Analysis of sea surface temperature pattern, variability and their teleconnection with rainfall dynamics over the Gulf of Guinea

Abstract

Spatial pattern and variability of sea surface temperature SST and their teleconnections with rainfall dynamics in the Gulf of Guinea (GOG) were examined in this study. SST and rainfall data of 50 years (1970–2022) were obtained from ERA5 and NOAA CPC at 0.25° x × 0.25° and 0.5° × 0.5° spatial resolution from longitudes 10°W and 8°W, 6^oW, 4^oW, and 2^oW and latitudes 15°N, 5^oN, 3^oN, 15^oS, 5^oS, and 3^oS, distributed along the Gulf of Guinea (GOG) respectively. Analysis was further carried out on twelve rainfall gridded stations distributed along the Gulf of Guinea (GoG) for the characterization of the Rainfall-SST teleconnection across the region while the relationship between the rainfall-SST anomalies, seasonal, inter-annual, and decadal scales was carried out using correlation analyses and composites. Interpolation of the meteorological variables was carried out using Inverse Distance Weight (IDW) from the ArcGIS Spatial Analyst Tool, Ferret, and CDO which were further employed to generate the seasonal and decadal rainfall and SST maps and statistical analysis of the study area. The result of the decadal and seasonal analysis of SST variability from 1970–1980,1980–1990, 1990–2000,2000–2010,2010–2020 and 2022 indicate that SST was highest from 2010 to 2020 at 28.91 °C and fluctuated between (28.49 °C) in the 1970–1980 and (28.08 °C) for the 1980–1990 decade While the seasonal pattern of SST showed marked variability with March–April and May(MAM) recording 29.34 °C with the lowest being in June-July–August(JJA) at 28.7 °C. In terms of decadal analysis of rainfall, the period 2010–2020 recorded the highest amount of rainfall along the coast (3,145.5 mm-3,928.3 mm while 1970–1980 recorded the lowest amount of rainfall (2,650–3.310 mm. To investigate the teleconnection between of SST and rainfall dynamics, statistical analysis was used where the SST values were plotted against seasonal rainfall in 11 stations namely Abidjan, Banjul, Accra, Guinea, Conakry, Cotonou, Dakar Doula, Freetown, Lagos, Lome, and Monrovia. The outcome of the statistical analysis and Standardized Anomaly Index used indicate that Banjul, Cotonou, Dakar, and Doula exhibited statistically insignificant correlation at 0.05 confidence level while Abidjan, Accra, Lagos, Lome, Freetown, and Monrovia showed positive and statistically significant correlation. The spatial pattern of seasonal rainfall climatology categorized into DJF, MAM, JJA, and SON reveals that JJA and SON produced 80% of rainfall in the Coastal GOG followed by MAM. The study affirmed that warm and cold tongues exist in the GOG alongside positive teleconnection and that the spatial variability of SST observed in this study corresponds positively with the decadal and seasonal variability of rainfall.

1 Introduction

The ocean plays an important role in the earth’s climate system owing to their large heat storage capacity with approximately 3.5 m of water containing as much energy as the entire atmosphere column (Deser et al. 2010). Similarly, more than 70% of the entire Earth's surface is covered by the ocean and therefore exerts an enormous influence on boundary layer atmospheric processes (Siedler et al. 2013). Ocean circulation therefore has profound impacts on the mean state and variability of the climate system. This is evidenced in the meridional transport of heat to the poles from the equator and its loss to the atmosphere moderates the climate, particularly in the mid to high latitude. In addition, changes in equatorial upwelling and currents play a critical role in driving El Nino and La Nina phenomena and thus wield significant influence on global climate from inter-annual to decadal timescale as well as modulating the intensity of human-induced climate change (Kosaka and Xue 2016).

The mechanism that accounts for the observed high and increasing SST anomalies emanates from the excessive solar radiation and heat that dominates the tropics. The surplus heat available in the tropics is then transported to higher latitudes through atmospheric and oceanic circulations which allows for interaction between the ocean and the atmosphere through deep atmospheric convection (Lubbecke et al. 2018). It is the coupled ocean–atmosphere interaction that influences the atmospheric wind field which in turn, strongly impacts sea surface temperature (SST) in the tropics, especially in the eastern equatorial Atlantic and Pacific, where it controls both the upwelling of cold water from below the sharp, shallow thermocline as well as the depth of the thermocline. SST anomalies in the tropics can thus have far-reaching effects on the climate system.

Scientist over the years has focused their attention on sea surface temperature anomalies in their effort to unravel the driving mechanism behind the observed change and variability of the West African precipitation system characterized by extreme rainfall with implications on severe floods and droughts in the region. (Vizy and Cook 2001, Paeth and Stuck 2004, Rodwell 2013 and Meynadier et al. 2016). This is premised on already established scientific reports that there exists a statistically significant relationship between sub-Saharan rainfall variability and global sea surface temperature pattern and distribution. The pioneering work of Lamb (1978) who was able to successfully link Saharan rainfall deficiency in 1960 to warn SST anomalies over the equatorial Atlantic and periods of normal and heavy rainfall to cold SSTA and vice versa attracted the attention of another scientist with keen interest in the African climate system. Lübbecke and McPhaden (2017) following previous investigations further asserts that the fluctuations in interannual and seasonal SST influence the winds and precipitation regime over South America, the Indian Ocean, the Sahel and even Europe as changes in ocean temperature and upwelling of nutrients have the propensity to affect the marine ecosystem. Lamb and Peppler (1992) in another study revealed a distinctive SSTA pattern over the Atlantic in comparison with the three driest years in Sub-Saharan Africa notably 1972,1977 and 1984 since 19840 to rainfall climatology unfortunately the 1983 drought broke the pattern. The high SST anomalies of the Atlantic Ocean have been described as an “El Niño like” phenomenon in the eastern equatorial Atlantic and the Gulf of Guinea as the “El Merle (1980) and Hisard (1980). Other studies that have been carried out with respect to SST anomalies of the East Mode and rainfall variability over West Africa include (Vizy and Cook 2001, Paeth and Stuck 2004, Rodwell 2013 and Meynadier et al. 2016) in their various research findings linked were the historical droughts that characterized West Africa and recent flood frequencies to SST anomalies. Some of the notable historical droughts with devastating impact in West Africa are shown in Table 1 below (Grist and Nicholson 2001; Odekunle and Eludoyin 2008).

