Water-Bearing Complexes and Fault Breccia Aquifers of the Sanaga Valley Fault System: New Insight from Outcrops, Wells, and Geophysics Data

1. Introduction

The Sanaga (Figure 1) is the longest and largest river system in Cameroon. It flows through a contorted drainage pattern characterized by the association of three main structural trends, including NW-SE, N-S, and NE-SW fault traces. The regional study carried out by [1] indicated that these fault lines are commonly associated with linear depressions of up to 550sqkm and igneous intrusions of Paleozoic age [57]. According to [2,3], these depressions hosted continental and shallow marine sediments of 100 m to more than 400 m thick. Also, during the 1970 and 1980 hydrogeological research missions, these features were targeted as clastic aquifer systems [4,5,6,7,8,9,10]. To date, several of these surface structures have been drilled (Figure 1), but a clear definition of their hydrogeological conditions and reservoir parameters prediction is still not documented or published. Collaborative projects carried out through funding from the public administrations, groundwater companies, and international cooperation, have enabled new data to be collected. The study involves analysis of wells and geophysics data derived from these projects. The produced cross-sections show prominent features that can help aquifers risk assessment of the Sanaga water-bearing complexes and fault breccia.

2. Geological Setting

The Sanaga Valley Fault System (Figure 1) trends east-north-east, west-south-west for approximately 510km long and 85km wide across the Pan-African belt of the north Congo Craton. It’s one of the geotectonic elements of the Central African shear zone [11,12,13,14,15]. The westernmost segment (Figure 2) of this structure corresponds closely with the Paleozoic rift trend of the South Atlantic Ocean [16]. Its eastern segment runs from Puma to Lom basin (Figure 2) passing through the Kikot-Ndockoa plain and is most famous for alluvial gold production in the Betare Oya district. Studies carried out by several authors [17,18,19,20] indicated that the gold-bearing rock formations in the Lom basin consist of argillaceous sandstones with intercalations of volcanic sequences. The regional geology (Figure 2) of the study is characterized by a predominance of Meso- and Neoproterozoic bedrocks [21,22,23]. Field observations conducted by many authors [24,25,26,27] show that these bedrocks are subdivided into two series. Namely, a high-degree suite [23,24,25,26,27,28,29] represented by granitic to migmatitic gneisses belt of Bafia, Lom, and Mbalmayo-Bengbis domains; and a low-grade series represented by schists, phyllite, and slate formations [30,31]. According to some previous studies [2,9,10,32] Paleozoïc to early Mesozoic sedimentary rocks covered part of basement formations. Also, many areas with fault rocks and areas in which the basement sequences are covered with a thick stratum of alteration or alluvium have been documented [9]. The data presented in this article provide more arguments about the deposition architecture of these sequences.

3. Data and Methods

The data used for this study consists of outcrop photographs, electrical resistivity images, and cuttings samples acquired over buttes and season river spreading topo-graphic areas. These data were collected through five rural hydraulic projects, financed by local administrations, individuals, and international cooperation. The main work carried out during these projects includes geological and hydrogeological site recognition, geophysical measurements, drilling operations, and water flow rate evaluation. In addition to these recordings, the free-air satellite gravity data version (V21.1, Figure 1) derived from the global gravity database [33] was used in this study for the spatial reconstruction of basement deformation around the Sanaga and adjacent areas. According to [34], this gravity database contains a shorter wavelength (< 14 km) and spatial resolution (~7 km) adequate for the study of subsurface deformation.

3.1. Site Recognition

The site survey consists of a 1 km radius walk around the perimeter of the project area to identify the nature of the rock outcrops encountered, calculate the valley floor elevation, and make a catalog of the previous groundwater engineering projects. This step also provided the constraints of the geophysics survey design.

3.2. Data Acquisition and Processing

Electrical prospecting, and in particular the resistivity tomography imagery, is widely known for its applications in the delineation of hydrogeological features within the basement and sedimentary formations [35,36,37,38,39,40,41]. In this study, measurements of the resistivity were performed to reach hydrogeological targets that could be located at a depth of 120 m. The choice of this depth range was guided by the results of previous research [9] and the valley floor elevation of 80-100 m measured during the site recognition. Also, to fully describe the lithological succession over this depth interval, several lengths of devices and inter-electrode spacing were used. The most important exploration system used, consisted of a 1000 m long streamer, composed of 50 traces spaced 20 m apart (Table 1).

