The thickness of the fractured horizon (fractured-weathered horizon) is the intermediate layer between the sound substrate and the weathered material ( Figure 3 ). It is characterized by the presence of fractures in sound rock, the density of which decreases with depth [12] .

The fractured horizon located beneath the base of the weathered zone is where most of the water inflows observed in a borehole occur. The fractured horizon can be determined by plotting the linear flow rate on an arithmetic graph as a function of the depth of the borehole beneath the weathered zone. This is an effective graphical method for identifying the thickness of the highly fractured bedrock horizon [3] and [13] . The linear flow rate represents the flow rate per meter of drilling depth below the base of the weathered rock.

Following the same principle of mapping the thickness of the fractured horizon, the method consisted of calculating the cumulative percentage of linear flow using the following formulas [14] :

$q_i\left(l_i\right)=Q_i/l_{i}i=1,n$(1)

where $Q_i$ is the instantaneous flow rate obtained at the end of drilling (m³/h) at borehole $i$; $l_i$ is the depth of borehole i below the base of the weathered zone (m); $q_i\left(l_i\right)$ is the linear flow rate (m³/h/m).

$P_q\left(L\right)={\displaystyle {\sum }_{l=Min}^L{q}_i\left(l\right)/{\displaystyle {\sum }_{l=Min}^{l=Max}{q}_i\left(l\right)}}$ (2)

where:

L: maximum depth (m) given for drilling below the base of the weathered rock;

Pq(L): cumulative percentage of linear flow rate (%) obtained with the sample of boreholes whose depth is less than L.

It should also be noted that the cumulative percentage of the number of boreholes is also determined. This will enable us to identify the shallowest boreholes whose total linear flow rate will intersect the base of the weathered rock.

After the previous step, a graphical representation of the cumulative percentages (linear flow rate and number of boreholes) as a function of the depth of the borehole below the base of the weathered rock was plotted on an arithmetic graph ( Figure 4 ). Analysis of this graph made it possible to deduce the linear flow rates and flow rates corresponding to the thicknesses of the bedrock read on the same graph [13] .

The thicknesses of the base were determined by identifying different slopes observed on the graph. The intersection of the segments of the different slopes is read on the axis of total drilling depths below the base of the weathered rock.

The linear flow rate (flow rate of the useful fractured medium relative to the thickness of the useful fractured horizon) and the flow rate of the useful fractured medium are calculated as follows:

$q_M\left(L\right)={\displaystyle {\sum }_{l=Min}^L{q}_i\left(l\right)/j\left(L\right)}$(3)

where:

$q_M\left(L\right)$: average linear flow rate (m³/h/m) calculated for the depth range L (m) located below the elevation using the slope method; $j\left(L\right)$: corresponding number of boreholes.

$Q_M\left(L\right)=q_M\left(L\right)\times L$(4)

where:

$Q_M\left(L\right)$: flow rate of the useful fractured horizon, the product of the previous parameter and the thickness defined by the slope method.

After estimating the various useful depths and their corresponding flow rates, a conceptual model was developed. This model highlighted the layout of the various thicknesses of the useful horizons determined beneath the base of the weathered rock using the slope method.

Validation consisted of verifying the reliability of the method used to map the thicknesses of useful fractured horizons and their productivity. Logs and technical drilling equipment were used to validate the model. Surfer 11 software was used to implement the conceptual model and validate it.

Figure 5 shows the vertical distribution of the linear flow rate for 414 boreholes drilled in the study area. The linear flow rates range from 3.4 × 10⁻³ to 4.33 m³/h/m, with an average of 0.19 m³/h/m. The depths of the boreholes below the base of the weathered rock range from 2 to 90.5 m, with an average of 26.83 m. The linear flow rate decreases gradually with the depth of the boreholes below the base of the weathered rock. In addition, we also note that the best linear flow rates are observed around the first 50 meters below the base of the weathered rock.

The parameters sought were determined using the slope method. This method made it possible to identify different horizon thicknesses, including their corresponding flow rates and linear flow rates. These classes of fissured horizon thickness were defined by the presence of slopes observed on the curve of cumulative percentages of linear flow rates as a function of drilling depth below the base of the weathered rock ( Figure 6 ). It revealed the existence of four (4) distinct zones:

The results of the slope method indicate that the desired thickness of the fractured horizon is the sum of the first two zones. Consequently, the thickness of the fractured horizon is 35 m below the weathered rock, with a linear flow rate of 0.23 m³/h/m and a flow rate of 8.15 m³/h. Furthermore, this thickness corresponds to 76% of shallower boreholes that pass through the weathered rock.

