Characterization of land use dynamics from 1997 to 2017 in the baya river basin (East of the Ivory Coast)

 Corresponding author: N’GUETTIA Kobenan Germain. Copyright © 2025 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution License 4.0. Characterization of land use dynamics from 1997 to 2017 in the baya river basin (East of the Ivory Coast) N’GUETTIA Kobenan Germain 1, *, Mangoua Oi Mangoua Jules 2, Diallo Seydou 1 and et Gone Droh Lanciné 1 1 Géosciences and Environnent Laboratory (LGE), UFR Environmental Sciences and Management (SGE), Nangui Abrogoua University (UNA), 02 BP 801 Abidjan 02, Ivory Coast. 2 UFR Environment, Jean Lorougnon Guedé University (UJLoG), Daloa, BP 150 Daloa, Ivory Coast. GSC Advanced Research and Reviews, 2025, 25(02), 016-024 Publication history: Received on 25 September 2025; revised on 02 November 2025; accepted on 04 November 2025 Article DOI: https://doi.org/10.30574/gscarr.2025.25.2.0334 Abstract The landscapes of the baya watershed have evolved under the direct and indirect effect of climatic and anthropogenic factors during episodes of drought and the advent of arboriculture as well as the agrarian reforms that have affected the country since the early 1980. in this context, this study assesses the dynamics of land use over the entire watershed (6325.48 km2) between 1979 and 2017 using satellite data (Landsat images). The methodology relies on the use of remote sensing and GIS to analyze the dynamics of the major land cover classes in the study area. These classes concern agricultural plots, natural vegetation, bare soils, water surfaces and urbanized areas. The supervised classification approach of Landsat TM satellite images from 1997, 2007 and Landsat 8 from 2017 made it possible to characterize the vegetation cover with cartographic accuracies of 98% and 97% respectively. Following this treatment, it appears that forest (44.97%) and fallow land (33.41%), which were the most represented vegetation types in the basin in 2007, experienced a significant regression between 2007 and 2017 in favor of crop classes (42.58%) and bare soils or urban areas (13.15%). The regression of plant formations is due to rapid urbanization and the expansion of agricultural areas, in particular the search for fertile land with still traditional cultivation techniques Keywords : Caractérisation; Dynamics of land use; Baya watershed; Eastern Côte d'Ivoire 1. Introduction The Baya watershed is one of the major basins located in eastern Côte d'Ivoire. Due to its location and the excellence of its land suitable for commercial agriculture (Judge, 2019), it experienced a significant demographic expansion between the 1980s and 2000s, which led to landscape transformations (Mangoua, 2013). The Baya watershed, dedicated to agriculture, is subject to various forms of attack due to the interaction between various climatic conditions and the human impact on vegetation and soil cover (N’Guettia, 2023). The main objective of this study is to analyze the évolution of land use in the Baya watershed during the years 1979 to 2007 and 2017. Knowledge of land use is very important information for the characterization of a region and for the development of plans for better land management Biga et al. (2020). It provides an illustration of the continuous confrontation between society and its environment (Belghiti et al., 2014), expressing at a given moment a state of these relations which involve structures changing in time and space. It involves the use of remote sensing data, together with high-resolution satellite imagery (Diallo et al., 2025), to finely track the dynamics of land use and, subsequently, environmental pressures. In order to meet the objective of this work and thus allow a good under standing of the evolution of the landscape in the region, we will first present the studied area and the methodology applied before proceeding to the analysis of the dynamics of the land use through the spatial changes and the observed trends. GSC Advanced Research and Reviews, 2025, 25(02), 016-024 17 2. Materials and Methods 2.1. Study area The Baya watershed is located in eastern Côte d'Ivoire between longitudes 2°38' and 3°33' W and latitudes 6°35' and 8°26' N. It covers an area of approximately 6,324 km2 (Figure 1). The population of the basin is estimated at 267,263 inhabitants with an increase of 2.8% of the population (INS, 