Table 1 Historical droughts and flood years over West Africa

Given the scale and intensification of extreme climate change impacts which are already manifesting in Africa owing to her vulnerability to human induce global warming and climate change and low adaptive capacity, West Africa the most populated regions in the continent of Africa and the world are projected to experience a reversal of their developmental gains that will result in loss of livelihoods and infrastructure damages (UNFCCC 2021). Furthermore, the coastal regions of the GOG have seen an upsurge in rainfall in recent years culminating in devastating floods with huge loss impact on critical economic systems and human lives. The death toll and infrastructure loss triggered by flooding indicate high level of vulnerability of the coastal regions of West Africa due mainly to extreme rainfall and ocean–atmosphere interaction (The World Bank 2020).

Arising from above, it has become imperative to investigate the influence of local SST anomalies of the gulf of Gulf of Guinea on rainfall dynamics over West Africa. This is aptly important as rainfall variability is a major meteorological parameter that is critical to human life and socio-economic systems in West Africa. Most of the region’s population is largely dependent on rain-fed agriculture and pastures for livestock and the water resources sector, thus making them highly vulnerable to rainfall fluctuations. Consequently, it has become pertinent to understand the pattern and variability of SST anomalies on rainfall dynamics over the Gulf of Guinea and this is the relevance of the study.

Evidence in the literature affirms that previous studies conducted on sea surface temperature and their influence on West African rainfall include Ginanni (2010) on the relationship between anomalies of sea surface temperature (SST), perturbations in the atmospheric circulation, and response in precipitation forms the physical basis of seasonal to interannual prediction where the focus was on large scale SST variability. Other scholars include Grist and Nicholson 2000 on the dynamic factors influencing rainfall variability in the Sahel in which soil moisture, wind, temperature, and other instability mechanisms a. The outcome of their study confirms that the weaker African easterly jet (AEJ) and stronger tropical easterly jet (TEJ) and their latitudinal location rather than intensity during wet years are some of the dynamic factors driving rainfall in the Sahel. Vizy and Cook 2001, Vizy and Cook 2001 focused their studies on the mechanism by which the Gulf of Guinea and Eastern North Atlantic Sea surface temperature anomalies influence African rainfall by employing GCM model simulations. The study like previous ones was on large-scale atmospheric circulation and moisture anomalies and its association with precipitation in the Gulf of Guinea and the Congo Basin. Deser et al. 2010 also adopted a global outlook in their studies of large-scale atmospheric circulation perturbation and their influence on SST variability, patterns, and mechanisms in different on both the Northern and Southern oceans. Others include Rowell 2013, Akinbobola et al. 2015 Odekunle and Eludoyin 2008 and Wainwright 2019 have also investigated at various times the influence of tropical Atlantic SST on African rainfall variability. However, most of these studies failed to demonstrate the teleconnection of SSTA and rainfall variability in coastal cities along the Gulf of Guinea and this is the relevance of the study.

2 Materials and methods

The study used SST and rainfall data of 50 years (1970–2022) from ERA5 and NOAA CPC at 0.25⁰ × 0.25⁰ and 0.5⁰ × 0.5⁰ resolutions from specific locations namely longitudes 20°W and 10°E and latitudes 15°N and 2°N along the gulf of Guinea obtained at https://cds.climate.copernicus.eu/ and https://iridl.ideo.columbia.edu/sources/noaa as shown in Table 1 below. The inverse distance weighted (IDW) method of interpolation, ArcGIS spatial Analyst tool, Ferret, Panoply, and climate data operator CDO were used in the study to produce the spatial patterns of SST and rainfall variability at decadal and seasonal timescale covering decade 1 (1970–1980), Decade 2 (1980–1990), Decade 3 (1990–2000), Decade 4 (2000–2010), Decade 5 (Decade (2010–2020), and the year 2022 for the characterization of extreme events in West Africa within the study period (1970–2022) while December–February (DJF); March–May (MAM); June–August (JJA- summer), and September–November (SON- autumn) constitute the seasonal pattern for the Northern hemisphere while statistical and standardized anomaly index was used for the teleconnection analysis on twelve names namely Abidjan, Banjul, Accra, Guinea, Conakry, Cotonou, Dakar Doula, Freetown, Lagos, Lome, and Monrovia which are points within the 0.25 × 0.25 grides of the datasets as the mid-point of the grid were extracted to form the said twelve points as shown in Fig. 1b. Secondly, the choice of the twelve points was based on their coastal locations within West Africa at the intersection of the land–ocean interface which forms the focus of this study as shown in Fig. 1b while Table 2 highlights the Rainfall characteristics of all the sampled points. Consequently, statistical tools involving means and averages were used to generate monthly, annual, and inter-seasonal rainfall and SST for the study period. The choice of the dataset is based on the spatial characteristics and resolution of ERA5 Copernicus data which has been widely used for rainfall studies in Africa. In addition, both ERA5 and NOAA CPC data have high reliability and are up-to-date for hourly, synoptic, daily, and monthly data fractions for trend and ocean/weather monitoring and forecast particularly for a data-sparse region like Africa (Table 3).

Fig. 1

a) Map of the Gulf of the Guinea showing study Area

Table 2 Historical datasets used in the study

Table 3 Rainfall characteristics of sampled stations

2.1 Standardized rainfall and SST anomaly

The Standardized Precipitation Index (SPI) developed by McKee et al (1993) was used to identify variations and fluctuations in the rainfall values, for regional climate change and for changes in rainfall distribution. The Short-term and long-term rainfall anomalies were plotted for annual and inter-seasonal variations. SP values with its corresponding categories as shown in Table 4 below give a better representation of abnormal wetness and dryness than the Palmer indices. Similarly, an SST value greater than + 0.5 °C or less than -0.5 °C was considered abnormal. Sea Surface Temperature anomalies are a commonly used index of the frequency and magnitude of El Niño and La Niña events. The SPI table as shown below in Table 1 contains both negative and positive values with the negative values as rainfall anomalies while the positive values indicate normal rainfall incidents. The formula for computing the SPI values is expressed in Eq. 1 below.

