. The
![]()
Figure 1. Sea surface temperature in the Gulf of Guinea (courtesy from [1] ). The two regions of weak temperature represent respectively the Atlantic cold tongue along the equator and the coastal upwelling off Cote d’Ivoire and Ghana.
coastal area and particularly, the region of Cote d'Ivoire and Ghana retaining the attention of this work are more complex because of being the principal region of the occurrence of the coastal upwelling and also the area of freshwaters intrusion from the tropical rainfall, the rivers and the lagoons [4] [8]. This coastal upwelling is very important for the climate of the bordering countries of the GG, the marine ecosystem of this region, the phytoplankton dynamic and, the fishery [4] [5] [6] [7] [9] [10] [11]. Unfortunately the physical mechanisms of this upwelling are still discussed [5] [12] [13]. Many mechanisms have been put forward to explain this upwelling, among them, the cooling by the mixing due to current shear in the mixed layer seems to be a dominant process [13]. Few studies on the mixed layer depth have shown that this depth is very shallow during the whole year [14]. But no other study has tried to understand why the mixed layer is shallow, and especially the contribution of the Buoyancy flux in the mixing of the mixed layer in this area. This is one of the main reasons of this work.
The Bo is important when analyzing the mechanisms of the ocean-atmosphere interactions [15]. The Bo plays an important role in water mass transformation, mixing, stratification of water levels [15] [16], gas transfer in ocean [17], vertical nutrient transport [11] and, then on the upwelling. This important parameter depends on the heat and the freshwater fluxes [15]. One of the challenges involving the Bo study for a particular region is relative to the determination of the quantity that allows capturing of the relative contribution of the two components (thermal and haline) of the buoyancy flux in one single quantity [15] [18] [19] [20]. The net Bo is dominated over a larger part of the ocean area by the thermal component Bq [21]. But it is the concomitant action of both that determine the impact of atmospheric and oceanic changes on the ocean ventilation [15] [22] [23]. Unfortunately, the previous studies on ocean-atmosphere interaction did not focus on the Bo in the GG. So which of the thermal or the freshwater buoyancy flux dominates the buoyancy flux in the GG? Thus, the present study focuses on the: 1) net Bo and, 2) determination of the dominant component of the Bo in the GG. An application of the Bo to the determination of the mixing layer has been further evaluated in this region. Besides, the significance of these results for the dynamic of the mixed layer and then for the coastal upwelling of this region has been discussed.
In this work, we have first combined the oceanic and atmospheric reanalyses in order to compute the Bo and from the sign of the Bo, we have indirectly estimated the contribution of the Bo to the oceanic mixing, the mixing layer and the coastal upwelling of this region. The outline of this work is presented as follows. Section 2 describes the data and gives an overview of the methods used to estimate the buoyancy flux and the mixing layer. The results are presented in Section 3. The results and the significance of this work have been discussed and concluded respectively in Sections 4 and 5.
2. Data and Methods
2.1. Data
This study combines the atmospheric and the oceanic reanalyses data provided respectively by the European atmospheric reanalysis ERA5 and the new eddy-permitting ocean reanalysis called Ocean Reanalysis Pilot (ORAP5). These reanalyses have been both completed at the European Centre for Medium-Range Weather Forecasts (ECMWF) and they span respectively over the period 1979-2012 for the ORAP5 [24] and 1979-onwards for the ERA5 [25]. The ECMWF produces the ERA5 and the ORAP5 reanalyses within the Copernicus Climate Change Service. The ERA5 provides global hourly estimates of the atmospheric variables at a spatial resolution of 31 km and 137 vertical levels from the surface to 0.01 hPa. The ERA5 has been improved from its predecessors with some improvements in the models, the dynamic cores and the data assimilation [25]. The atmospheric data used are the daily precipitation, the evaporation and the wind at the sea surface. The unit of the precipitation and the evaporation is m/day. The oceanic ORAP5 reanalysis is produced using the NEMO ocean model [26] at a spatial resolution of 25 Km between two consecutive grid points and 75 levels in the vertical [25]. For this study, the oceanic data used are the net heat flux Q0 at the sea surface and, the sea surface salinity (S) and temperature (SST). Q0, S and SST are monthly data. The units of Q0 and S are respectively W∙m−2 and psu. The data used for this study cover the period 1995-2005.