2014). From a soil point of view, this basin is covered with ferralitic soils occupied by cash and export crops (coffee, cocoa, cashew) and food crops (Yves et al., 1995). The main geological formations encountered can be grouped into three large lithological ensembles (Youan Ta et al., 2015). A Tarkwaian and volcano-sedimentary ensemble consisting mainly of Shale, Amphibolite and Metadolerite. From a hydrogéologiques point of view, there are two types of aquifers in the study area. These are the aquifers of alterites, the aquifers of cracks and fractures. This hydrogeological framework means that water resources are mainly used for drinking water supply and for hydroelectric production, hence a basin characterized mainly by rainfed agriculture subject to climatic hazards (http://www.water.gov.ma/). Climate governs the spatial and temporal distribution of water resources (Figure 1) and thus influences land use, particularly in agricultural and forestry areas. It is distinguished on the basin by a wet season, from October to April, during which 90% of rainy episodes take place, and a dry season, from May to September (Mangoua, 2013). 2.2. Study data Figure 1 Location of the study area The methodology used in this work is based on the use of satellite data and field observations. These data were then integrated and analyzed using remote sensing tools and Geographic Information Systems (GIS) to map the study area. Two scenes covering the basin. These are the scenes p195r54 and the scenes p195r55 was produced after the mosaicing of the scenes p195r54 covering the northern part of the basin and the scenes p195r55 covering the southern part of the study area. GSC Advanced Research and Reviews, 2025, 25(02), 016-024 18 2.3. Methods 2.3.1. Unsupervised and supervised classifications The unsupervised classification was an intermediate step and made it possible to better and more easily interpret the radiometric reality to the thematic reality (Kone et al., 2007 ; Diallo et al., 2019). Eight thematic classes were defined at the launch of the unsupervised classification. From the results of the unsupervised classification and the type of land use on the ground was made on the basis of the training plots. The supervised classification algorithm chosen is the “maximum likelihood” considered as one of the most efficient algorithms in the production of thematic maps in the field of land use (Diallo et al., 2025). Indeed, the classification supervised by the maximum likelihood was used, because it allows a good distinction of the themes of land use. This method is based on the probability that a pixel will belong to a given class (Djohy et al., 2016). The colored composition ETM+7(R), ETM+5(V), ETM+3(B) was used for the classification. 2.3.2. Cartographic precision The overall accuracy of the classification or Kappa Coefficient (CK) is valid when this value approaches 100%. The CK translates the ratio between the total number of well-classified pixels and the total number of pixels in an image (Benmessaoud et al., 2008). It takes into account errors of omission. These two parameters make it possible to better appreciate the classification carried out. 2.3.3. Land use between 2007 and 2017 The balances of the evolution of the land-use units were calculated using the software Envi 4.7; i.e. V.2007 and V.2017, the proportion of a land-use unit respectively in 2007 and 2017. The variation TV is the area of a land use unit between two dates is given by equation 2 and FIG. 2 summarizes the supervised classification. TV = (𝑽𝟏−𝑽𝟎) 𝑽𝟎 ∗100 TV: Rate of vegetation change between the two observation dates ; V0: the proportion of the land cover class taken in the initial state ; V1: the proportion of the land cover class taken in the final state. Figure 2 Simplified Methodological Scheme of Supervised Classification (N’go, 2015) GSC Advanced Research and Reviews, 2025, 25(02), 016-024 19 3. Results 3.1. State of land use in 1997, 2007 and 2017 in the Baya river basin Land use maps from 1997 to 2017 were used to characterize changes in vegetation cover and anthropogenic activities in the Baya watershed. 