Table 4 The standardized precipitation index (McKee et al. 1993)

The SPI equation is stated as follows

$$\text{SPI}=\frac{x-{x}^{\prime}}{\varphi }$$

(1)

x is the annual mean rainfall,

x’ is the mean of the entire series.

φ is the standard deviation from the mean of the series.

Similarly, the Regression model was also used to examine the relationship between the two variables of interest (SST and Rainfall in all the stations considered in the study and establish a teleconnection. Linear Regression model can identify the cause and effect of one variable against the other. The formula for Linear Regression is given below:

$$\text{y}=\text{mX}+\text{b}$$

(2)

where Y = Dependent variable.

X:

Independent variable (Predictor).

M:

estimated slop.

b:

intercept.

2.2 The study area

The study area is the GOG region of the Atlantic Ocean (Fig. 1). The area lies between longitudes 10°W and 8°E and latitudes 5°N and 5°S, and thus, south of West Africa. The hydrography of the GOG is directly affected by five main currents, namely, Benguela, the South Equatorial, the Canary, the Counter Equatorial, and the Guinea Currents (Longhurst 1962) as shown below in Fig. 1a while the sample points namely Abidjan, Accra, Banjul, Cotonou, Conakry, Dakar, Doula Freetown, Lagos, Lome, and Monrovia are shown in Fig. 1b. The tropical marine (mT) air mass and the tropical continental (cT) air mass are the two main wind currents that have the most impact on the climate of GOG and all of West Africa. The Equatorial Easterlies, a third wind current, is only significant over the continent of West Africa (Ojo 1977; Iloeje 1981). The southern high-pressure band off the coast of Namibia is where the mT air mass starts. On its path, the air mass takes up moisture over the South Atlantic Ocean, crosses the Equator, and then penetrates GOG and West Africa. The high-pressure area to the north of the Tropic of Cancer is where the cT air mass is found. It picks up little moisture along its path and is thus dry. The two air masses (mT and cT) meet along a slanting surface called the Inter-tropical Discontinuity (ITD). The Equatorial Easterlies are rather erratic cool air masses, which come from the east and flow in the upper atmosphere along the ITD.

3 Results/discussion

The results of the spatiotemporal variability of SST and rainfall patterns are presented as follows.

It is well known that variations in sea surface temperatures (SSTs) are partly responsible for large anomalies of seasonal mean rainfall over many areas of Africa. Generally, as can be seen from Fig. 2 above. SST decreases westward and the northern parts with a significant increase in the south and eastern axis of the Gulf of Guinea. The 1970 -1980 (Fig. 2a), 1980–1990, and 1990–2020 periods (Fig. 2b and c) were characterized by below-normal SST mean values across the Gulf of Guinea while substantial increase in SST was observed from 2000–2010, 2010–2022 and 2022 respectively (Fig. 2d-f) The anomalously low SST observed in the periods corresponds well with the periods of different drought episodes that characterized West Africa in the early 1970 and 1980s. The result further reveals that the Guinea coast has experienced accelerated and fluctuating SST increase over the study period leading to intensification of rainfall anomalies. A reversal of this narrative was observed in the Western flank with anomalously cold SST leading to reduced rainfall anomalies and drought. The result established in the study are in agreement with extensive studies conducted by Odekunle and Eludoyin 2008; Akinbobola et al. 2015, Rowell 2013; Vizy and Cook 2001 and Giannini 2016 on the contribution of sea surface anomalies on West African climate.

Fig. 2

Decadal pattern of SST variability: a) Climatological 1970–1980 b) 1980–1990 SST c) 1990–2000 SST d) 2000–2010 SST e) 2010–2020 SST and f) climatological 2022 SST

Furthermore, the pattern of decadal variation observed in this study confirms the influence of El-Nino on the warming of the tropical Atlantic Ocean and the continuous rise of SST of the GOG. An ENSO episode manifests through the appearance of SST anomalies, and this was evident in the analysis performed in the study.

3.1 Seasonal pattern of SST variability

Seasonal pattern of SST variability was also conducted in the study and the result of the geospatial analysis as shown in Fig. 3 below revealed marked variability with June-July and August (JJA, Fig. 3b) recording the lowest (28.18 °C for the Guinean coast and 24.70 °C in the western Sahel) followed by 28.71 °C for the Guinean coast in fall months of December, January and February DJF (Fig. 3a) and 20.85 °C for the Western Sahel. Conversely, the spring months of March, April, and May MAM recorded the highest as shown in (Fig. 3c) given the variability values of 29.34 °C for the Guinean coast and 20.08 °C in the Sahel while the variability values for the fall months of September, October, and November SON exhibited 28.74 °C for the Guinean coast and 25.62^OC for the low rainfall region (Fig. 3d). Generally, the result of the seasonal pattern indicates that SST in the Guinean Coast was generally low in June, July, and August, JJA as shown in Figure 3b followed by September, October, and November (SON) in Figure 3d. Similarly, March. April and May MAM  and December, January, and February (DJF) were characterized by high SST values as presented in Figure 3a and 3c. The anomalously cold SST observed in June July August, JJA and September, October November along the Guinea coast is attributed to the influence of coastal upwelling of cold water from below the sharp, shallow thermocline. The upwelling of nutrients has been reported to impact marine ecosystems negatively (Odekunle and Eludoyin 2008, Lubbeck et al. 2018).

Fig. 3

Seasonal pattern of SST Variability: Climatological a) December-January- February SST b) June-July August SST c) March–April-May SST and d) September–October-November SST

3.2 Rainfall variability

Precipitation variability is one of the most important climate variables that impacts human activities, agriculture, and the environment in West Africa, particularly seasonal rainfall. Therefore, assessing rainfall patterns and variability is critical in managing climate risk and improving resilience.

The analysis of the decadal variability of rainfall was carried out in this study to ascertain the degree of interannual variability of rainfall in the Gulf of Guinea. Although the period of study is from 1970–2022, the analysis was divided into 4 phases namely 1970–1980, 1980–1990,1990–2000, 2010–2020, and 2022 at ten years intervals as shown in Fig. 4 below.