In order to characterize useful and productive fissured horizons, the synthesis of the results of the slope method made it possible to obtain a three-dimensional (3D) conceptual hydrogeological model. Under the weathered rock, the conceptual models identify four (4) fractured horizons based on fracture density. Starting from the base of the weathered rock at the bedrock, we have:

This is the point of contact between the alteration zones and the granite massif. In our case, it represents an area of uncertainty in determining the depth of the alteration zone wall or the roof of the healthy bedrock. Figure 7(a) revealed that the alteration zone-granite interface is observable in the south at depths of 250 m above sea level in Téhini, 290 m above sea level in Trikongo, 400 m above sea level in Garankodouo, and in the east at around 330 m above sea level between the towns of Yolonkora and Doropo. The depth of the alteration-granite interface can also be identified at an altitude of around 340 m north of Tchamino and at an altitude of 220 m west of Téhini ( Figure 7(b) ). The gradient method has shown that this zone is productive, with an estimated flow rate of 4.27 m³/h for a thickness of 7 m.

This horizon is located below the zone of uncertainty. The conceptual model showed that its roof is located to the south at altitudes of 240 m in Téhini, 280 m in Trikongo, 390 m in Garankodouo, and to the east, it is around 310 m between the towns of Yolonkora and Doropo ( Figure 7(a) ). The roof of this horizon is also located at 200 m west of Téhini and 330 m north of Tchamino ( Figure 7(b) ).

This is the lower part of the highly fissured horizon. It is therefore the extension of the highly fissured horizons at depth. The 3D model showed that in the east, the roof of this horizon is at a depth of about 270 m between the localities of Yolonkora and Doropo, and in the south, it is at an altitude of 220 m in Téhini, 250 m in Trikongo, and 350 m in Garankodouo ( Figure 7(a) ). To the north, the roof is at an altitude of 275 m in Tchamino and 200 m west of Téhini ( Figure 7(b) ).

It is characterized by a very slightly fractured granite substrate. These fractures are the result of tectonic phenomena that have affected the sound rock. The lack of drilling data at this depth prevents us from determining whether the flow rate is constant or decreasing. In other words, we cannot reliably estimate the productivity of this horizon.

After analyzing the 3D conceptual model, drilling productivity is calculated as the cumulative estimated flow rates of the various horizons identified using the slope method. To verify the performance of the catchment structures, validation is performed by comparing the drilling logs with the conceptual model.

The total depths of the boreholes range from 262.25 m above sea level at Yolonkora to 321.9 m above sea level at Garankodouo ( Figure 8 ). In the bedrock region, most of the water inflows identified in the boreholes are observed beneath the weathered layers. Given the depths of the roof and wall of the fractured horizon, the Garankodouo, Yolonkora ( Figure 8(a) ) and Tchamino ( Figure 8(b) ) boreholes intersect these down to the sound bedrock. The Trikongo borehole, however, only intersects the first two horizons without reaching the slightly fractured horizon ( Figure 8(a) ). All water inflows from the boreholes are observed in the strongly and slightly fractured horizons of the model. This confirms the presence of fractured horizons at these depths in our study area.

However, the flow rates of some boreholes do not match those estimated using the slope method. These are the locations of Garankodouo, Yolonkora, and Tchamino. They obtained flow rates of 1.5 m³/h, 2.05 m³/h, and 1 m³/h, respectively. These low flow rates could be due to a lack of interconnections between fractures or the presence of dry fractures. These flow rates are well below 8.15 m³/h, which is the threshold flow rate for a borehole located in the useful fractured horizon. In contrast, the Trikongo and Gnondo Coté boreholes recorded flow rates of 12 m³/h and 9.9 m³/h, respectively. These are in line with the flow rates estimated using the slope method because the boreholes in these locations obtained 5 and 4 water inflows, respectively ( Figure 8(b) ). Unlike the latter boreholes, the Yolonkora locality benefited from a low-flow catchment structure (2.05 m³/h) with five water inflows. This low flow rate can be explained by the fact that the borehole probably intersected one or more wet fractures, suggesting that water is being collected from them.