3.2. Discrimination of types of land use by colored compositions The color composition ETM+5/ETM+4/ETM+7 (FIG. 32) is the one which has allowed a better discrimination of the types of land use. It was possible to distinguish on this composition, various colorations that highlight the types of land use. The major colors in this image are represented by yellow, black and cyan indicating urban areas and bare soils. Bright magenta indicates wetlands like forest. Also, the mosaics of red-green, green-blue colors indicate crops and fallow land. Finally, black indicates the reservoirs and streams. The vegetation cover (degraded magenta) of the study area is very important, but little varied between 1997 and 2007. Areas of forests (dark magenta), cultivated mosaics (light magenta) and bare cyan habitats or soils can be observed. In 2017, the analysis of the colored composition carried out on the same zone shows a very significant variation in the vegetation. You can see a significant disappearance of the yellow color with its variants, including green yellow, orange yellow. The mosaics of forests, crops or fallow crops observed in 1997 and 2007, have evolved to make way for forest complexes. The cyan themes show the spatial distribution of urban areas and bare soils that progress and are concentrated mainly in large. agglomerations and a proliferation of small hamlets, especially in the Bondoukou and Agnibilékrou regions (Figure 4 (a, b, c). There is also a decrease in wetlands. Crop and/or fallow areas have a light magenta hue. They have taken place gradually with the destruction of forest areas. In all the images, the change in vegetation cover can be observed in the space of 20 years. This degradation affects the different types of vegetation in the Baya watershed. Figure 3 Colored composition of bands 5, 4 and 7 of the Landsat TM images from 1997 (a), Landsat TM+ images from 2007 (b) and Landsat 8 images from 2017 (c) 3.3. Supervised Classification To better discriminate the different types of land use, a directed classification with field observation was carried out. This classification was carried out on the basis of 5 types of land use: forest, crops, fallow land, water and urban areas or bare soils. The confusion matrices (Table 1, 2 and 3) express the pixel rates of a class actually assigned to this class by the maximum similarity classification method. The different rates outside the diagonal reflect those of the poorly classified pixels, that is to say pixels that have been assigned to a class to which they do not belong. GSC Advanced Research and Reviews, 2025, 25(02), 016-024 20 The matrices obtained show overall cartographic accuracies of 98.6% ; 99.12% and 98.47% for Landsat 7 TM+ of 1997, 2007 and Landsat 8 ETM for 2017 with respective KAPPA coefficients of 0.981, 0.988 and 0.980. The values in bold in the diagonal (Table 1, 2 and 3) correspond to each type of land use with well-classified pixel rates. The fallow class has the best accuracy rate in 1997 with 100% pixel well ranked. Next are the forest, bare soil/habitat and crop classes (Table 2). In 2007, the crop, urban or bare soil and water classes had the best mapping accuracy rates of 99.42%, 99.27% and 99.97% respectively. On the other hand, the rate of cartographic accuracy recorded in the forest class, although important, remains the lowest with 97.31%. This low rate records a confusion of 2.65% with the fallow class. With regard to the confusion matrix from the Landsat 8 image of 2017 (Table 3), the best mapping accuracy rates were observed at the fallow class (99.76%), forest class (99.23%), habitat/bare soil class (100%). The culture class displays the lowest rate of well-ranked pixels with (95.57%). This low rate recorded a confusion of 1.02, 1.29 and 1.96% respectively with the forest, water and habitat/bare soil class. The fallow class has the best accuracy rate in 1997 with 100% pixel well ranked. Next are the forest, bare soil/habitat and crop classes (Table 2). In 2007, the crop, urban or bare soil and water classes had the best mapping accuracy rates of 99.42%, 99.27% and 99.97% respectively. On the other hand, the rate of cartographic accuracy recorded in the forest class, although important, remains the lowest with 97.31%. This low rate records a confusion of 2.65% with the fallow class. With regard to the confusion matrix from the Landsat 8 image of 2017 (Table 3), the best mapping accuracy rates were observed at the fallow class (99.76%), forest class (99.23%), habitat/bare soil class (100%). The culture class displays the lowest rate of well-ranked pixels with (95.57%). This low rate recorded a