Fig. 4

Decadal pattern of rainfall variability: a-1970-1980; b-1980-1990; c-1990-2000; d-2000-2010; e-2010-2020; f-2022

The results of the geospatial analysis presented above provide an insight into the areas of low and high precipitation concentration with the Western Sahel characterized as a region of low and erratic rainfall. The variability values in the 1970–1980 decade as presented in Fig. 4a are shown in Table 2 below.

Rainfall distribution in five different decades of the Gulf of Guinea is shown in Table 2 with the 1970–1980 (Fig. 4a) recording the lowest rainfall concentration based on the variability values exhibited. There was however an upward trend in the 1980–1990 phase (Fig. 4b) as the variability values improved. In contrast, a reduction in the closing values of 1980–1990 was observed compared to the 1970–1980 era implying low rainfall in the rainfall-abundant regions of the study. 1990–2000 (Fig. 4c) was characterized by a remarkable improvement in rainfall as the variability increased from 21.4–691.4 to 2.071.5–3.371.4 mm. In comparison with the 1990–2000 decadal analysis, the 2000–2010 epoch (Fig. 4d) can be described as a decade of reduced rainfall as the variability values decreased from 21.4–691.4 to 13.9–687.2 mm. The result further indicates that the Western Sahel and the Guinean coast recorded significant improvement in rainfall in 2010–2020 (Fig. 4e) given the fact that the variability values were which is the highest in all the five 5 decades considered in the study (Table 5).

Table 5 Decadal rainfall variability values(mm)

Arising from above, it’s obvious that the GOG decadal rainfall pattern was characterized by a high degree of fluctuation and variability with 1970–1980 considered as the decade with the lowest rainfall concentration followed by the 1980–1990 phase. Finally, a cursory look at the 2022 analysis as presented in Fig. 4f reveals rainfall reduction in the dry regions as the mean annual contrasted from 21.7- 695.4 in 2010–2020 to 17.4- 659.8 mm in 2022. The rainfall-abundant regions also experienced a drastic reduction as maximum rainfall decreased from 2,7166–3.390.1 to 2557.2–3.229.6 mm in 2022.

The low rainfall variability values shown in the decadal analysis are indicative of drought vulnerability that dominated West Africa. There was however rapid improvement and recovery in rainfall in 1990–2000 while 2000–2010 witnessed reduced rainfall. The 2010–2020 was characterized by increased rainfall both ion the Sahel and the Guinean coast followed by the year 2022. The increased rainfall observed in 2010–2020 and 2022 are indicative of wet years and flood risk in the study domain.

3.3 Seasonal pattern of rainfall variability

In many hydro-climatological investigations, precipitation is a crucial meteorological parameter that determines regional and global climate change risks (Soltani et al. 2016). Evidently, precipitation is very critical in the land surface-water balance studies as it replenishes soil moisture, supports vegetation growth and, given its role in replenishing soil moisture, supports vegetation growth, and in keeping surface flows and groundwater resources and serving as a link between atmospheric and land surface processes (Alimonti et al. 2017; Sarkar et al. 2021 and Abed-Elmdoust 2016). Therefore, seasonal pattern of rainfall variability assessment and distribution is imperative in mitigating against the impact of climate change and improving resilience to extreme weather events. The benefits of analyzing seasonal rainfall variability cannot be overemphasized given its role in climate risk management in the agricultural and water resources sectors.

The outcome of the geospatial analysis conducted on seasonal patterns of rainfall variability as shown in Fig. 5 above indicates almost total dryness along the West African Sahel given the variability values of 0–23.5 for DJF in Fig. 5a. In the coastal regions, rainfall ranged from 47.2 to 70.6 to 117.7 mm. Although this is the peak of the dry season, the coastal regions experience rainfall at any time of the year. March April and May MAM (Fig. 5b) represents the commencement of the rainy season in most parts of West Africa which accounts for the high variability values from 0.4–50.2 mm to 199.6–249.2 mm in the coastal regions.

Fig. 5

Seasonal pattern of rainfall variability over the Gulf of Guinea: a: DJF; b: MAM; c: JJA; d: SON

The overall improvement of rainfall in the JJA (Fig. 5c) as demonstrated in Table 6 is directly associated with the peak of the rainy season in West Africa. JJA is also characterized by increased flash floods in many regions in West Africa as shown by the high variability values in Table 6 above. However, rainfall continues to improve in the coastal regions particularly Nigeria, Sierre Leon, Liberia, and Cameron, with 3.1.4–392.8 mm as maximum rainfall for the months of September–October and November (SON, as shown in (Fig. 5d) with September as the peak of the rainy season, the reversal was what played out in the Western Sahel, further reduction in rainfall was experienced given the fact that the variability value of 0.9-151 mm recorded previously in JJA further decreased to 0.4–78.9 mm. In West Africa, SON is normally the transition from the wet to the dry season.

Table 6 Seasonal pattern rainfall variability values(mm)

3.4 Statistical analysis of sst-rainfall teleconnection in West Africa

Statistical analysis of annual sea surface SST and rainfall teleconnection of the sample points namely Abidjan, Accra, Banjul, Conakry, Cotonou, Dakar, Doula, Monrovia, Freetown, Lome, and Lagos located along the Gulf of Guinea was carried out using standardized SST anomalies and Regression analysis for the study period 1970–2022 starting with Abidjan in that order. It is important to emphasize that the ERA5 dataset was used for the teleconnection analysis.

The standardized SST and rainfall anomaly and Regression analysis for Abidjan as shown below in Fig. 6 indicates that the early 1970s and 1980 had negative SST values indicative of anomalously low SST with implication on rainfall while positive SST observed late in 2000 signifies warm tongue which could also positively influence rainfall in the station. The model summary result of the regression analysis was however positive with a coefficient of determination of R² = 0.0625. The standardized rainfall anomaly agrees with the trend analysis as negative SST was recorded in the late 1970s while positive SST values were observed from 2020–2022 indicative of increasing rainfall. The analysis of the monthly SST climatology further revealed that SST increased in the winter and spring months of January, February March, April, and May while a downward and anomalously low was recorded in the summer months of June, July August, and September which lends credence to the decadal and seasonal SST pattern earlier presented.

