Layout 6 1 ANNALS OF GEOPHYSICS, 62, 5, SE564, 2019; doi: 10.4401/ag-7877 “RECENT SEISMOTECTONIC STRESS REGIME OF MOST SEISMICALLYACTIVE ZONES OF GULF OF GUINEA AND ITS KINEMATIC IMPLICATIONS ON THE ADJOINING SUB-SAHARA WEST AFRICAN REGION„ Ayodeji Adekunle Eluyemi1,2,3, Santanu Baruah3, Sangeeta Sharma3, Saurabh Baruah2,3,* (1) Division of Environment and Earth Science., Centre For Energy Research and Development (CERD), Obafemi Awolowo University (OAU), Ile−Ife, Nigeria (2) Academy of Scientific and Innovative Research (AcSIR), CSIR− North East Institute of Science and Technology (CSIR−NEIST) Campus, Jorhat−785006 (3) Geosciences and Technology Division, CSIR. North−East Institute of Science and Technology, Jorhat−785006, Assam India 1. INTRODUCTION The South Atlantic Ocean around the equator is known as the Gulf of Guinea. This gigantic water body borders most of the West African nations in the south and few of the central Africa nations in the west. Most of the coun- tries located on the adjoining area of the gulf of Guinea are mega cities / national capitals and are high- ly populated, having their various boundaries to the south-Atlantic ocean (gulf of Guinea) and are regarded as stable and aseismic in nature (Figure 1). Though, ev- idence of tectonic activities are numerous on the ocean- ic floor of this zone including submarine volcanic fields, seismicity and the mid Atlantic ridges, the tectonic ac- tivities / seismicity within the Atlantic ocean have a sig- nificant contribution to the intraplate seismic activities of the adjoining areas pertinent to the continental crust [Shimazaki, 1976]. This has resulted into accumulation Article history Received August 6, 2018; accepted April 1, 2019. Subject classification: Gulf of Guinea stress tensor inversion; migration of earthquakes; Cameroun volcanic line; sub-Sahara West African region. ABSTRACT The Tectonic stress regime of Gulf of Guinea region has been studied by stress tensor inversion analysis for the area bounded by lati− tude −10.0 0 S to 4.0oN and longitude −25 0 W to −11.0 0 E. A total of one hundred and four focal mechanism solutions, pertaining to the earthquake events which have occurred in this region, were used for this study. In order to decipher the stress pattern of this region, we have divided the region into four fracture zones namely Romanche, Chain−Romanche, Charcot and Ascension fracture−zones based on seismicity clustering, tectonics and available focal mechanism solutions. The seismicity pattern indicates that none of the nearby coun− tries on the border line between the west Africa region and the gulf of Guinea is devoid of seismicity. Simultaneously, the stress tensor inversion in the four subzones of investigation indicates different types of stress orientations. All of these zones are characterized by vary− ing principal axial directions. The orientation of the principal axial direction along Romanche, Chain−Romanche, Charcot and Ascen− sion fracture−zones are along NE−SW, NE−SW, ENE−WSW and ESE−WNW respectively. The stress tensor inversion results indicate that Chain−Romanche, Charcot and Ascension fracture zones are characterized by extensional stress regime while Romanche fracture zone is characterized by strike−slip stress regime respectively. These patterns imply sea floor spreading activities in and around the mid oceanic ridges and the general orientation of the extensional stress regime is found towards the continents, along the line of migration / pro− gression of earthquakes. of stress and successive migration / progression of oc- currence of earthquakes toward the continental crust from its inherent secondary faulting. Stress accumulated within the earth is often released in form of earthquakes, which take the form of foreshocks, main-shocks, aftershock and earthquake swarms [Michael, 1987]. This has been reflected in the identified seismi- cally active zones of the gulf of Guinea, extending from the north of Ascension Island to the mid Atlantic ridge province. The seismicity pattern indicates clusters of earth- quakes that are well observed within this region (Figure 1). This is substantiated by the evidence of earthquake migrations and successions. On the other hand, it was argued that the intra plate seismic activities on the con- tinental crust of the West Africa is greatly affected by the seismic activities within the oceanic crust which is still very active. In general, stress accumulation in the province of intra plate zones takes a longer period of time to accumulate, hence the recurrence time for intraplate earthquakes is often found to be longer than the inter plate zones Shimazaki [1976] as a result, the continen- tal crust of the west African sub region is seen to be seis- mically quiet (Figure 1). Consequently, the orientation of its seismotectonic stress regime is poorly known. Over the years, the information of the stress regime of a particular region could be obtained from stress ten- sor inversion of focal mechanisms of given sets of earth- quake events occurring from a particular region, which is based on determination of stress field orientation [Mar- tinez-Garzon et al., 2014]. However, the conventional methods of stress tensor determination are through in- version of focal mechanisms. Efforts towards writing up of a code and developing a more precise and accurate means of determining stress tensor have been carried out by various authors [Michael, 1987; Kiratzi, 1999; Abers, 2001; Yamaji, 2006; Zalohar and Vrabec, 2007; Angelier, 2008; de Vicente, 2008; Delvaux and Barth, 2010; Tirifu, 2011; Baruah et al., 2013 and Martinez- Garzon et al., 2014]. In recent past, seismotectonics of the intraplate earth- quake of the west Africa has been linked to the adjoining area of the gulf of Guinea province [Kutu, 2013 and Ak- pan et al., 2014]. Attempts to determine the stress ten- sor and its orientation through in-situ and mechanical methods within this region were made to ascertain the bore-hole lithological-wall stability of the Asahnti deep gold mine in Ghana [Affam and Achibald, 2012]. In-situ stress determination, through ocean drilling pro- ject (ODP) integrated with focal mechanism solutions of earthquake events were carried out in order to have