2.2. Method
The surface buoyancy flux depends on the net heat flux or the thermal condition (Bq) and the variability of the freshwater amount or the haline component (Bp). The Bo is calculated from the atmospheric and the oceanic parameters at the sea surface [15] [16] [18] [19]. The buoyancy flux is calculated as follows according to [18]:
The gravity is g = 10 m∙s−2; Q0 the net heat flux at the sea surface, ρ the ocean density computation is based on the equation of sea water, while constant values are α the effective thermal expansion coefficient and β the effective haline contraction coefficient. The evaporation (E) and precipitation (P) data are divided by 86,400 s to have the m/s unit. Finally, the unit of the Bo is m2∙s−3.
A negative (i.e., downward) buoyancy flux, due to either the surface warming or the precipitation, tends to make the ocean surface more buoyant and subsequently stable. Conversely, a positive Bo, due to either the surface cooling or the evaporation, tends to make the ocean surface less buoyant. As the water column loses buoyancy, it can become unstable with heavy water lying over lighter water [18].
One important application of the Bo is the determination of the mixing layer, the depth at which the turbulence generated by the wind is balanced by the Bo.
The wind friction is
with τ corresponding to the wind stress at the
surface and ρo the density of surface water. The wind data are obtained from atmospheric reanalysis ERA5. The wind stress is calculated from the wind at 10 m above the sea surface using the drag coefficient determined by the relationship of [27]:
The mixing layer depth or the Monin-Obukhov length L [28] is:
;
with k = 0.41 the Von Karman’s constant.
3. Results
3.1. Temporal Variability of the Buoyancy Flux and Its Components
Figure 2 shows the climatology of the Bo, Bq, Bp and the ratio of the thermal over the haline buoyancy fluxes along the northern coast of the GG. The Bo is negative in the GG. Its weakest values are around −4 × 10−8 m2∙s−3 and they are observed over the region of the coastal upwelling of the GG. These values increase offshore tending to positive values. The highest values around −0.5 × 10−8 m2∙s−3 are observed over the warm pool between 2˚N and 4˚N.
The thermal buoyancy flux Bq also displays similar patterns as the net Bo with a core of negative values ranging between −3 × 10−8 m2∙s−3 and −2 × 10−8 m2∙s−3 in the coastal upwelling region. These values increase southward and reaching −0.1 × 10−8 m2∙s−3 around 2˚N. The positive values are observed in the West of 9˚W over the warm pool.
The haline buoyancy flux Bp is negative in the GG. The highest values of Bp are recorded along the northern coast while the minimum values are in the south and particularly over the warm pool. The values along the coast are higher than −1 × 10−8 m2∙s−3 and they decrease to values more and more negative offshore. The ratio of the thermal over the haline buoyancy flux is maximum along
![]()
Figure 2. Spatial distribution of the climatology of the net buoyancy flux (panel (a)), the thermal buoyancy flux (panel (b)), the haline buoyancy flux (panel (c)) and, the ratio of the thermal over haline buoyancy flux (panel (d)) off Cote d’Ivoire and Ghana. The data have been averaged over the periods 1995-2005 to build the climatology. Intervals between contours are 0.2 × 10−8 m2∙s−3 for the different buoyancy fluxes. The core of negative Bo is observed off Cote d’Ivoire and Ghana in the region of the coastal upwelling.
the northern coast of GG and minimum in the south. The ratio is greater than 3 along the coast and decreases southward reaching values less than 1. The Bt flux is more important at the coast and contributes more to the net buoyancy flux there. While south of 3˚N, the Bp becomes more important than the Bq. The ratio of the thermal over the haline buoyancy fluxes shows that the haline buoyancy flux represents one fourth of the thermal buoyancy flux along the coast. In the South of 3˚N, the Bp is greater than the Bq and the ratio shows that Bp corresponds to one half of the Bq.
Figure 3 shows the seasonal variability of the Bo and its components along the northern coast of the GG. The Bo and the Bq are characterized by a seasonal variability with maximum values in June and two minima in March and October. The Bp is negative all the year and it has two maxima respectively in January and August-September. The Bo decreases from January to reach a minimum and negative value of −4 × 10−8 m2∙s−3 in March before increasing until June when it reaches an absolute maximum value of 2 × 10−8 m2∙s−3. It decreases from its maximum value in June to reach minimum values in July, August, September and October. The absolute minimum value around −5 × 10−8 m2∙s−3 is observed in October. It increases from October to a maximum value in December-January. The thermal buoyancy flux governs almost 70% of the variability of the net buoyancy flux with the peak of highest values around 1 × 10−8 m2∙s−3 when the net buoyancy flux reaches 2 × 10−8 m2∙s−3 and the haline buoyancy flux is around −1 × 10−8 m2∙s−3.