Fig. 6

Regression and Standardized Anomaly Index of SST-Rainfall teleconnection in for Abidjan Cote’d Ivoire

In Accra Ghana, the trend analysis indicates negative SST in 1976–1979 with a corresponding impact on rainfall during the period. Similarly, rainfall anomaly was recorded in 1997 implying anomalous SST. Furthermore, the positive rainfall trend that characterized the station from early 2000 to 2022 represents an increased rainfall trend. The summary result of the regression model indicates a statistically significant relation given the coefficient of determination R² = 0.1345 as (P < 0.05) as shown in Fig. 7d above. The monthly climatology plot in Fig. 7e shows anomalously low SST in the summer months of June, July, and August.

Fig. 7

Regression and Standardized Anomaly Index of SST-Rainfall teleconnection for Accra Ghana

The trend analysis for Cotonou in Benin as shown in Fig. 8a above reveals a significant decrease in both SST and rainfall in 1970 and 1980 decade which indicates the influence of SST on rainfall in the Guinean coast. However, an upward SST trend was observed in early 2000 before decreasing sharply again in 2014 and 2016. The summary result of the regression analysis in Fig. 7d indicates a weak and statistically insignificant given the coefficient of determination R² = 0.0151 (P < 0.05). The monthly climatology in Fig. 7e shows again anomalously low SST in June, July, August, and September.

Fig. 8

Regression analysis and Standardized Anomaly index of SST-Rainfall Teleconnection over Banjul

Marked temporal annual variability continues for both SST and rainfall as shown in Fig. 9. SST was all time low in 1972 -1974 before accelerating in 1984 with a corresponding increase from early 2020 as shown in (Figs. 9b and c). Rainfall fluctuates all through the period recording its lowest in 1984 and from 2002–2009 before showing signs of recovery in 2001. The regression result indicates a weak correlation coefficient of (0.00266) in Fig. 9e which is not statistically significant at a 0.05 confidence level (p < 0.05). The monthly climatology in Fig. 7e shows again anomalously low SST in June, July, August, and September.

Fig. 9

SST and Rainfall trend analysis Banjul Gambia

In Fig. 10 above, significant variation in both SST and rainfall was observed all the period of study in Dakar. The result further indicates a rapid downward trend in rainfall. As shown in Fig. 10a. The standardized SST and rainfall anomaly computed for Dakar as shown in Figs. 10b and c indicates that the early 1970s and 1980 were characterized by SST anomalies indicative of decreased SST and erratic rainfall and dryness while positive SST was observed late in 2000 while the positive values are considered upward trend of rainfall. The model summary result of the regression analysis in Fig. 10d revealed a weak and insignificant correlation of 0.006 at a 0.05 confidence level (p < 0.05). This implies that other variables outside the regression model could be responsible for the observed variability in rainfall in the SST. The monthly SST climatology showed anomalously low SST in the summer months of June, July, and September thus agreeing with the SST and rainfall seasonal analysis as shown in Fig. 10e above.

Fig. 10

SST and rainfall trend and anomaly for Dakar

The trend analysis for Monrovia in Liberia reveals decreasing SST in 1976, 1994, 1990,1992, 2008–2014 while the increasing trend was recorded in 1972 and 1974, 2010, 2020, and 2022 as shown in Fig. 11b and c below. Rainfall anomaly recorded in 2012–2014 which corresponds to anomalous SST. Again, the analysis showed marked interannual variability in both meteorological variables. The standardized SST and rainfall analysis indicate negative SST and rainfall in 1972 and 2008 implying anomalously dry years. However positive SST was recorded for the period 2018–2022. The summary result of the regression analysis as shown in Fig. 11d indicates a coefficient of determination R² = 0.092 was statistically significant at 0.05 significance level which we can infer that SSTA can be used as an explanatory variable in predicting rainfall in the station while the monthly climatology of the mean SST shows that SST was anomalously low in July August and December as captured in Fig. 11e

Fig. 11

Regression and Standardized Anomaly Index analysis of SST-Rainfall teleconnection SST-Rainfall for Monrovia

The trend analysis for Conakry in Guinea reveals decreasing SST in the 1970s and 1990s while the upward trend was observed in early 2000 and late 2000, The station exhibited the same trend observed in other cities along the West African coast considered in this study characterized by abnormally low SST in the 1970s and 80 which correspond well with declining rainfall as shown in Fig. 12a. The SST and rainfall anomaly indicates that 1974–1976 and later 1994 were periods of decreasing rainfall and dryness as shown in Fig. 12b and c, although increasing trend of SST and rainfall was observed from 2018–2022. The summary result of the regression analysis indicates a coefficient of determination R² = 0.007 which is insignificant at 0.05 as shown in Fig. 12d. The monthly climatology plot indicates anomalously low SST in January February and March before peaking up in April, May, June July. Anomalously low SST was observe in August and September while November and December were associated with an increase in SST as shown in Fig. 12e.

Fig. 12

Regression and Standardized Anomaly Index analysis of SST-Rainfall teleconnection SST and Rainfall for Analysis Conakry Guinea

The SST and rainfall trend analysis for Doula Cameron as shown in Fig. 13a above reveals spatial variation for both SST and rainfall. The standardized rainfall anomaly indicates that 1974 and 1976 were years in which SST was negative as shown in Fig. 13b and c. However, 2002 -2006 were also periods of positive SST while abnormal rainfall was also recorded in 2018. The regression model shows a downward and negative slop with a coefficient of determination R² = 0.08 as shown in Fig. 13d which is statistically insignificant at 0.05 confidence level while the monthly climatology shows downward slop in the months of June, July, August, and September as captured in Fig. 13e

Fig. 13

Regression and Standardized Anomaly Index analysis of SST-Rainfall Teleconnection for Doula, Cameron

The result of the SST rainfall teleconnection for Freetown Sierre Leon as shown in Fig. 14a revealed spatial variability with anomalously dry conditions from 1993 to 1994 as rainfall was negative (Fig. 14b and c). Increasing SST was also observed from 2003 to 2012 with corresponding upward trend in rainfall. The summary regression model shows a positive correlation with an upward slop of 0.0683 as coefficient of determination which is statistically significant at 0.05 confidence level as captured in Fig. 14d while the monthly climatology of shows anomalously low SST in the months of February with a peak in March, April May. The summer months June July and August were characterized by anomalously low SST while the fall months of October, November and December were associated with high SST.

