The dataset used for the current study is a combination of recent and historical data. The recent data have been collected in March 2015, using a Lacoste & Romberg G-823 gravimeter. The historical data come from 183 gravity stations that are part of the dataset collected by ORSTOM in 1967. The recent dataset corresponds to additional 223 gravity stations realised in the Kribi-Campo-Ma’an region (which until this gravity campaign done, was not covered by ground geophysical studies). This led to a resulting dataset of 406 gravity measurement sites.

The recent geophysical campaign involves two collecting data profiles, along the N-S and W-E direction as well. The N-S survey profile follows the coastal Kribi-Campo road while the W-E (perpendicular to the first) was realised from Campo to Ma’an, passing through the Memve’ele hydro-electrical dam site. The distance between different stations ranges from 0.5 to 2 km (black circles on Figure 2). The base station was

set at Kribi, as for past gravity campaigns [20] [21] , in order to increase the accuracy of these new collected data and also, because the Kribi base station is linked to the international base station of Douala. The irregularities in the distribution of the gravity points are due to the difficulties in accessing the area, which belongs to the tropical rain forest. Before merging data, the recent measurements were corrected from the luni-so- lar tidal effects and instrumental drift (herein assumed to be linear in time). Free-air and Bouguer reductions based on a mean density of 2.67 g/cm³ were applied and the simple Bouguer anomalies have been derived. The resulted Bouguer anomalies are the basic data used in this study context (Figure 2). Since data was collected onshore, the sea-data have not been included in this study.

The simple Bouguer anomaly map (Figure 2) of the region is obtained after interpolation by combining the existing data set and the one collected during the 2015 campaign, and then applying a minimum curvature gridding algorithm implemented in Oasis Montaj 8.0 software. The Bouguer anomaly values in the grid are encompassed in the −68.7 mGal −2.0 mGal interval.

This study applies the horizontal gradient technique to upward-continued gravity grids at various heights, in order to assess the peaks in the concerned grids. The technique is referred to as the multiscale analysis [9] [22] . It appears straight forward that the base analytical tools are the Upward Continuation (UC) and the Horizontal Gradient (HG) methods. In addition, we performed the 3-D Euler Deconvolution in an attempt to find depth to basement of investigated lineaments.

Figure 3 below shows the Bouguer anomaly map of the study area. It reflects lateral variations of density in the basement and shows several anomalies with various values and shapes. It is produced from the interpolation of the gravity map from Figure 2

which enabled to fill the grid dummies. It is characterized by gravity values ranging from −68.7 mGal to more than −2.0 mGal. This is in accordance with the fact that gravity data have been collected onshore (and by the fact that the continental crust is less dense that the oceanic crust).

The Bouguer map displays three main types of gravity anomalies over the study area. These are accordingly high, intermediate and low values gravity anomalies. High gravity anomalies (values ranging from −38.5 to −2.0 mGal) occupy the broad western side of the study area. The broad zone characterised by high gravity anomaly values also exhibits four peaks. The peaks located near the margin (three peaks) are circular and line up along the N-S direction, from Campo to Kribi. The broadest peak is shifted to the right, east of Kribi; it is oval and nearly curved to the left at its upper part. All these peaks may correspond to dense or basic intrusive bodies within the main formation broadly characterised by the high intensity anomaly values. The anomaly field herein most likely corresponds to the signal of Paleoproterozoic terrains that occur in the study area’s part. Alternatively, the two peaks observable in Kribi neighbourhood may rather be considered as a unique very high-valued anomaly corresponding to a ring complex affected by a nearly N-S trending discontinuity. In Addition, a moderate high gravity anomaly occurs at the SE edge of the Bouguer map. This is a NW-SE elongated anomaly that is probably caused by topographic effects or an intrusive body. The high gravity anomaly values domain is flanked to the east by an intermediate anomaly domain whose values are between −50.8 mGal to −38.5 mGal.

The domain characterised by intermediate Bouguer’s anomaly values forms the centre domain. This domain is separated from the previous described domain by a N-S trending gravity gradient west from Nyabessan (Figure 3). This suggests a discontinuity which seems gradational to south and steep to the north of the study area. Globally, the area is fairly quiet although it is disturbed by a high NW elongated anomalies (−40.1 mGal to −38.5 mGal) at Ma’an and a trough at Nyabessan, respectively. The anomaly peak at Ma’an is affected by a NE discontinuity that can be interpreted as a lineament. Careful observation of the Bouguer gravity map tends to suggest the existence of NE-SW discontinuity between Akom II and Nyabessan. This is probably a shear zone or faulting that affects the deep basement. The gravity anomaly field herein may correspond to the signature of the Archaean basement of the study area. The intermediate gravity anomaly field domain narrows to the north at the vicinity of Bipindi. This domain is separated to the east by a NW-SE trending domain of very low gravity values herein referred to as the NE domain.