an ELUYEMI ET AL. FIGURE 1. Tectonic setting and epicentres plot of the gulf of Guinea and the adjoining continental crust of the countries of sub−Sa− hara west Africa. In order to depict the seismicity pattern of the region of study, seismic data during the period of 1900− 2016 were utilized. The epicentres are shown in red to green circles (the red circles represent events with depth range of 0−33 Km, and the green circles represent events with depth range of 33−70 Km). Dashed lines indicate the fault lines, the volcanic zones are depicted in yellow triangles. in-depth knowledge of the structural, thermal and sed- imentary processes that take place at the transform margine [Mascle et al., 1998]. Since the region around the gulf of Guinea is comprised of numerous geologi- cal features with active tectonics: including Cameroun volcanic line (CVL), which is estimated to one thousand and eight hundred kilometers long and made up of con- tinental and oceanic sectors Tokam et al. [2010] and Adams et al. [2015] seismically active mid Atlantic ridge and the array of submarine volcanoes on the oceanic floor of the Gulf of Guinea. However, the Adjoining region of the gulf of Guinea (sub Shara West Africa) is far away from the plate bound- aries but it has experienced devastating earthquakes on five occasions, thrice in Ghana, once in Togo and Guinea. During the period of 18th December, 1636 (Ms = 5.7), 1862 (ML ~ 6.5 and Ms ≥ 6.5), 11 th February, 1907, on 22nd of June, 1939 (Ms ~ 6.5 and Mb ~ 6.4) Accra earthquake in Ghana killed sixteen (16) people and one hundred and thirty three (133) others were severely injured. On 22nd December, 1983 in Conakry-Guinea (Mw ~ 6.3) caused the death of three hundred (300) people and several oth- ers were severely injured including loss of properties [Juner, 1941; Blundell, 1976; Bacon and Quah, 1980; Yarwood and Doser, 1990; Amponsah, 2004 and Ram- dani et al., 1984]. Similarly, in recent years, the region has also suffered from devastating volcanic eruption with estimated death of one thousand and seven hundred (1,700) people, only on 21st of August 1986 [Fomine, 2011]. Other similar volcanic eruptions took place in the year 1999 and 2000 [Such et al., 2003] respectively. In this paper, we aim to interpret the active defor- mation of Gulf of Guinea bound by the latitude -10.0o to 4.0oN and longitude -25o to -11.0oE through estima- tion of seismotectonic stress based on inversion of earth- quake focal mechanism solutions, with emphasis on re- construction of the consistent seismotectonic stress state. In fact, prior to stress estimation, we required delineat- ing the major structural domain through an in-depth un- derstanding of present geodynamics. Thus, this study at- tempts to determine the stress tensor inversion of the in- vestigated zones exclusively for: (1) the dimension of az- imuthal variation of compressional axis related to ex- isting geodynamics of the region and (2) the significance of seismotectonic mechanisms related to four major frac- ture lines. The aforementioned features are directly or in- directly connected to stress accumulation and dis- charge of the intra-plate earthquake on the adjoining ar- eas of the gulf of Guinea (continental crust of West Africa), with the scope of seismic hazard assessment and its mit- igation pertaining to the occurrence of future large earth- quakes. We finally intend to better characterize the in- terplate deformation in gulf of Guinea and its kinemat- ic implications on the sub-Sahara West African region, with emphasis on the west African rift system along the Cameroun volcanic line (CVL). 2. TECTONIC SETTINGS Freeth [1977] described the regional tectonics of the west Africa and the gulf of Guinea from a “membrane tectonics” perspective: The West Africa and the Gulf of Guinea with peripheral “compression” was transformed into peripheral “tension”, as African plate moved across the equator having Africa and south America split apart, where a major rift opened up into west Africa rift sys- tem during Hauterivian to Turonian times [Burke et al., 1971; Burke and Whiteman, 1973]. The rifts are the low- er Benue-rift and the Yola rift [Burke et al., 1972]. This involved the separation of continents and zones of ma- jor weaknesses in between the eastern ends of the Yola rift and the Red sea rift which is termed as the effective centre of the African plate known as “Jebel Marra” which is a major geographical feature. This “effective centre” moves towards the equator, as the portion of the plate on the same side of the equator known as the “effective centre” experiences peripheral compression, towards south of the equator during early Cretaceous to Eocene times [Freeth, 1977]. Post early Cretaceous, to Eocene times, “Jebel Mar- ra” moved away from equator, therefore, peripheral ex- tension was experienced, north of the equator. Having acted upon the west African rifts are evidences of pe- ripheral extension or compression, in which resulting membrane effects were therefore completely swamped by the stresses, which opened the Atlantic and the west African rift systems. During the last 30 m.y. “Jebel Mar- ra” has been moving northwards, away from the equa- tor. This is the period of the formation of rifts opening up in west Africa and the Gulf of Guinea, with major rift formation known as the Cameroun volcanic line (CVL) extending up to one thousand and eight hundred kilo- meters (1,800 Km), from the offshore (gulf of Guinea) up to the central African republic [Tokam et al., 2010 and Adams et al., 2015]. Cameroun volcanic line (CVL) is further explained as a subset of the Pan African belt formed by the collision of Sao Francisco, Congo and west African cratons dur- ing Neoproterozoic formation of Gondwana that lies within the Obanguides belt, with multiple shear zones, consisting of central African shear zones, in close as- sociation with Pernambuco lineament in Brazil [Brown and Fairhead, 1983; Dorbath et al., 1986 and Adams et 3 TECTONIC STRESS REGIMES OF THE GULF OF GUINEA ELUYEMI ET AL. 4 al., 2015]. CVL comprises of a chain of Tertiary to re- cent alkaline volcanoes, which split into two arms. One running northward into the northeast of Nigeria, form- ing Biu plateau, while the other arm of CVL runs east- wards through Nagoundere plateau of eastern Camer- oun [Fitton, 1980] thereby forming a “Y” shaped vol- canic feature. Based on long-baseline geodetic GPS data from Africa to Eurasia Kreemer et al. [2014] there is indication that within the framework of northerly extension of gulf of Guinea (25 mm/yr), the present day deformation is sta- tistically significant, relative to mainland Africa which comprises of Guinea, Liberia, Ivory-coast, Ghana, Togo, Benin, Nigeria and Cameroun. 