Fig. 14

Regression and Standardized Anomaly Index o analysis of SST-Rainfall teleconnection SST-Rainfall for Freetown

The statistical analysis of rainfall and SST teleconnection conducted for Lagos Nigeria as shown in Fig. 15a above exhibited the same pattern and trend of SST and rainfall like other coastal cities in West Africa. SST was anomalously low as captured in Fig. 15b and c in 1974–1976 and later assumed an upward trend from 2004–2022 implying periods of increased wet years. The summary regression statistics was positive with a coefficient of determination of R² = 0.0848 at 0.05 confidence level indicating a statistically significant relationship and teleconnection as shown in Fig. 15d while the monthly climatology in Fig. 15e shows that SST was anomalously low in the months of June, July, August, and September corresponding well with the decadal and seasonal SST pattern earlier explained. This explains why SST can be used as explanatory variable to predict rainfall variability in the summer months of June, July, August and September in Lagos Nigeria and across West Africa.

Fig. 15

Regression and Standardized Anomaly Index of SST-Rainfall Teleconnection over Lagos Nigeria

The result of the trend analysis for Lome as presented in Fig. 16 above indicates that 1974–1978 were anomalously dry years as SST was negative. Also, abnormal rainfall was observed from 1996–1998 implying increased drought episodes. However, there was an upward SST and a progressing SST trend from 2018–2022. The summary regression output showed a positive correlation at insignificant at r = 0.0043. The monthly climatology reveals anomalously low SST in June, July, August, and September with peaks in March, April, May, October, November, and December.

Fig. 16

Regression and Standardized Anomaly Index calculation of SST-Rainfall teleconnection a SST-Rainfall trend anal over

The composites of the SST anomaly, wet and dry years over the Gulf of Guinea is shown in the table below.

The table above summarizes wet and dry years over the Gulf of Guinea and the accompanying years of SST anomaly showing the influence of ENSO events which could be disastrous to the water resources and the agricultural sectors in the study domain. The Wet years as shown in Table 7 are years of El Nino which is associated with warmer SST anomaly over the gulf of gulf of Guinea and driven by ENSO in the central and Pacific. The resultant effect manifests in wetter years experienced in West Africa while La Nina conditions favor Cold SST anomalies resulting in drier and sort rains years as demonstrated in the various statistical analyses. The outcome of this analysis as shown in this study is consistent with previous studies conducted by (Akinsanola et al. 2018; Biao and Alamou 2018).

Table 7 Rainfall and SST years of anomaly

4 Discussion of findings

4.1 Decadal SST variability

It is well known that variations in sea surface temperatures (SSTs) are partly responsible for large anomalies of seasonal mean rainfall over many areas of Africa. Knowledge of these teleconnections is increasingly used to provide probabilistic predictions of such rainfall anomalies, to help vulnerable African communities, and assist national resource management (Lubbecke et al. 2018). The outcome of the decadal analysis of SST variability in the Gulf of Guinea was generally on the increase with 2010–2020 as the period with the highest warming followed by 2022 (28.91 °C) while the decade with the lowest SST was 1980–1990 followed by the decade of 1990–2000 phase. with consecutive increases.

Generally, SST decreases westward and the northern parts with a significant increase in the south and Eastern axis of the Gulf of Guinea. The 1970 -1980 and the 1990 periods are associated with low SST which corresponds well with the periods of different drought episodes that characterized West Africa in the early 1970s and 1980s. The dominant variability of SST is seasonal. Studies has shown that SST anomaly arises from several mechanisms at different timescales and that Inter-annual SSTA variability at a given location tends to be greater for some months of the year than for others, and also varies from place to place (Czaja 2004). The warming trend shown in this study is consistent with other studies made during the satellite era (1979–2012) as cited in the Intergovernmental Panel on Climate Change (IPCC 2022 special report Ocean. Seas and cryosphere).

The result further shows that coastal SST has increased rapidly over the study period leading to intensification of rainfall anomalies while cold SST dominates the Western part leading to reduced rainfall and drought. Extensive studies conducted by Odekunle and Eludoyin 2008; Akinbobola et al. 2015, Rodwell 2013; Vizy and Cook 2001 and Giannini 2016 affirmed the existence of interannual variability of SST and its attendant influence on the regional climate system of the Guinean Coast and the Sahel. Furthermore, the pattern of decadal variability observed in this study confirms the influence of El-Nino phenomenon on the tropical Atlantic Ocean and the continuous rise of SST of the GOG. The warming pattern demonstrated in this study An ENSO episode established in this study is a robust response to anthropogenic induced climate change with significant impact on the climate system.

4.2 Seasonal SST variability

The analysis of the seasonal pattern of SST variability conducted in this study indicates anomalously low SST (28.18 °C and 2o) for the summer months of June July and August, JJA 29.34^OC, 28.71 °C and 28.74^OC for March April May, MAM, December, January February DJF, and September, October November SON respectively for the Guinean Coast. In the Southwestern flank bordering the Sahel region, the variability values are 20.08^OC for MAM, 20.85^OC for DJF while JJA recorded 24.7 °C and SON 25.82^OC over the Guinean Coast in the months of July and August and highest in March, April and March MAM followed by December, January and February DJF and September, October, and November SON respectively. The anomalously cold SST observed in the months of June July August, JJA and September, October November along the Guinea coast is attributed to the influence of coastal upwelling. The coastal upwelling is associated to gyre circulation and the Guinean current in the North South equatorial current in the south. What is not clear however is the mechanism that drives the cold water which is the focus of ongoing scientific investigation as previous studies are yet to fully resolve teleconnection between local winds and upwelling effect (Odekunle and Eludoyin 2008; Rowell 2013).