Table 1 provides a summary of the experiments. The equipment system used to collect these data was a Syscal R1 equipment. Each profile was recorded at eleven levels and with a listening time of four seconds. Also, to ensure the proper working of the system and the accuracy of the recording data, a quality assurance procedure was implemented before, during, and after each acquisition (Figure 3). This procedure includes:

• The identification of potential parasitic sources along the survey line;
• The pushing of the electrodes steel 3/4 in the ground;
• The resistance checking of the connections;
• The visual inspection of the dispersion coefficient (Dev, Figure 3a) during the machine running process. Indeed, during this stage, the machine calculates a Dev and displays it on the screen for each data point or stack. This provides information on the quality of the measuring sequence. The number of stacks (minimum and maximum) and the minimum value of the Dev parameter are defined when the sequence is created. The measurement is repeated as many times as necessary until the Dev value reaches the setting value (Dev ≤ 20 %, Figure 3). If this value is not reached, the system performs additional stacks, up to the defined maximum number;
• The quality control of the measured apparent resistivity pseudosection (Figure 3b) at the end of the process to detect any problems that may have occurred during the survey.

The analysis of the data grid recorded (Figure 3b) revealed that from 449 resistivity data points measured in the ER-03 profile (Table 1), 445 recorded have a Dev value of less than 20%. The processing of these measurements consists of converting the measured apparent resistivity pseudosection (Figure 4) into an interpretable geological model using the inversion program proposed by [36]. Therefore, in addition to the definition of inversion parameters, the processing sketch implemented includes, the examination of the modelled sections (Figure 4). This procedure allows us to select the inversion result that best correlates with the drilling data.

As the overall dip of the geological structure of the study areas was unknown, the method adopted to optimize the geophysical measurement was to collect data along lines arranged at 90°. Therefore, the preliminary phase of this campaign consisted of carrying out electrical profiling to locate anomalies of lateral variations of geological formations [42]. The device used for all these measurements was the Wenner-Shlumberger four poles array [43]. According to [43], this configuration is a combination of two measurement systems and combines sensitivity and signal amplitude. Another advantage is that the operator can use it in noisy terrain and without prior knowledge of the geometry of the geological body to be studied.

4. Results

Resistivity sections presented in Figure 5 describe the profile of the subsoil comprised between the ground surface and a maximum depth of 90 m. Also, using the electric character of each line, we have distinguished five main resistivity facies on the mapping deposystem. These facies are described below in terms of geoelectric units (Table 2) from depth to the topographic surface.his section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.

• Unit U1

Its electrical attribute is most developed in sections ER-01, ER-02, and ER-03 (Figure 5). It corresponds to a very conductive body with a constant resistivity character ranging between 10-300 Ω.m. The base of this unit is not specially imaged because of the lengths of the survey array used. Its top appears at very variable depth due to its configuration marked by anomalies of plume corps which cross vertically the whole sequence in the form of a diapir or dyke. The results obtained on profile ER-03 show that the height of these anomalies can reach 80 m locally.

• Unit U2

This unit corresponds to a very resistant unit compared to its overlying U1 sequence. Its thickness in profile ER-01 (ER-01, Figure 5) varies between 23 and 45 m, while its resistivity over the ER-01 and ER-04 sections varies between 409 and 5020 Ω.m. Its external configuration with the form of a progradation wedge is locally restructured by deep deformation.

• Unit U3

This consists of a conductive terrain with resistivities varying between 60 and 700 Ω.m and corresponds to the most widespread facies of the study area. Its base follows the morphology of U2 on almost all the profiles, whereas it seems to be only slightly eroded on profile ER-04 (Figure 5). Regionally, its external structure corresponds to a bed-ding sequence whose thickness varies between 25-40 m. Its internal structure near the floodplain (ER-04, Figure 5) is characterized by a most continuous sequence of conductive character.

• Unit U4

This unit is semi-resistant and is more developed in the areas where the valley floor elevation is higher (30 to 80 m). Its electric character (ER-03, Figure 5) is almost uniform and reflects a resistivity range between 669 and 6031 Ω.m. On profile ER-04 its base shows signatures that could correspond to erosion marks. The limit of its top locally coincides with the topographic surface, especially in areas of low valley floor elevation (10-15 m). Also, its thickness varies laterally from 20 m in the transition zone to 70 m in the source zone of the Sanaga River channel. Regionally, its internal structure is characterized by a vertical alternation and includes resistive and conductive sequences. The resistive sequences appear locally in the form of small lens structures (ER-03, Figure 5).

• Unit U5

U5 (ER-03, ER-05, Figure 5) is a semi-conductive unit with resistivity values comprised between 594 and 874 Ω.m, located directly in contact with the topographic surface. The electric pattern observed on profile ER-05 (Figure 5), shows that U5 is more developed at the hilltop and along the slope of topographic depressions. Its thickness reaches locally 15 m.

5. Discussion

5.1. Data Accuracy and Deep Attributes of the Sanaga Buttes and Season River Spreading Topographic Features

Several authors have presented inversion software and algorithms for processing electrical resistivity tomography data [36,38,44,45,46]. The objective of these processing approaches is to produce a model of the subsurface attributes that is close to the geological subsoil crossed by the electric current. This study uses the algorithms developed by [36] to process the data.