3.2. Variability of the Mixing Layer
Figure 4 shows the climatology of the mixing layer along the northern coast of the GG. The mixing layer is very shallow along the coastline north of 3˚N and
![]()
Figure 3. Seasonal variability of the buoyancy, the thermal and the haline fluxes along the northern coast of the Gulf of Guinea. The data have been averaged in the coastal domain [7˚W-3˚E × 3˚N-5˚N] for the respective months of the calendar to get the monthly climatology.
![]()
Figure 4. Spatial distribution of the climatology of the mixing layer along the northern coast of the Gulf of Guinea. The monthly data have been averaged over the period 1995-2005 to build the monthly climatology. Intervals between contours are 2 m.
is deeper south of this latitude. It stays almost constant along the coastline from the Cap Palmas to Ghana. It increases southward to bigger values. The values are less than 5 m North of 3˚N while it increases quickly South of this position reaching 30 m around 2˚N over the warm pool. The weakest values of the mixing layer are recorded over the region of the weakest Bo while the strongest values of the mixing layer coincide with the region of the highest Bo in the GG.
4. Discussion
This study uses the net heat flux at the sea surface, the evaporation and the rainfall data from the oceanic (ORAP5) and the atmospheric (ERA5) reanalyses to describe the net buoyancy flux Bo along the northern coast of the GG. The analysis of the results showed that the Bo is negative and the mixing layer is shallow along the coast. The Bo increases the stratification which reduces the mixing and consequently the mixing layer [15] [18]. This could explain the shallow mixed layer observed along the northern coast of the GG [14]. The peak of the negative Bo is noticed since the period of the major upwelling in July-August-September till October [1] [4] [5] [6] [7]. This shows that the Bo will reduce the intensity of the mixing during the upwelling season and consequently works again the mechanism of cooling by mixing during this upwelling [13]. The role of the buoyancy flux on the mixed layer along the northern coast of the GG would have a negative impact on the phytoplankton dynamic and the fishery of this area [11]. The negative buoyancy is due to the heat flux that warms strongly the mixed layer during the upwelling season. The increasing trend of the global warming will increase the role of the buoyancy flux and consequently the stratification of the mixed layer along the northern coast of the GG.
The thermal buoyancy flux Bq dominates the Bo off the coast but offshore, the freshwater buoyancy flux Bp is very important. The ratio Bq/Bp is in favor of the Bq, but this method of analyzing the flux ratio [16] cannot be fully used in the GG because the two components of the Bo are both negative over large domain of the GG excepted in the west of 9˚W over the warm pool where the Bp is positive. Also this ratio method presents some disadvantages in the GG because, it gives sometimes nil values [15]. Most of the studies on the surface ocean buoyancy forcing focus on the thermal component because it controls the seasonal variability of the Bo [21]. In the GG, the freshwater buoyancy flux represents one fourth of Bq and then could not be neglected. Besides, [22] and [23] sustain that the haline buoyancy flux is crucial for long-term variability of the atmosphere/ocean system. In the Indian Ocean, [20] didn’t include evaporation in the freshwater component of the Bo. In the GG, the evaporation is very important for the estimation of Bp and for the dynamic of the coastal upwelling [7]. [7] attributed the SST cooling during the major coastal upwelling along the northern coast of GG to the evaporation showing then the role of the mixing due to positive buoyancy flux and therefore the crucial role of Bp among the physical mechanism of this coastal upwelling. Unfortunately, our study shows that buoyancy flux inhibits the upwelling.
5. Conclusion
This study describes the buoyancy flux and its components in the GG and particularly along the northern coast of the GG using the oceanic and the atmospheric reanalysis data. The Bo is negative all the year excepted during June. It is characterized by a seasonal variability modulated by the seasonal variability of the rainfall and the net heat flux in the GG. In this region, the thermal and haline buoyancy fluxes play an important role and both should be used to determine precisely the Bo. The mixing layer in this region is shallower close to the northern coast of the GG because of the negative Bo; while offshore toward the warm pool it becomes much deeper because of the maximum values of the Bo tending to positive values. In this region, the Bo induces a stratification of the mixed layer and then reduces the mixing which is very important to sustain the cooling during the upwelling season. The Bo doesn’t enhance the vertical mixing and the phytoplankton bloom. With the increasing trend of the global warming, the Bo will contribute to reducing the coastal upwelling intensity along the northern GG.
Acknowledgements
The Atmospheric and the Oceanic reanalyses data are freely available through the Copernicus website. Authors thank the anonymous reviewers for their helpful comments.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.