3. DATA USED The global centroid moment tensor (GCMT) and the International seismological centre (ISC) data catalogue revealed the presence of four types of available focal mechanism solutions (strike- slip, normal, thrust and oblique) that are Present along the existing fracture zones of the gulf of Guinea region. Altogether 104 FMS data used were obtained from global centriod moment ten- sor (GCMT) known as Harvard CMT Dziewonski et al. [1981]; Ekström et al. [2012] and International seismo- logical centre (ISC) catalogue project where magnitude (Mw) ranges from 4.8-7.1 and depth range (Km) varies from 3.0-33.2 Km. From the time period of 1977-2016. Since different types of focal mechanism solutions are available for the study region, specific types of focal mechanism solutions are more prevalent than the oth- ers. Therefore, the focal mechanism solutions are not uni- form and neither are these focal mechanism solutions are likely to be controlled by a single fault system with sim- ilar orientations. The epicentral pattern of the studied region can be broadly divided into three, namely: area around the Cameroun volcanic line (CVL), with moderate to active seismicity, the rest of the adjoining area of the gulf of Guinea (west Africa sub-continent) with less seismicity and the gulf of Guinea region, which its earthquake sources has been categorized by the ISC and the GCMT as well as other seismological organizations into: The north of Ascension island, Ascension island and central mid Atlantic ridge with active seismicity. However, due to non availability of FMS data, stress tensor inversion could only be carried out on the gulf of Guinea region alone. The FMS data of the studied zone is mainly char- acterized by normal, strike-slip, oblique and very few thrust fault mechanism. 4. METHODOLOGY 4.1 SEISMICITY In order to completely understand the seismicity pat- tern of the Gulf of Guinea and its implications on the sub Sahara west African region, a total of one thousand and one hundred (1,100) numbers of earthquake events were collected from International seismological centre (ISC), which comprised of historical and recent seismic data from the period of 1900-2016 and magnitude (Mw) range of 3.0-7.1. Figure 1 presents the seismicity pattern of the studied area. The corresponding epicenters are located inside a quadrangle from -10.0 0 N to 17.0 0 N in lati- tude and -25.0 0 E to 17.0 0 E in longitude, which rep- resent a total area of nearly 14,193,792 Km2. Figure 1 summarizes the distribution of the earthquake data as a function of magnitude and depth. The epicentral map, indicate clusters of earthquakes in the gulf of Guinea, within the area of longitude -25 0 E to -11.0 0 E and lat- itude -9.0 0 N to 4.0 0 N. The overall seismicity pattern shows a progression / migration of earthquakes from clus- tered zone in the gulf of Guinea to the adjoining sub- Sahara West African region. 4.2 QUALITY ASSESSMENT OF THE EMPLOYED FOCAL MECHANISM SOLUTIONS The earthquake events of which the fault plane so- lutions were used in our stress tensor inversion (STI) anal- ysis were recorded teleseismically around the world. The global centriod moment tensor (GCMT) catalogue con- tains two descriptions of moment tensor for every earth- quake source. The first one contains the azimuths (jT, jB, jP) and plunges (aT, aB, aP) of the T, B and P axes respectively. Their associated eigenvalues is represent- ed as (lT, lB, lP). The second description is in form of the six independent components of the matrix (m11,m22,m33,m12,m13, and m23) which represent the mo- ment tensor (M) and their respective standard error es- timate as follows: u11,u22,u33,u12,u13 and u23 the stan- dard error estimates is made up of the standard error ten- sor denoted by U. In this study, we employed various sta- tistical approaches that are useful tools in measuring the quality of the focal mechanism solutions employed. The first parameter we considered is the ƒCLVD. This is the sign ratio of the amplitudes of the intermediate and largest principal moments, an indicator used to determine the similarity of a particular focal mechanism solution (FMS) to a compensated linear vector dipole (CLVD), which is same as earthquake produced by faulting occurring si- multaneously along series of properly oriented non par- allel surfaces or more similar to a double-couple solu- tion (emergence of earthquake by slip on a planar sur- 5 TECTONIC STRESS REGIMES OF THE GULF OF GUINEA face). Sanchez and Nuuez [2009] define ƒCLVD statistic as: ƒCLVD = (1) Whenever a compensated linear vector dipole (CLVD) nears ±0.5 the extreme value of ƒCLVD, this implies that the orientation of the T, B and P axes are not conclu- sive. Simultaneously, we also consider evaluating a rela- tive error (Erel) for the respective focal mechanism so- lutions employed in the stress tensor inversion analysis. The Erel compares the relative size of moment tensor M and its standard error tensor U [Davis and Frohlich, 1995; Frohlich et al., 1997 and Sanchez and Nuuez, 2009]. Mathematical expression for Erel is given below as: (Erel) = = (2) The use of colon denotes the tensor scalar product, if A snd B are tensors, their respective elements are aij and bij, then A:B = Si.jaijbij. We therefore consider (Erel) as the norm of U divided by the norm of M. 0 ≤ Erel ≤ 1. In practice, it is usually advisable to seek for values of Erel as low or small as possible. The next parameter that we examined for the qual- ity assessment of the employed FMS is the nfree. This takes into consideration the numbers of moment tensor components of an earthquake source that is not equal to zero. For an instance, if a particular element mij of the moment tensor is equal to zero, the