The observed SST pattern demonstrated in this study indicates that the anomalously cold SST emanates from the advection of cold upwelled water that flows from the south equatorial current and the northward extension of the cold Benguela current that develops into an equatorial current during the summer leading to a shoaling of the Thermocline near the northern coast of the GOG and is what triggers the upwelling effect. The outcome of the decadal and seasonal SST anomalies in this study agrees with the explanation espoused above as causative factor responsible for the anomalously cold SST observed in the JJA and SON. The observed seasonal pattern of SST shown in this study corroborates earlier studies conducted by the Mercator operational oceanographer in 2006 cited in Odekunle and Eludoyin 2008, Rowell 2013 and Vizy and Cook in 2001, Dippe et al. (2018) reports that the cold tongue in the summer months of June, July and August manifest due to the northward migration of the inter tropical convergence Zone ITCZ and its interaction with the south trade winds which in turn leads to the shaolin of the thermocline. It is this process that gives rise to the development of upwelling phenomenon that persist up to September.

Furthermore, the pattern of the decadal and seasonal SST variability observed in the study lends credence to the much-publicized anthropogenic induced global warming as stated in the (IPCC 2022 WG II 4R6 report). The study revealed that between 1970–2022, SST increased by an average of 0.5° with 2010–2020 as the decade with the highest SST variability values.

4.3 Decadal pattern of rainfall variability

Precipitation variability is an important climate factor that impacts human activities, agriculture, and the environment in West Africa. The outcome of the decadal rainfall variability from 1970–1980 shows a maximum of 2.707.4–3.81 mm with Nigeria, Cameron, Guinea, Sierre-Leon, and Liberia as areas of major concentration. Area with latitude 20^oN constitute the low rainfall region. In the second decade, 1980–1990, was characterized by a period of high rainfall variability as significant reduction in rainfall was observed declining from 2.7074–3.381.8 mm in the 1970–1980 phase to 1.906.6–2.538.8 mm in the 1980–1990 phase. The pattern of rainfall variation observed in the 1980–1990 era synchronized well with the decadal pattern of SST in 1980–1990. In the discussion on SST pattern, mention was made of anomalously low SST between 1970–1980 and 1980–1990. Remarkably, the anomalously low SST reported for the first and second decade correspond well with the rainfall variability of the 1980 decade thus providing further insight on the SST rainfall teleconnection over the Gulf of Guinea which is the motivation for the study.

The pattern of SST-rainfall teleconnection revealed in this study supports previous studies by Rodwell 2013; Vizy and Cook 2001; Vizy and Cook 2001; Nicholson 2013; Akinbobola 2015 and Meynadler et al. 2016 on recent air–sea coupling of West Africa Precipitation.

Rainfall variability for the period 1990–2000 witnessed remarkable improvement in both the Sahel and the Guinean coast. The variability values increased from a minimum of 21.4–691.4 mm to 2.7101.5–3.371.4 mm with areas of high concentration been around Sierre Leone, Liberia, Nigeria and the borders with Equatorial Guinea. The Western Sahel from Senegal experienced rainfall recovery as the variability values increased from 9.5–641.6 mm in the preceding phase (1980–1990). The improvement in rainfall recorded this period is in response to accelerated ocean warming evidenced by SST over the tropical Atlantic during the period of 1990–2000.

The 2000–2010 decade was also characterized by intense rainfall in the Guinean coast as the variability values increased to 2.707–3.380.5 mm. The high rainfall recorded this period could be attributed to increased SST, topography, and the West Africa Monsoon WAM and the position of the ITCZ as previously reported by (Akinsanola and Zhou 2019, Grist and Nicholson 2001). Conversely, the regions around the Western Sahel experienced rainfall reduction with values ranging from 13.0–687.2 mm. The observed decline in rainfall in the Western Sahel could be explained by the weaker easterly jet (TEJ) and weaker low level southerly monsoon flow, equatorial Westerlies, and the position of the ITCZ that significantly modulates the inter annual rainfall variability in West Africa as stated by  (Nicholson 2013).

Therefore, the increased precipitation variability reproduced in this study can trigger reduce agricultural yields as shown by (Rowhani et al. 2011). Also, precipitation variability as demonstrated in this study is strongly connected to extreme wet and dry events, floods, and droughts, which pose great threats to the environment and society, and can also have devastating consequences on ecosystems, food supplies, and economies at the local and regional which is in agreement with (Easterling et al. 2000).

The analysis revealed that the period 2010–2020 was also characterized by increased rainfall as the variability values increased from 21.7–695.5 along latitude 12^o (Western Sahel) to 695.5–1.369.1–2.042.6. 2.7166.5–3.390.1 in the Guinean coast. This represents a significant increase compared to the preceding decade of 200–2010, However, the year 2022 experienced rainfall reduction as the variability values of 2.587.2–3.220.6 were lower as against 2010–2020. The decline observed in 2022 corresponds to lower SST in 2022. The pattern of rainfall variability could be linked top internal variability of the Atlantic Multidecadal Variability (AMV), previously known as the Atlantic Multidecadal Oscillation (AMO) (Zhang and Delworth 2006) The AMV is a basin-wide SST variability pattern with its strongest signal located in the North Atlantic. The pattern of decadal SST analysis produced in this study agrees with empirical studies conducted by (Bertnttelle 2023, Ting et al. 2011, Williams et al. 2022, and Zhang et al. 2021) that AMO exerts a significant influence on the climate of adjoining continents including rainfall over West Africa.

4.4 Seasonal pattern of rainfall variability

The seasonal pattern of rainfall variability from 1970–2022 was categorized into December-January–February DJF, March–April-May MAM, June-July and August JJA and September, October and November SON.

Findings from the spatial analysis indicate high seasonal variability with September–October and November SON recoding high rainfall intensity as the variability values increase from 0.4–76.9 mm in the Western Sahel 79–157.4 mm, 157.5–235.8 and 235.9–314.3 mm been the maximum. This further affirms that the rainy season in West Africa peaks up in September, October, and November SON, followed by June-July and August JJA and March, April, and May MAM respectively. Conversely, the seasonal distribution of rainfall in the Western Sahel was poor, particularly in the months of SON which could be devastating to crops and pasture development, particularly in Northern parts of Senegal, Southern Mauritania, Central and Northern Mali, Northern Nigeria, Niger and the lake Chad Basin. This finding agrees with the World Food Program (WFP 2021).

