Located in the south of Côte d’Ivoire, the study area lies on the northern edge of the Ebrié Lagoon, between 4˚15'57.6'' and 3˚45'36'' West and 5˚17'20.4'' and 5˚27'57.6'' North ( Figure 1 ). It belongs to the Autonomous District of Abidjan, which comprises ten (10) municipalities and four (04) sub-prefectures covering an area of around 1.160 km² and with an estimated population of 29 million [9] .

It has three main lagoons. The Ebrié lagoon, the largest, is a vast expanse of navigable water covering 566 km² and connected to the Atlantic Ocean via the Vridi canal. It receives water from the Comoé river, the Agnéby and Mé rivers [10] . The Potou and Aghien lagoons are small internal lagoons running north-south. They flow into the Ebrié Lagoon [11] ( Figure 2 ).

The most important rivers are [12] :

The geological context of the study area generally belongs to the sedimentary basin. The sedimentary formations in this basin are grouped into two main formations: the Continental Terminal in the northern part of the lagoon fault and the Quaternary to the south of this fault [14] . The Continental Terminal or Tertiary is made up of sandy-clay sediments, gravels and conglomerates, while the Quaternary is composed of marsh muds and sand ( Figure 3 ).

The hydrogeological overview of the Ivorian coastal sedimentary basin at the northern boundary of the Ebrié Lagoon reveals the existence of a Continental Terminal aquifer. It comprises four horizons superimposed from top to bottom [15] :

According to the same authors, horizons 3 and 4 make up the bulk of the Continental Terminal aquifer. Generally unconfined, lenticular clay banks located at the top of horizon 3 make the aquifer captive in some areas. These horizons form a single aquifer when these clay banks are absent. Horizon 2 is a sterile aquifer and horizon 1 contains the water table at the base of the Tertiary period.

Exploitation of the Continental Terminal aquifer, commonly known as the “Abidjan aquifer” accounts for 68% of Côte d’Ivoire’s drinking water supply. This water table also feeds the Maestrichtian one at its base [16] because of the thickness of the clay and ferruginous sandstone separating the two aquifers.

The various processes involved in the mineralization of groundwater were determined by studying the spatial and temporal evolution of electrical conductivity, piezometry and correlations between the major elements. Water mineralization is often assessed by measuring electrical conductivity [4] . This is because most of the dissolved matter in water is in the form of electrically charged ions [8] .

In this study, electrical conductivity was used as a marker of the mineralization of the Continental Terminal groundwater.

Electrical conductivity measurement campaign was carried out using a sound piezometric probe at twenty (20) piezometers ( Figure 4 ) covering the study area. It took place from 02 to 05 March 2022, a month that corresponds to the dry season.

The piezometric probe used is a Solinst TLC Meter model 107, 300 m long and mounted on a coil. It is ideal for profiling conductivity and temperature in piezometers for saltwater intrusion studies and a general indication of contamination levels. Conductivity and temperature readings are displayed on a rotating LCD screen on the face of the coil The depth of the water and the depth of the readings displayed are read off the tape. Conductivity accuracy is 5% of reading or 100 μS (whichever is greater). A light and buzzer are briefly activated when the zero point of the probe enters the water.

The GARMIN Map64s GPS (Global Positioning System) is used to determine the coordinates (longitude, latitude and altitude) of the measurement points (piezometers).

The probe is lowered progressively into the piezometer (vertical profiling) and measurements are taken in steps of approximately two (02) meters. At each depth, the electrical conductivity and the temperature values are displayed on the screen of the coil.

The electrical conductivity of a liquid depends largely on its temperature. That’s why these two parameters were measured simultaneously. While standard temperatures such as 20˚C or 25˚C are used for calibrations with conductivity standards, conductivity measurements are generally carried out at different temperatures. Identifying water types from conductivity therefore requires recalculating the conductivity measured at a standard temperature, for example 25˚C. This correction is based on the following equation [17] :

$E{C}_{25°C}=E{C}_T\times F_{25}$(1)

where: $E{C}_T$ = the measured electrical conductivity et $F_{25}$ = temperature correction factor.

A classification of electrical conductivity has been established in order to define water types. This is the Milot and Renaud (2020) classification ( Table 1 ).

Water mineralization is assessed by measuring electrical conductivity. It is a good indicator of groundwater mineralization. This is because most of the dissolved matter in water is in the form of electrically charged ions [8] .

In this study, Continental Terminal water mineralization is assessed by studying the spatial evolution of the electrical conductivity of water from this aquifer. Electrical conductivity data are acquired at each piezometer and along with six (06) aligned profiles (horizontal profiling) ( Table 2 ). The horizontal profiles were defined according to the alignment and orientation of the piezometers.

This study focuses on the horizontal spatialization of electrical conductivities in the surface part of the water table and on the distribution of electrical conductivities as a function of depth (vertical profiling).

Data processing was carried out with software. This software is chosen in accordance with the different objectives set. These are:

Table 3 shows low standard deviations overall. This reflects a high concentration of measured temperatures around their mean values. In fact, the almost zero standard deviation values indicate that the temperatures measured as a whole are similar. Also, the standard deviations of approximately one (01) obtained only at the level of three piezometers (PZ-C2D8, PZ-C2D9 and PZ-C2D12), means that the temperatures measured are distributed more or less uniformly over a range of one (01) unit.