associated stan- dard error uij is also equal to zero. Hence, nfree is de- termined by counting the number of elements of U that satisfy the condition uij ≠ 0. This is an important fac- tor for shallow earthquake sources, whereby the CMT constrains the m12 and m13 elements of the moment ten- sor to be exactly zero. The geometrical effect of restricting one of the P, T or B axes to the vertical and the two oth- ers are confined to the horizontal plane. However, con- straining the P, T or B axes to the vertical in the case of pure double couples for shallow earthquakes is equiv- alent to having a perfect strike-slip, normal or thrust fault. Out of the 463 FMS reported by the ISC and GCMT for the gulf of Guinea region, only 104 satisfy the three aforementioned conditions. Our computed values for ƒCLVD for the gulf of Guinea region range between -0.007- 0.275; (Erel), varies between 0.096-1.00 and earth- quake events that are made up of moment tensor com- ponent of nfree = 6 was also considered for the STI anal- ysis. These are the criteria upon which the selected 104 FMS data for this study were based upon. 4.3 FMS DATA PREPARATION AND STRESS TENSOR INVERSION (STI) Stress tensor inversion analysis was done by selec- tion of one hundred and four (104) FMS data on glob- al centroid moment tensor (GCMT) catalogue Dziewon- ski et al. [1981] and Ekström et al. [2012] as compiled in Table 1. Statistical analyses of these data were pre- pared for: depth, magnitude and time which are depict- ed through histogram (Figure 2). Figure 3 represents the beach ball presentation of the focal mechanism solutions associated with the earthquake events in respective tec- tonic zones of the investigated region. Respective beach balls were generated using Rake software Louvari and Kiratzi [1997]. − B max( T . P l l l U :U M :M u112 +u222 +u332 +2u122 +2u132 +2u232 m112 +m222 +m332 +2m122 +2m132 +2m232 Zone Number of Events Depth (Km) Magnitude (Mw) FMS Type Min Max Avge Min Max Avge NF SS TF OB Romanche 36 4.0 33.2 18.6 6.1 7.1 5.2 - 27 - 9 Chain / Romanche 34 3.0 25.0 14.0 4.8 6.8 5.8 3 17 - 14 Chacotte 10 10.0 23.2 16.6 4.9 5.6 5.25 5 - - 5 Ascension 24 08.0 33.0 20.5 4.8 6.5 5.65 9 6 1 8 TABLE 1. Shows distribution of earthquake focal mechanism solutions (FMS) within subzones of the investigated zones of the gulf of Guinea. The following acronyms are contained in the above table and are explained as follows: NF indicates normal fault, SS indicates strike−slip fault, TF indicates thrust fault and OB indicates oblique fault mechanism respectively. ELUYEMI ET AL. 6 FIGURE 2. Shows the statistical analysis of 104 focal mechanism solutions selected for STI analysis for the clustered zone of the gulf of Guinea. The following histograms are presented: (a) depth (b) magnitude and (c) time. Database of the focal mecha− nism solutions for the period of 1977−2016 used for the stress tensor inversion analysis. FIGURE 3. Map showing the selected 104 focal mechanism solutions in the study region. All focal mechanism solutions are shown by conventional beach−ball illustrations of global CMT solutions. The beach balls were divided into four subzones, as- sociated to four fracture zones, based on the corre- sponding tectonic domain. The focal parameters of the fault plane solutions utilized in this study (supplemen- tary) and the direction of the stress tensor were calcu- lated and determined using the inversion algorithm of Michael [1984, 1987] also known as Michael’s method and Zalohar & Vrabec [2007] known as Gauss method, simultaneously, in order to constrain the consistency of inferred results. Inversion results of the 104 fault plane solutions are zone-wise tabulated in Table 2. The oper- ational principles of the Gauss method is based upon the concept of best fit of stress tensor, with consideration for angular misfit between resolved shear stress and actu- al direction of movement of fault plane. Secondly, it also takes into account the ratio between the normal and shear stress on fault plane. The optimal stress tensors for each homogeneous sub-system of faults are found by maxi- mizing the object function, which is defined through a summation of the compatibility function for all the fo- cal mechanism solutions. Michael’s principle works upon a condition that must be satisfied by the input data. Dur- ing the observed time interval, stress is assumed to be uniform in respective zones of investigation, earthquakes are assumed to be shear-dislocated on pre-existing faults and slips occur in the direction of resolved shear stress on fault plane. Michael method accounts for determi- nation of three principal stresses: maximum stress, in- termediate stress and minimum stress, from statistical technique of bootstrapping re-sampling. Orientations of the principal stresses s1 and s3 of the stress tensor in- version results derived from Gauss and Michael meth- ods are illustrated in Figure 4. The world stress map of the study region (gulf of Guinea) Heidbach et al. [2016] is presented In figure 5 and the geodetic GPS data from Africa to Eurasia Kreemer et al. [2014] movement of African plate towards the Eurasia plate is indicated with dashed arrow line, in Figure 6. 5. RESULT 5.1 STRESS INVERSION AND EARTHQUAKE MECHANISMS: SEISMOTECTONIC CONSIDERATION In order to examine the homogeneity or variations in stress regime within the identified active seismicity zone (clusters of earthquake events) in the Gulf of Guinea (Fig- ure 1). Methods of Michael [1984, 1987] and Gauss by Zalohar and Vrabec [2007] have been implemented. The results of stress tensor inversion for each of the mentioned four fracture zones are shown in Table 2. The principal stresses and obtained orientations from Michael and Gauss methods are illustrated in Figure 4. Romanche fracture zone comprised of stress tensor inversion of 36 focal mechanism solutions which shows that the principal ex- 7 TECTONIC STRESS REGIMES OF THE GULF OF GUINEA Zone Method Used (σ1) (σ2) (σ3) ɸ ɸ Avg Variance Std Dev. Azim (°) Pln (°) Azim(°) Pln (°) Azim (°) Pln (°) Romanche Fault zone Michael 135.9 11.3 -52.7 78.6 45.7 1.2 0.75 0.63 0.019 0.138 Gauss 127.0 7.0 272.0 81.0 37.0 5.0 0.5 - - Chain/Romanche Fault Zone Michael 159.9 87.8 -39.8 2.1 50.2 