The NE domain is characterised by very amplitude anomaly values within the interval −68.7 mGal to −50.8 mGal interval. It is a NW-SE trending domain at the NE margin of the study area, thus implying a NW-SE contact limit with the centre domain to its west flank. The domain forms a NW-SE basin-like environment NE of Akom II. The basement of this basin is made up of low density formations, probably granites or granitoids of the Ntem complex.

The general disposition of anomalies on the Bouguer map from the study area shows a decrease in the anomaly field values from the west to east. This corresponds to a decrease in density from west to east. We therefore assume that formations to the west are denser than those to the east, as we move from the shore to the hinterland.

The upward continuation of the gravity field at increasing heights highlights the gravity effect of deep sources. This is because upward continuation suppresses the signals due to small, shallow bodies [38] .

The map above (Figure 4) shows the Bouguer gravity anomalies map upward continued to 10 km from the study area. Although continuations are realised for various heights and used in the subsequent sections, the Figure 4 (corresponding to the highest level) above has been chosen to illustrate the behaviour of gravity anomaly affecting the basement. Some previous observations made on the simple Bouguer anomalies (Figure 3) are confirmed although differences are noticeable on this map. The moderate gravity anomalies observed in the study area map are well smoothed and could reflect the gravity effects of deep or regional structures, while other anomalies are better amplified.

This proofs that, they are also regional anomalies probably caused by deep or regional basement structures.

All these gravity anomalies tend to expand as the upward continuation height increases. The heavy anomalies trend broadly N-S. Besides there is a NW trough observed on both maps (Figure 3 and Figure 4) probably caused by faulting processes.

To enhance lateral boundaries of density contrast in potential field data, we calculated the horizontal gradient of the Bouguer gravity anomaly. The horizontal gradient of the Bouguer anomaly data was calculated in the frequency domain and its amplitudes provided the horizontal gradient map presented in Figure 5 below. The horizontal gradient magnitude map obtained enhances some short wavelength features and highlights areas with steep density variations, which could be interpreted either as faults, geologic contacts, or as intrusive formations as well. It is observed that the pattern of the high gradient anomalies is broad, not like sharp ones of ideal vertical boundaries of density contrast.

Analysis of the whole horizontal gradient magnitude map (Figure 5) highlights

various geological features, which are characterized by differences in gradients values. The map exhibits different gradients values zones with various amplitudes, shapes and little coherent regional fabric. These different zones are bounded by relative steep or abrupt gradients in some places. The horizontal gradient values range from 0.2 mGal/km to 5.7 mGal/km. It appears on Figure 5 that there is a broad region of high horizontal gradient anomalies (magenta to red colour) that dominates the entire map. This region departs from the background gradient anomaly mainly characterised by greenish colours.

The high horizontal gradient anomalies domain is characterized by a relative high gradient value (magenta to red colour) disseminated in the whole map within the moderate background (greenish). This domain is dominantly spans from the west to nearly centre-east part of the study area. The trends of these high amplitude horizontal gradients are approximately NNE, NNW to N-S in certain places. Besides, there are many very low intensity gradient anomalies that are also noticeable on the horizontal gradient magnitude map (Figure 5). These anomalies are characterized by troughs (blue colour) that affect the background field. The low amplitude anomalies are majorly closed and elongated along the NW-SE and E-W directions, though other trends are seemingly acknowledgeable.

To sum up, the appearance of the horizontal gravity map (Figure 5) suggests that the study area underwent an intense tectonic activity that affected its basement. This tectonic activity is testified by the fault and fracture occurrences which are represented by elongated high gradients; many intrusions represented by close-contoured high amplitudes anomalies. The horizontal gradient map shows that the west to centre of the study area is strongly marked by that brittle tectonics more than the eastern side.

Accordingly with the geological information, in the whole map, the gravity gradient highs, are bounded by relatively steep gradients. The presence of high gravity gradients, which occur over these basic rocks, suggests the existence of a suture zone between two blocks of the crust or the existence of dense mantle rock or uprising of some mantle materials in these places. It might also represent the thinning of upper crust due to the mantle materials uplift. Then, a relative elongated high gravity gradient anomaly at the north-western and middle southern portion could be due to the uplift of deeper mantle materials, dense than the surface materials i.e. seems to represent the signature of dense mantle materials or rocks in this zone. This zone with a strong gradient could also corresponding to contact or fault structures.

In addition, the relative low gradient value also highlighted on the horizontal gradient of the Bouguer gravity anomaly, might indicate the presence of lower density materials. It uplifts zones of high and moderate gradients. The long linear gradient anomalies are indicative of intra-basement faults or crustal block edges.

The maxima of the horizontal gradient of a the Bouguer anomalies shown in Figure 6 below help to locate contacts associated with abrupt changes in density and the mul-

ti-scale analysis of these maxima involves upward continuation of the gravity field to different heights with a view to characterize the vertical extension of anomalous structures. Faults and contacts are expressed by a quasi-linear disposition of many maxima and horizontal limits of intrusive bodies are showed by their quasi-circular disposition [9] [29] [39] . In general, the local peaks in the magnitude of the horizontal gradient give the locations of the large density contacts. The steepest horizontal gradient of a gravity anomaly will be located directly over the edge of the body if the edge is vertical and far away from any other edge or source [40] .