Significant rainfall deficits were also observed in JJA across the coastal Guinea regions of Ghana, eastern Cameron, Northern Nigeria, Burkina Faso, Togo, and Benin. The phenomenon of reduced rainfall in the months of JJA is called the little Dry season LDS. The Little Dry Season LDS associated with anomalously low SST in the months of JJA and the southward migration of the tropical rain belt ITCZ all contributed to the high variability of rainfall during this season (Adejuwon and Odekunle 2006, Odekunle and Eludoyin 2008, Chineke et al. 2020, Paker and Diop-Kane 2017 and Wainwight et al. 2019. The length and severity of the LDS could be detrimental to crop productivity and pasture which are the major source of livelihoods in West Africa. The seasonal variability of rainfall further indicates that JJA and SON account for 80% of total rainfall in West Africa which is attributed to WASM, a critical aspect of the West Africa Monsoon (Akintomide and Zhou 2019).

The anomalously low SST also increases static stability over the Gulf of Guinea region leading to the suppression of convective-producing rainfall. Findings from the decadal and seasonal distribution of rainfall reveal that mean annual rainfall in West Africa was generally zonal with rainfall amount decreasing from the coast to the inland regions. The results of the analysis performed here are consistent with previous studies that rainfall intensity is dominant along the Guinea coast in West Africa (Nicholson 2003, Afiesimama et al. 2006, Gbobaniyi et al. 2013 and Akinsanola et al. 2006). Palmer et al. 2023 was emphatic of the role of seasonal rainfall to the health and economic prosperity of the West Africa sub-region and such disruptions to the rainfall cycle could result in wide range economic and environmental impacts.

Furthermore, the results of the analysis on spatial variability of SST and rainfall anomalies demonstrated in this study lend s credence to previous studies that SST is believed to be strongly influenced by the interannual variability of the West African Moonson (Caminade and Terray 2010; Folland et al. 1986; Losada et al. 2001; Palmer 1986; Rodríguez-Fonseca et al. 2011; Zhang and Delworth 2006. Additional observational data carried out on the SST pattern of the South Atlantic Ocean indicates that the SST anomalies in the tropical Atlantic are associated with the ENSO phenomenon which reinforces previous studies by Nnamchi and Li 2011, Cheneke et al 2020, Rohli and Vega 2017.

4.5 Rainfall and SST anomalies

The outcome of both rainfall and SST regression analysis and the Standardized anomalies conducted in the study revealed fluctuating and progressive increases in both SST and rainfall in the study domain from 1970–2022. This agrees with previous studies that SST fluctuation is more pronounced in the coastal regions and arises from dynamic features such as Kelvin waves and Tropical Instability Waves (TIWs) in the equatorial Pacific Ocean. (Bulgin, et al. 2020). The regression analysis further indicates a weak and insignificant coefficient of relationship for Banjul, Conakry, Cotonou, Dakar, and Doula while the positive and significant relationship was established in Abidjan, Accra, Lome, Freetown Monrovia, and Lagos. The weak correlation exhibited by some of the coastal stations demonstrates the fact that precipitation in the coastal regions are also influenced by other processes such as coastal topography, thunderstorms, as well as prevailing wind and tropospheric stabilitywhich corroborates Akinbobola et al. 2015.

The standardized rainfall and SST anomaly carried out reveals composite wet and dry years with anomalously low SST years more pronounced in the 1970 and early 1980 across all the stations. The SST anomaly captured in this study is in direct response to various ENSO episodes as evidenced by years of below and abnormal SST and rainfall variability. Similarly, plots of the SST monthly climatology revealed that SST was anomalously low in the summer months of June, July, August, and September before peaking up in the fall months of October, November and December as this was evident across all the stations except Conakry. The SST monthly variability is greatest in summer than winter which is associated when the marked layer of the ocean significantly responds to its internal forcing. This is partly linked to the El Nino Southern Oscillation (ENSO), which is the largest mode of SSTA variability involving remote teleconnections mediated by ENSO influences on atmospheric circulation (Deser et al. 2010).

5 Conclusion

The study has examined sea surface temperature spatial pattern, variability, and the teleconnection with rainfall dynamics over the Gulf of Guinea from 1970–2022. In achieving this, both NOAA CPC and ERA5 data sets were used for analysis of decadal and seasonal variability for both SST and rainfall using ten year moving average for both SST and rainfall while three-month interval was used for the seasonal analysis. To establish the teleconnection of SST response to rainfall, statistical analysis were employed for trend, anomaly and linear regression for Abidjan, Accra, Cotonou, Dakar, Doula, Banjul, Lome, Freetown, Conakry, Lagos and Monrovia with results showing increasing trend, weak and insignificant relationship in Banjul, Dakar, Doula, Conakry, and Cotonou which are clear signals that other processes such as prevailing winds, coastal topography and thunderstorms could play a part. The monthly SST climatology indicates that June, July, August, and September are the months where SST is anomalously low in all the stations considered except Conakry thus agreeing with previous studies that SST monthly variability depends on the location though the changes are much greater in the summer months than winter thus further providing lucid explanation to rainfall variability during the summer months in West Africa.

In view of the forgoing, the study concludes that the spatial variability of SST established in this study has enormous implication on rainfall dynamics and the West African climate system. The anomalously low SST established in the study indicates that coastal regions will experience below normal rainfall in the months of JJA, and SON irrespective of its geographical location in the coast. Coastal upwelling has been implicated as the causative mechanism for the anomalously low SST thus driving erratic rainfall in mid-summer giving rise to the phenomenon of Little Dry Season LDS. The study also revealed that SST decreases westward and increases along the coast.

The rainfall variability analysis also indicates that 2010–2020 recorded the highest concentration of rainfall. The pattern of the SST observed may create north–south rainfall dichotomy, a situation in which when precipitation is below normal in the Sahel it will tend to be above normal rainfall in the Guinean coast. The study further affirms that the general increase in the SST of the GOG in the 52 years of the study period (1970–2022) may lead to a dipole structure of rainfall changes over West Africa and the intensification of extremes and a further reduction of the average severity of the LDS by creating the moisture-laden wind more buoyant in terms of moisture and energy and thus in agreement with Chineke et al, 2020