0.7 0.8 0.70 0.089 0.293 Gauss 305 65.0 145.0 24.0 51.0 8.0 0.6 - - Chacotte Fault Zone Michael -71.8 66.5 137.8 20.6 43.8 10.6 0.44 0.52 0.084 0.289 Gauss 226.0 76.0 339.0 6.0 71.0 13.0 0.6 - - Ascension Fault Zone Michael -172.6 19.6 0.1 70.0 96.4 2.3 0.82 0.76 0.100 0.316 Gauss 2.0 86.0 189.0 4.0 98.0 0.0 0.7 - - TABLE 2. Result of stress tensor inversion for the four subzones of the studied region of the gulf of Guinea using Michael and Gauss methods. ELUYEMI ET AL. 8 FIGURE 4. Stress tension inversion results of the four fault zones for the cluster of events therein using Michael and Gauss meth− ods. Syntheses of stress pattern at the associated fault zones / clusters are depicted with symbols. The black convergent and divergent arrows indicate compression and extensional stress regimes. FIGURE 5. World stress map release, 2016 edition showing the tectonic of the studied region in rectangular box. 3 types of tecton− ics have been identified within the studied region by WSM project: strike−slip (SS), normal (NF) and thrust (TF) [Heid− bach et al., 2016]. The small red box indicated a thrust fault but with only two FMS data, which were found on ISC and GCMT data catalogues for the period of 1977−2016. Perhaps an indicator to the development of a new tectonic domain, these were synthesized with the result of our study in figure 6. tensional axis s3 is along N-E direction. Michael method revealed azimuthal, plunge and phi measurement of this zone, for principal extensional axis (s3) as follows: 45.7 0, 1.7 0 and f = 0.75. The dominant extensional axis (s3) measurement, using Gauss method is as follows: az- imuthal value of 37.0 0, plunge of 5.0 0 and f = 0.5 re- spectively, while the dominant compression axis s1 is along NW-SE direction. Michael method shows azimuth and plunge measurement for s1 as follows: 135.9 0, 11.3 0 while Gauss method revealed azimuth and plunge for s1 as 127.0 0 and 17.0 0 respectively. Chain-Romanche fault zone is made up of stress ten- sor inversion of 34 focal mechanism solutions, Michael method shows s3 with an azimuth of 50.2 0; Plunge of 0.7 0 and f = 0.8 while Gauss method shows, s3 with an azimuth of 51 0; plunge of 8.0 0 and f = 0.6. s3 tends towards north-east direction, Charcot fault zone comprised of stress tensor inversion of 10 focal mechanism solu- tions which revealed that the zone is governed by near- ly east-west extension s3. The dominant principal ex- tensional axis (s3) measured in this zone using Michael method is as follows: azimuthal value = 43.8 0, plunge = 10.6 0 and phi = 0.44. While Gauss method showed az- imuthal value of 71.0 0, plunge = 13.0 0 and phi = 0.6. Ascension fracture zone is made up of stress tensor in- version of 24 focal mechanism solutions, Michael method estimates s3 with azimuth of 96.4 0; plunge of 2.3 0 and f=0.82 0 while Gauss method shows s3, with an azimuth of 98.0 0; plunge of 0.0 0 and f = 0.7. Out of the four fracture zones, only Romanche fracture zones is characterized by strike-slip (SS) stress regime while the rest of the fracture zones are characterized by extensional stress regimes only. 5.2 SYNTHESIS OF STRESS TENSOR INVERSION RESULT AND WORLD STRESS MAP PROJECT The obtained stress tensor Inversion results of the pre- sent investigation have been supported by data of world stress map (WSM) project Heidbach et al. [2016] (Fig- ure 5). The derived stress map from the result of the pre- sent study as it is, in Figure 4 was synthesized with the existing global stress map of the studied region (Figure 5). On a broader scale, two main tectonic stress regimes [strike-slip and extensional stress regime] have been iden- tified that are primarily associated with seismic activ- ities in the Gulf of Guinea. However, two of the exist- ing stress regimes have been well indicated in this study. There is possibility of development of a new stress regime perhaps in its initial stage (Figure 5). The red small rect- angular box, contains just two focal mechanism (FMS) 9 TECTONIC STRESS REGIMES OF THE GULF OF GUINEA FIGURE 6. VMap depicting the synthesis of main stress regime derived from stress tensor inversion associated with four fracture zones. The inputs from global stress map are included as indicated in legend. The symbols are explained in details in the text. Yellow triangle denoted with VP represent the volcanic points; Pairs of convergent arrows denoted by STIZD stands for compression stress regime and divergent arrows denoted by STIZC stands for extensional stress regimes are shown as in Figure 9 are the final determination. Extensional stress regime accounts for 72 per cent of the events out of the total data set. Compiled GPS observation in adjoining region (see text), the large dotted arrow marks on the right top corner of the figure indicate the velocity vector of the African plate with respect to the Arabian and Eurasian plate at the same scale. data belonging to Thrust fault. As such, no stress ten- sor inversion could be done, due to limited availabili- ty of FMS data. Romanche fracture zone is characterized with strike slip stress regime and its stress orientation is well lo- cated, while Chain-Romanche, Charcot and Ascension fracture zones are characterized by extensional stress regimes respectively, each with different degrees of ori- entation. The outcome of our stress tensor inversion re- sult and the WSM data map project of the same region have been synthesized together to generate a detailed and current seismotectonic map of the investigated re- gion. A comprehensive explanation of the present day seismotectonic stress regimes of the gulf of Guinea and its kinematic implication on its adjoining region is pre- sented in Figure 6. 