The maxima of the horizontal gradient of the Bouguer anomalies showed in Figure 6 below exhibit most structures features as faults/fractures and geologic boundaries. The linear local peaks of horizontal gradient magnitude trending in NNW, NNE and N-S directions. While the circular disposition of local peaks of horizontal gradient magnitude occurred at the west side present the southward extension.

Multi-scale edge mapping is a process of edge detection in potential field data. During the process, the positions of gradients maxima are detected across multiple heights of upward continuation. The multi-scale analysis horizontal gradient map is obtained by showing the maxima from various levels on the same plot. The offsets observed on the arrangement of maxima inform about the dip of the structure outlined [41] [42] .

Figure 7 below is the multi-scale edge map from horizontal gradient of upward continued Bouguer anomaly map from 0 to 10 km height respectively (colour description is provided with the map of Figure 6). The analysis of the multi-scale map (Figure 7) below enables to outline many geological features, especially lineaments. According to Figure 8, the map enables to highlight many lineaments. The features outlined correspond probably to the major elements that affect the basement. These are lineaments which follow the E-W, N-S, NE and NW directions. However, the NW directions are the more represented. Some of these lineaments are consistent with the features outlined from the Bouguer anomaly map (Figure 3).

The lineaments highlighted on Figure 7 can be interpreted as contacts or faults, though the interpretation as contacts seems more consistent. The multi-scale analysis shows that most the local maxima plotted are clustered along well designed lines. In

addition, these maxima don’t appear to be much offset from the cluster lines outlined. These allow us to interpret the lineaments as contacts, whose directions are E-W, N-S, NE-SW and NW-SE (the dominant directions being NW-SE and N-S as shown on the map from Figure 7). The good clustering or few offsets observed on local maxima enable to suggest that the contacts outlined have vertical to subvertical dips. However, probable interpretation of these features is that they are vertical to subvertical faults which follow the aforementioned directions.

In Figure 8, there is a major lineament that occurred at the Akom II North-eastern side which could be interpreted as the contact. In the other hand, this major lineament seems to be a buried fault named F1, directed NW-SSE. The same observations are made at Campo (fault F6) and Kribi (F8 and F9 faults) neighbouring. The anomaly at Bipindi could be considered as an intrusion while a major NE-SW deep fault occurs eastern of Kribi.

Close inspection of the features on Figure 8 reveals that, the study area is affected by deep-seated structural lineaments oriented in the NW-SE, E-W and N-S directions. The NW-SE trend is strongly developed than the other identified trends. It could represent the prevailing tectonic trend in the area, and it could have played an important role in the formation of its tectonic framework.

In summary, there is: 1) a family of E-W, N-S, NE-SW and NW-SE structural features, which corresponds to contacts associated to several deep-buried faults (F1, F6, F8 and F9); and 2) circular structures interpreted as circular contacts (C1 to C4) or delineation of intrusive bodies (I).

In this work, to enhance the estimation of the depth to basement solutions, the 3D Euler Deconvolution technique was performed on the Bouguer data in an attempt to find depth to basement of lineaments. The procedure is realised with a window size of 10 grid cells, a depth tolerance of 15% and, structural indices 0 and 0.5 to efficiently locate contact boundaries, faults and dykes. These structural index values are commonly used in gravity studies for that purposes [21] [43] - [47] .

Figure 9 and Figure 10 below show the results of the Euler method applied to the

Bouguer map (Figure 3) for the SI value of 0. The structural interpretation map of the area (Figure 9) shows the different features affecting the area. The computational results reveal that the depths obtained range between 2 and more than 16 km. Outputs from the technique show that there are many linear segments trending in NW, NNW and N-S directions. While linear elements correspond to faults or basement fractures, curved elements express the presence of a probable deep contacts or geological boundaries. Most of these lineaments ( interpreted as F1’, F2’ and F9’ faults) trends are consistent with some previous directions highlighted by the maxima of the horizontal gradient magnitude. The recorded depths (up to the [8 km - 11 km] interval) show they could be considered as buried faults. In addition, Euler solutions show non-uniform depth distribution.

Euler solutions for the structural index 0.5 (Figure 10) also show good clustering along the linear and circular segments which trend in the NW, NNW, NNE, E-W and N-S directions. There are many new structural features/ lineaments highlighted (Figure 10). The depth estimates range from 2 km to 21 km for the subsurface structures.

Comparison of the two Euler Deconvolution solutions maps (Figure 9 and Figure 10) enhances the delineation of most structural features. Lineaments L1 and L2 line up in the same direction as F2’ and F1’, while L7, L19 and L22 are superimposed on F3’, F9’ and F8’ faults. These significant facts exhibit the real location and extension of these vertical-dipping features (L7 and L19, as well) which trend NW, NNE and N-S.