Electrical conductivity measurements are analyzed according to their spatial distribution (mapping) and their behaviour at depth (electrical conductivity logs).

The electrical conductivities measured in the first few meters of the piezometers are all low (less than 200 µS/cm). These values are typical of freshwater. However, their spatial distribution highlights three classes of electrical conductivity ( Figure 5 ).

The first class, the most important in terms of surface area, is that of water with very low electrical conductivity (between 20 and 80 µS/cm). The second class is characterized by electrical conductivities ranging from 80 to 120 µS/cm and is located along the Ebrié lagoon to the west and between the Banco and Gbangbo rivers. Finally, the third class, comprising water from piezometers PT1, PZ_C2D 13 and PZ 19, is characterized by electrical conductivity values of between 120 and 200 µS/cm. This class is contained within the previous class. The piezometers in these last two classes are located downstream of surface water flows towards the Ebrié Lagoon (PT1) and Banco Bay (PZ_C2D 13, PZ_C2D 14 and PZ 19).

The vertical variation of electrical conductivity was characterized using conductivity logs. The results are summarized in the appendix. These were analyzed using (6) aligned piezometer profiles ( Table 4 ).

It comprises piezometers PT1, PZ_C2D 26 and PZ_C2D 21. The electrical conductivity logs are shown in the appendix. Its values range from 30 to 1900 µS/cm. Piezometers PT1 and PZ_C2D 21 have very low electrical conductivity ranging from 30 to 200 µS/cm (PT1: 30 < EC < 74 µS/cm and PZ_C2D 21: 31 < EC < 250 µS/cm). It’s the same for the electrical conductivities of the CT sands at PZ_C2D 26, which range from 58 < EC < 66 µS/cm. However, there is a sudden jump in conductivity from 140 m corresponding to the depth of the bedrock. These electrical conductivities reach 1900 µS/cm, indicating the presence of brackish water.

It includes piezometers PZ_C2D12, PZ_C2D13, PZ_C2D14, PZ19 and PZ_C2D47. The electrical conductivities range from 34 to 1600 µS/cm and the logs of electrical conductivity are shown in the appendix.

The electrical conductivities of the water in the CT sands define two classes of water. “Fresh water”, relatively low (CECT < 200 µS/cm), is found in all the piezometers except PZ19. Slightly higher conductivity values (200 < CECT < 1000 µS/cm) corresponding to “mineralized water” class were measured in piezometers PZ_C2D13, PZ19, PZ_C2D14, PZ_C2D12 and PZ_C2D47. This class is marked by significant and sudden variations of conductivity at 23 m and 80 m, 40 m, 80 m, 100 m and 138 m respectively.

In the basement, brackish water is found in piezometers PZC2D12 and PZ_C2D13, at 120 m and 100 m respectively. These depths correspond to those of the basement.

Three piezometers (PZ_C2D 11, PZ_C2D 9 and PZ7_2010) make up this profile. The electrical conductivity logs are shown in the appendix. Piezometers PZ_C2D 11 and PZ7_2010 record very low electrical conductivity values ranging from 33 to 200 µS/cm with depth. As for piezometer PZ_C2D 9, three (03) electrical conductivity classes are defined. These are:

To sum up, the water in the CT for this profile is made up of fresh water, mineralized water and brackish and saline water.

It comprises piezometers PZ_C2D 10, PZ1 and PZ7_2010. The electrical conductivity logs are shown in the appendix. The water in the CT is characterized by low electrical conductivity (29 to 94 µS/cm) and is classified as freshwater.

It consists of piezometers PZ_C2D5, PZ_C2D10 and PZ_C2D2. The electrical conductivity logs are shown in the appendix. The water in the CT has very low electrical conductivity (30 - 62 µS/cm) to a depth of around 120 m. This electrical conductivity is characteristic of ‘fresh’ water. However, ‘weakly and moderately mineralized’ water, with electrical conductivity values of less than 1000 µS/cm, is encountered in PZ_C2D 5 and PZ_C2D2.

Brackish water was also detected in piezometer PZ_C2D5 at a depth of 126 m. At this depth, the bedrock is found. Relatively high electrical conductivity values (1400 µS/cm) were measured from this depth.

This profile consists of piezometers PZ_C2D4, PZ_C2D6, PZ_C2D7 and PZ_C2D8. The electrical conductivity logs are shown in the appendix. Low electrical conductivities were measured at depths of 32 m, 98 m, 47 m and 116 m. These depths correspond to those of the CT. We therefore conclude that at the level of this profile, the water in the CT is ‘fresh water’.

Below, in the basement, the electrical conductivity values of the water indicate ‘moderately mineralized water’ and ‘brackish water’.

Analysis of all the electrical conductivity logs indicates high electrical conductivity values (greater than 1000 µS/cm) in seven (07) piezometers ( Table 5 ). These are piezometers PZ_C2D 26 (profile 1), PZ_C2D 12 and PZ_C2D 13 (profile 2), PZ_C2D 9 (profile 3), PZ_C2D 5 (profile 5), PZ_C2D 6 and PZ_C2D 7 (profile 6).

The CT waters have not yet been affected by the marine intrusion, the effects of which are perceptible at basement level with the presence of brackish waters. Only the CT waters at piezometer PZ_C2D 9 are brackish and saline.