6. DISCUSSION The study region is quite large enough and has sig- nificant implications for seismic hazard assessment of the adjoining continental region of sub-Sahara West Africa because of the presence of five regional fracture zones that are oceanic and continental by extension. The Michael and Gauss methods of stress tensor inversion analysis employed in this work provide reliable and sta- ble results. The results of the two methods are relative- ly comparable in terms of stress axes orientation and stress regimes determination (Table 2). The synthesized stress pattern from focal parameters and associated earthquake events of the seismogenic zones could be inferred. These seismogenic zones can be categorized into three major zones, namely: area around and along the Cameroun volcanic line (CVL) or the west African rift zone, the rest of the adjoining re- gion (West African continental crust) and the gulf of Guinea region, (involving the central mid Atlantic ridge, mid Atlantic ridge, north of Ascension Island and the Ascension Island). Seismicity around the CVL region is directly connected to the Gulf of Guinea region, since the region is the zone of weakness arising from the rifting and the separation of the south American plate from the African plate, there- fore, build up of stresses from the Gulf of guinea region are easily released along the CVL zone in form of earth- quakes more than the adjoining areas. Migration and suc- cession of seismic events in linear pattern are evident to- wards the CVL and the adjoining area, around the west African continental crust, in conformity with Gubin [1960] and Nikonov [1976] from the clusters of the earthquake events within the gulf of Guinea. It is obvious that tec- tonic stresses are being released on a regular basis with- in the studied zone (gulf of Guinea). However, amidst and closer to the clusters of the seismic events, lies an array of submarine volcanic field, within the oceanic segment, which is obvious on the continental segment extending up to Cameroun, Sudan Chad and Algeria. Historically, volcanic eruption along the CVL, on mount Cameroun a subset of the continental segment of CVL took place in 1986, 1999 and 2000 [Fomine, 2011; Suh et al., 2003]. 7. CONCLUSION We have determined the regional stress regime and its orientation in the Gulf of Guinea as well as its kine- matic implications on the adjoining sub-Sahara West Africa region, which was not previously known. The re- sult obtained from our investigation gives a better un- derstanding to the tectonics stress regime of the Gulf of Guinea and a better insight to the work of the earlier re- searchers. Since the adjoining region of the Gulf of Guinea (the West African sub region) is characterized by regional fracture/fault zones that originated from the Gulf of Guinea. The resultant effects from the Gulf of Guinea are easily propagated through these medium to mani- fest themselves in various ways, which can be described as follows: First of all, the investigated region lies on the litho- spheric plate boundaries in between the south America and the African plates, there exists a plate boundary forces as well as potential energy and the ridge-push forces es- timated around 2-3×1012 N per metre of the ridge length [Lister, 1975; Parson and Ritcher, 1980 and Coblentz et al., 1994]. As a result of this, stresses are being gener- ated. The orientation of the generated stress on the Gulf of Guinea, which are mainly extensional stress regime became zones of weakness and tends to fracture as a re- sult of action of forces per unit area (stress = F/A). These explain the origin of the multiple and regional fracture zones in the gulf of Guinea (St Paul, Romanche, Chain, Charcot and Ascension). Interestingly, of all the inves- tigated sub-fracture zones, only Romanche fracture zone is made up of the strike slip regime, which can be ex- plained in terms of plate boundary forces. The rest of the sub fracture zones are made up of extensional stress regime, which is as a result of simultaneous effect of ridge push and a plate boundary forces. For as long as the stress orientation remains in the same direction, as it is, in this study. It is expected that the processes of the fracturing continue, hence the afore- mentioned fracture zones traversed several thousand kilo- ELUYEMI ET AL. 10 meters from mid Atlantic ridge in the Gulf of Guinea on- shore through west Africa into the African plate. The current seismicity being experienced on the con- tinental crust of the West African region could be ex- plained not solely because of transmission of motion from either Romanche or Chain fracture zones in the gulf of Guinea to the continental faults or an early stage in the development of plate boundary on the continent or de- flection of motion from Romanche or Chain fractures to coastal boundary fault as theorized by earlier re- searchers Burke [1971], Blundel [1976] and Yarwood and Doser [1990] but also the seismicity of the West African region can be further explained as a result of release of built up stress in the Gulf of Guinea along the axial ori- entation of extensional stress regimes, which are towards and along the existing fracture zones in the gulf of Guinea, up to the African continent. This is evident in the progression/migration of earthquake events from the gulf of Guinea to the West African sub region (Figure 1). The West African rift system, along the Cameroun vol- canic line (CVL) is mainly a result of ridge push forces exerted on the CVL from the mid Atlantic ridge in the Gulf of Guinea and gravitational potential energy along the mountainous region on the continent in Cameroun. We also observed a direct connection of the CVL to the mid Atlantic ridge system through Charcot and Ascen- sion fracture zones. The present extensional stress regime within the Charcot subzones is oriented towards and along the West African rift system or the CVL. This implies that the West African rift system is still in its ac- tive stage of rifting. Hence a linear progression and mi- gration of earthquakes are observed from the mid Atlantic ridge in Gulf of Guinea through the CVL up to Chad re- gion in (Figure 1). Results obtained from our work also suggest that the incessant eruptions of the CVL could be due to mechanical coupling between the Gulf of Guinea fracture zones (Chain, Charcot and Ascension) and magmatic systems, due to stress release from large earthquakes on to the mag- matic body, thereby perturbing the system and trigger- ing volcanic eruptions. For as long as the orientation of stress from the gulf of Guinea is towards the CVL, the CVL will remain an active volcano for years to come, un- less the stress regime axial direction is rotated away from the CVL. All in all, this work also corroborates