The wide use of this algorithm by many academics and private companies is a credit to its performance [40]. Table 2 shows the RMS values obtained at the end of the inversion process of each line and testify to the quality of the measurements made. Also, from a detailed analysis of all inversion results it appears that the resistivity values of 159-22011 Ω.m recorded in the ER-01 profile are affected by an error of a factor of 10% which is propagated over the entire section. This error is because the ground where the electrodes were placed at the time of the measurements was too dry and consisted of an indurated horizon. Therefore, for the interpretation of the highlighted formations, two types of inverse model sections display have been implemented. These included the logarithmic and the linear contour intervals display. The control of the accuracy of sections obtained from both methods allows us to discover that the modeling using linear contour intervals produced element size of geologic features that better correlated with the drilling results (Figure 6). This method is well known for the interpretation of induced polarisation tomography [40]. The results obtained in this study show that this method can be also applied to resistivity data for the recognition of a thin sand layer in weathering or massive clay succession. Thus, in addition to the fundamental aspect based on the determination of the geological model from the resistivity measurements, the results obtained provide new elements that contribute to the discussion on the lithology and deformation of the study area. From the spatial analysis of the data collected including electrical resistivity measurements (Figure 5), boreholes (Figure 6), and field observations (Figure 7), we find that the subsurface geology of buttes and season river spreading zones of the Sanaga Basin consists of sedimentary deposits. The diversity of facies observed on the ER sections (Figure 5) and their low level of compaction marked by drill-ing tool advancement speeds of less than one minute per meter suggests that they may be Paleozoic to Early Mesozoic sedimentary sequences described by previous studies [2,9,10,47]. The most remarkable facies are that of Unit 3 (Figure 7) which shows an almost rhythmic alternation of red, pink, gray, brown, and black shale facies. In contrast, the resistivity facies of unit U4 appear much more uniform. The overall sequence shows vertical intrusion anomaly marks. The shape of these anomalies recorded in ER-03 (Figure 8) as well as the very low resistivity values measured at the contact of these bodies suggest that they could be related to mud intrusion. Also, the identification of brackish saline water in some traditional boreholes of the study area suggests that this mud section may be salt-rich.

Table 3. Lithological interpretation of identified electrical units.

Table 3. Lithological interpretation of identified electrical units.

5.2. Hydrogeological System and Sub-Outcrop Aquifers of the Buttes and Season River Spreading Zones of the Sanaga Basin

Two approaches are generally used to interpret electrical units in terms of the hydrogeological situation. These approaches include the auxiliary method and the well-constrained method [48,49]. The auxiliary method requires a pre-calculated catalog of the field model prospected. By using this method, the hydrogeological context composed of a gneiss-weathering sequence is generally described by a C-R-type electrical model [49]. While shale and sandstone succession tie with the C-R-C type electrical model [48]. By considering these assumptions, it can be argued that the facies model of C-R-C-R-C (U1, U2, U3, U4, and U5) obtained from the records of this study describes sedimentary covers. This model seems to be in good agreement with borehole information (Figure 7) and reveals a relatively deep sub-outcrop hydrogeology composed of clastic and covered-fault aquifer systems. None of the wells drilled during the project underline the lithological nature of facies U1 (Table 3). However, the presence of a large proportion of mud in the sandstone level drilled in the borehole W-01 (ER-01, Figure 6) suggests that these facies possibly describe a mudstone sequence. Further, comparing ER results and well data allows us to discover that sedimentary buttes of the Sanaga Basin contain two potential hydrogeological systems. The first is represented by discrete levels of shallow (30-40m, 60-80m) and deep buried (>110m) water-bearing sand giving a yield of 3-5 m3/hr. The second system is found in areas where the deep sedimentary section (sandstone complex) is fractured and faulted. Detailed reconstruction of topographic anomalies (Figure 1a) of the study area shows that the ground pattern of this system ties with 030-0900 fault trends and describes a basement fabric having a wide range of tensional faults parallel to the Cameroon Volcanic Line (Figure1c). Within the littoral and marine regions, these features are not continuous but consist of 125-1650 fault segments parallel to the Equatorial Atlantic rift axis [50,51,52,53,54]. Also, when considering the age of igneous and volcanic provinces reported along the Central African regions [55,56,57], our observation is that this segmentation (Figure 1) developed in response to the Pan-African collision between the Brazilian and Oubangui-Chari craton.

6. Conclusions

Alluviums and unconsolidated sediments regularly cause problems of slumping and drilling failure when their presence in the crystalline bedrock is unknown in advance. Geophysical models of the C-R-C-R-C type recorded over 150 m depth and on five surface features of more than 50 km apart show that sedimentary deposits with a mud bedrock are scattered within the Sanaga Valley fault system. These deposits have a deep hydrogeological potential associated with sand channels which can be useful for the water supply of local populations and industries of the study area.