the hypothesis of Blundell [1976], which suggested an early stage of de- velopment of a new plate boundary, possibly through fault propagation but not rifting processes, along the Akwaipim fault zone (Ghana) into the continent, as a result of ac- tive tectonism from Romanche fracture zone to Akwaipim fault zone. Our work concludes that similar activities could be true of Kandi regional fault line, on the shore of West Africa, in between Benin republic and Nigeria up to Hogga, a volcanic zone in Algeria North Africa. DATA AND RESOURCES The Global Centroid Moment Tensor Project database was searched using http://www.globalcmt.org/CMT- search.html (last accessed on November 18th, 2016). The International Seismological Centre catalogue was searched using http://www.isc.ac.uk/iscbulletin/search/cat- alogue/ (last accessed was on October, 2nd, 2016). Acknowledgements. Eluyemi A.A. pays gratitude and apprecia- tion to The World Academy for Sciences (TWAS) Italy and Coun- cil for Scientific and Industrial Research (CSIR) India, for a CSIR- TWAS sub UNESCO Ph.D fellowship. The Academy of Scientific and Innovative Research (AcSIR) India is acknowledged by EAA for avail- ing the opportunity to carry-out Ph.D. Authors also acknowledge the International Seismological Centre, On-line Bulletin, http://www.isc.ac.uk Internatl. Seismol. Cent., Thatcham, United Kingdom, 2014 for enabling us access to recent and historic seis- mic data. Authors hereby acknowledge the criticisms and loopholes identified and presented by the anonymous reviewer in contri- bution to the quality of this manuscript. Those comments have sincerely improved the quality of this manuscript. Much profound appreciation is hereby extended. Funding. This work was fully supported by The Council For Sci- entific and Industrial Research (CSIR) India and The World Academy For Sciences (TWAS), Italy [CSIR-TWAS, FR number: 3240275042]. REFERENCES Abers G.A. and J.W. Gephart (2001). Direct Inversion of Earthquake First Motions For Both The Stress Ten- sor and Focal Mechanisms and Application to Southern California, J. Geophys. Res. 106 (B11), 26523-26540. Adams A.N., D.A. Wiens, A.A. Nyblade, G.G. Euler, P.J. Shore, R. Tibi (2015). Lihosphereic Instability and The Source of The Cameroun Volcanic Line: Evi- dence From Rayleigh Wave Velocity Tomography. J. Geophys. Res.: Solid Earth 120 (3), 1708-1727. Afam M., J. Achibald (2012). In-Situ Stress Determina- tion at the Ashanti Mine, IJMMP 3 (1), 1-12. 11 TECTONIC STRESS REGIMES OF THE GULF OF GUINEA Akpan O.U., M.A. Isogun, T.A. Yakubu, A.A. Adepelumi, C.S. Okereke, A.S. Oniku, M.I. Oden (2014). An Eval- uation of The 11th September, 2009 Earthquake and Its Implication For Understanding The Seismotec- tonics of South Western Nigeria, Open J. Geol. 4, 542-550, doi: 10.4236/ojg.2014.410040 Amponsah P.E. (2004). Seismic Activity In Ghana: Past, Present and Future, Ann. Geophys. 47 (2/3), 539-543. Angelier J., S. Baruah (2009). Seismotectonics In North- east India: A Stress Analysis of Focal Mechanism Solutions of Earthquakes and Its Kinematic Impli- cations, Geophys. J. Int. 178 (1), 303-326, https://doi.org/10.1111/j.1365-246X.2009.04107.x Bacon M., A.O. Quaah (1981). Earthquake Activity in Southeastern Ghana (1977-1980), Bull. Seismol. Soc. Am. 71 (3), 771-785. Baruah S., S. Baruah, J.R. Kayal (2013). State of Tectonic Stress In Northeast India and Adjoining South Asia Region: An Appraisal, Bull. Seismol. Soc. Am. 103 (2A), 894-910. https://doi.org/10.1785/0120110354 Blundell D.J. (1976). Active Faults In: West Africa. Earth Plan. Sci. 31, 287-290. Brown J., J. Fairhead (1983). Gravity study of the Cen- tral African Rift System: A model of continental dis- ruption, In: The Ngaoundere and Abu Gabra Rifts. Tectonophysics, 94, 187–203. Burke K.C., T.F.J. Dessauvagie, and A. J. Whiteman (1971). Opening of the Gulf of Guinea and geological his- tory of the Benue Depression and Niger Delta. Na- ture, Physical Science 233, 51-55. Burke K.C., T.F.J. Dessauvagie and A.J. Whiteman (1972). Geological history of the Benue Valley and adjacent Areas, African Geology 187. Burke K.C., A.J. Whiteman (1973). Uplift, Rifting and The Break-up of Africa. In Tarling, D.H and Runcorn, S. K. (Eds.), Implications of Continental Drift to the Earth Sciences, 735-755. Coblentz D.D., R.M. Richardson, M. Sandiford (1994). On the Potential energy of the Earth’s Lithosphere, Tectonophysics 13 (4), 929-945. https://doi.org/10.1029/94TC01033 Davis S.D., C. Frohlich (1995). A Comparison of Moment Tensor Solutions in the Harvard CMT and USGS Cat- alogs, EOS Trans. American Geophysical Mono- graphs 76, F381. Delvaux D., A. Barth (2010). African Stress Pattern from Formal Inversion of Focal Mechanism Data, Tectonophysics 482, 105-128. doi: 10.1016/j.tec- to.2009.05.009. de-Vicente G., S. Cloetingh, A. Munoz-Martin, A. Olaiz, D. Stitch, R. Vegas, J. Galindo-Zaldivar and J. Fer- nandez-Lozano (2008). Inversion of Moment Ten- sor Focal Mechanism for Active Stresses around the Micro continent Iberia: Tectonic Implications, Tec- tonics 27, 1-22. doi:10.1029/2006TC002093 Dorbath C., L. Dorbath, J.D. Fairhead, G.W. Stuart (1986). A Teleseismic Delay Time Study Across The Central African Shear Zone In The Adamawa Re- gion of Cameroon, West Africa, Geophys. J. Int. 86 (3), 751-760. Dziewonski A.M., T.A. Chou, J.H. Woodhouse (1981). De- termination of earthquake source parameters from waveform data for studies of global and regional seismicity. J. Geophys. Res. 86, 2825-2852. doi: 10.1029/JB086Ib04p02825. Ekström, G., M. Nettles, A.M. Dziewonski (2012). The glob- al CMT project 2004-2010: Centroid-moment ten- sors for 13,017 earthquakes, Phys. Earth Planet. In- ter. 200-201, 1-9. doi:10.1016/j.pepi.2012.04.002. Fittion J. G. (1980). The Benue Trough and Cameroun Line - A Migrating Rift System In West Africa. Earth Plan. Sci. Lett. 51, 132-138. https://doi.org/10.1016/0012- 821X(80)90261-7 Fomine F.L.M. (2011). The Strange Lake Nyos Co2 Gas Dis- aster: Impact and The Displacement and Return of the Affected Communities. The Australasian Journal of Disaster and Trauma Studies. ISSN: 1174 - 4707. Freeth S.J. (1978). Tectonic Activity In: West Africa and The Gulf of Guinea Since Jurassic Times-An Ex- planation Based On Membrane Tectonics, Earth Plan. Sci. Lett. 38 (2), 298-300. https://doi.org/10.1016/0012-821X(78)90103-6 Frohlich C., M.F. Coffin, C. Massell, P. Mann, C.L. Schu- ur, S.D. Davis, T. Jones and G. Karner (1997). Con- straints on Macquarie Ridge tectonics provided By Harvard focal mechanisms and telesiesmic earthquake locations, J. Geophys. Res. 102 (B3), 5029-5041. Gubin I.E. (1960). Regularities In: Seismic Manifestations of Tajikistan, Academy of Science, Moscow, 464. Heidbach O., M. Rajabi, K. Reiter, M. Ziegler, W.S.M. Team (2016). World Stress Map Database Release 2016, GFZ Data Services. doi: 10.5880/WSM. 2016.001. Isacks B., J. Oliver, L.R. Skyes (1968). Seismology and The New Global Tectonics, J. Geophys. Res. 73, 5855-5899. Juner N.R. (1941). The Accra Earthquake of 22nd June, 1939, Gold Coast Geological Survey Bulletin, 13. Kiratzi A.A. (1999). Stress Tensor Inversion In Western Greece Using Earthquake Focal Mechanisms From The Kozani-Grevena, 1995 Seismic Sequence, Ann. Geophys. 42 (4). Kutu J.M. (2013). Seismic and Tectonic Correspondence of Major Earthquake Region In Southern Ghana With Mid-Atlantic Transform-Fractured Zones, Int. J. Geosci. 1326-1332. ELUYEMI ET AL. 12 Kreemer C., G. Blewitt, E.C. Klein (2014). A Geodetic Plate Motion and Global Strain Rate Model, Geochem. Geophys. Geosys. 15, 3849-3889. doi: 10.1002/2014GC005407. Lister C.R. (1975). Gravitational Drive on Oceanic Plates Caused by Thermal Contraction, Nature 257, 663-665. Louvari E.K., A.A. Kiratzi (1997). Rake: A Windows Pro- gram To Plot Earthquake Focal Mechanisms and The Orientation of Principal Stresses, Comput. Geosci. (8), 851-857, doi: https://doi.org/10.1016/S0098- 3004(97)00070-8 Martinez-Garzon P., G. Kwiatek, M. Ickrath, M. Bohnoff (2014). A Matlab Package For Stress Inversion Com- bining Solid Classic Methodology, A New Simpli- fies User Handling and A Visualization Tool, Seis- mol. Res. Lett. 85 (4), 896-904, doi: https://doi.org/10.1785/0220130189 Mascle J., G.P. Lohmann, M. Moullade (Eds.) (1998). In- Situ Stress at the Cote D’Ivoire-Ghana Marginal Ridge from FMS Logging In Hole 959D1, Proceed- ings of the Ocean Drilling Program Scientific Re- sults, 159. Mckenzie D.P. (1969). Speculations on The Conse- quences and Causes of Plate Motions, Geophys. J. R. astr. Soc. (18), 1-32. doi: 10.1111/j.1365- 246X.1969.tb00259.x. Michael A.J. (1984). Determination of Stress From Slip Data: Faults and Folds. J. Geophys. Res. 89 (B13), 11517-11526. Michael A.J. (1987). Stress Rotation During The Coalin- ga After Shock Sequence. J. Geophys. Res. 92 (B8), 7963-7979. Mogi K. (1974). Active Periods in the World’s Chief Seis- mic Belts, Tectonophysics 22, 265-282, doi: https://doi.org/10.1016/0040-1951(74)90086-9 Grant N.K. (1971). South Atlantic, Benue Trough, and Gulf of Guinea Cretaceous Triple Junc- tion, Bull. Geol. Soc. Am. 82 (8), 2295- 2 2 9 8 , doi: https://doi.org/10.1130/0016- 7606(1971)82[2295:SABTAG]2.0.CO;2 Nikonov A.A. (1976). Migration of Large Earthquakes Along The Great Fault Zones In Middle Asia, Tectonophysics 31, 55-60, doi: https://doi.org/10.1016/0040-1951(76)90113-X Parsons B., F.M. Richter (1980). A Relation between Driv- ing Force and Geoid Anomaly Associated with mid- ocean ridges, Earth Plan. Sci. Lett. 51, 445-450. Ramdani M., B. Tadili, S.D. Ben (1984). A Strong Earth- quake Strikes Guinea (22nd DECEMBER 1983), Bull. Seismol. Soc. Am. 74 (5), 2045-2047. Rhodes R.C. (1971). Structural Geometry of Subvolcanic Ring Complexes As Related To Pre-Cenozoic Mo- tions of Continental Plates, Tectonophysics 12, 111- 117, doi: https://doi.org/10.1016/0040- 1951(71)90012-6 Sanchez J.J., F.J. Nuuez-Cornu (2009). Sesmicity and Stress in a Tectonically Complex Region: The Rivera Fracture Zone, the Rivera-Cocos Boundary, and the Southwestern Jalisco Block, Mexico, Bull. Seismol. Soc. Am. 99 (5), 2771-2783, doi:10.1785/0120080350 Shimazaki K. (1976). Intra-plate Seismicity and Intra-plate Earthquake. Historical Activity In South Western Japan, Tectonophysics 33, 33-42, doi: https://doi.org/10.1016/0040-1951(76)90050-0 Suh C.E., R.S.J. Sparks, J.G. Fitton, S.N. Ayonghe, C. An- nen, R. Nana, A. Luckman (2003). The 1999 and 2000 Eruptions of Mount Cameroon: eruption be- havior and petrochemistry of lava, Bull. Volcanol. 65, 267-281. doi. 10.1007/s 00445-002-0257-7 Tirifu C.I., V. Shumila (2011). The Analysis of Stress Ten- sor Determined From Seismic Moment Tensor So- lutions At Goldex Mine Quebec, American Rock Me- chanics Association, 11-584. Tokam K.A.P., C.T. Tabod, A.A. Nyblade, J. Julia, D.A. Wiens, M.E. Pasyanos (2010). Structure of The Crust Beneath Cameroon, West Africa, From The Joint In- version of Rayleigh Wave Group Velocities and Re- ceiver Functions, Geophys. J. Int. 183 (2), 106-1076, doi: https://doi.org/10.1111/j.1365- 246X.2010.04776.x Turcotte D.L., E.R. Oxburgh (1976). Stress Accumulation In The Lithosphere, Tectonophysic 35 (1-3), 183- 199, doi: https://doi.org/10.1016/0040- 1951(76)90037-8 Wright J.B. (1968). South Atlantic Continental Drift and the Benue Trough, Tectonophysics 6 (4), 273-352. doi: https://doi.org/10.1016/0040-1951(68)90046-2 Yamaji, A., K. Sato (2006). Distance For The Solution of Stress Tensor Invasion In Relation to Misfit Angles That Accompany The Solutions, Geophys. J. Int. 167(2), 933-942, doi:10.1111/j.1365- 246X.2006.03188.x Yarwood D.R., D.I. Doser (1990). Deflection of Oceanic Transform Motion at a Continental Margin as de- duced from Waveform inversion of the 1939 Ac- cra, Ghana earthquake, Tectonophysics 172 (3-4), 341-349, doi: https://doi.org/10.1016/0040- 1951(90)90040-F Zalohar J., M. Vrabec (2007). Paleostress Analysis of het- erogeneous Fault-Slip Data: The Gauss Method. J. Struct. Geol. 29, 1798-1810, doi: 10.1016/j.jsg.2007.06.009 13 TECTONIC STRESS REGIMES OF THE GULF OF GUINEA *CORRESPONDING AUTHOR: Saurabh BARUAH, Academy of Scientific and Innovative Research (AcSIR) CSIR - North East Institute of Science and Technology (CSIR-NEIST) Campus, Jorhat-785006, Assam India. Geosciences and Technology Division, CSIR - North East Institute of Science and Technology, Jorhat-785006, Assam India email: saurabhb_23@yahoo.com © 2019 the Istituto Nazionale di Geofisica e Vulcanologia. All rights reserved ELUYEMI ET AL. 14