confusion of 1.02, 1.29 and 1.96% respectively with the forest, water and habitat/bare soil class. Table 1 Confusion Matrix of the Supervised Maximum Likelihood Classification from the Landsat TM (1997) image of the Baya Basin Theme (in %) Waters Fallow land Forests Cultures Habitat/ Sol nu Waters 98.11 0.00 0.20 0.00 0.12 Fallow land 0.00 100 0.07 0.00 0.00 Forests 1.05 0.00 99.59 0.11 0.00 Cultures 0.00 0.00 0.14 97.47 0.84 Habitat/ Sol nu 0.84 0.00 0.00 1.26 98.95 Total 100 100 100 100 100 Overall accuracy = (12471/12648) 98.6 %, Kappa coefficient = 0.9818 Table 2 Confusion Matrix of the Supervised Maximum Likelihood Classification from the Landsat TM (2007) image of the Baya Basin Theme (in %) Waters Fallow land Forests Cultures Habitat/ Sol nu Waters 99.66 0.00 0.00 0.13 0.00 Fallow land 0.22 99.42 2.65 0.00 0.00 Forests 0.00 0.16 97.31 0.03 0.03 Cultures 0.11 0.00 0.00 99.27 0.00 Habitat/ Sol nu 000 0.42 0.04 0.56 99.97 Total 100 100 100 100 100 Overall accuracy = 2007 (11377/11478) 99.12%; Kappa coefficient = 0.98 GSC Advanced Research and Reviews, 2025, 25(02), 016-024 21 Table 3 Confusion matrix of the supervised classification by maximum likelihood from Landsat 8 (2017) image of the Baya basin. Theme (in %) Waters Fallow land Forests Cultures Habitat/ Sol nu Waters 98.05 0.00 0.08 1.29 0.00 Fallow land 0.00 99.76 0.00 0.16 0.00 Forests 1.38 0.00 99.23 1.02 0.00 Cultures 0.00 0.00 0.62 95.57 0.00 Habitat/ Sol nu 0.57 0.24 0.08 1.96 100 Total 100 100 100 100 100 Overall accuracy = 2017 (9470/9617) 98.47%; Kappa coefficient = 0.9802 3.4. Assessment of land use between 1997, 2007 and 2017 Compared to 2007 and 1997, 2017 (Figure 4) saw a wide variety of vegetation units. The evolution of different vegetation cover units between 1997 and 2017. This figure shows a continuous evolution of the bare habitat-soil classes from 1997 to 2017. The culture class, on the other hand, experienced a drastic fall from 1997 to 2007 and a galloping increase from 2007 to 2017.In addition, the forest and fallow classes experienced a reduction in their areas between 2007 and 2017. Proportions of land-use units are reported (Table 4). From 2007 to 2017, the forest increased from 2842.06 hectares to 1386.91 ha, corresponding to a deforestation rate of 21.93%. Crop classes showed an exponential change from 10.28% to 42.62%, i.e. a change rate of 314%. Similarly, urban areas and bare soil increased from 717.68 to 931.83 hectares, an increase of 15.9 per cent. There was also a slight increase in water areas, from 0.02% in 2007 to 4.78% in 2017. The order of predominance of units is defined by secondary forest or gallery (44.97%), fallow (33.41%), urban areas-bare soil (11.34%), crops (10.28%) and water (0.02%). As for the satellite image taken in 2017 in the basin, it shows a significant variation in vegetation cover compared to that taken in 2007 and 1997. The order of predominance of units is as follows : crops (42.58%), fallow forest (22.27%), forest (21.93%), urban areas and bare soils (13.15%) and water (0.07%). Figure 4 Land use in 1997 (a), 2007 (b) and 2017 in the Baya a b c GSC Advanced Research and Reviews, 2025, 25(02), 016-024 22 Table 4 Land cover classes in 1997, 2007 and 2017 Classe State in 1997 State in 2007 State in 2017 Units Area (ha) Proportion (%) Area (ha) Proportion (%) Area (ha) Proportion (%) Waters 8.02 0.13 2.3 0.02 4.71 0.07 Fallow land 2100.78 33.21 2112.96 33.41 1408.86 22.27 Forests 1879.71 29.72 2842.06 44.97 1386.91 21.93 Cultures 1949.68 30.82 650.48 10.28 2693.17 42.58 Habitat/ Sol nu 387.3 6.12 717.68 11.34 831.83 13.15 TOTAL 6325.48 100 632.48 100 6325.48 100 3.5. Summary of land use between 1997 and 2017 The regressive trends of land-use units in the 1997-2017 chronicle concern forest and fallow land with respective proportions -51% and -33%. These differences in area are very significant in the old loop of the cocoa coffee binomial of the basin, circumscribed in the localities of the South from Akobouassué to Tankessé and the new zone of cultivation of the cashew precisely in the localities of Tanda, Gouméré and Tabagne ; however, the differences in progressive areas are of 16%, 314%, and 105% and concern islets of plots which are observed scattered over the whole basin. These are bare habitats/soils, crops and water surfaces respectively. 4. Discussions The classification method supervised by the maximum likelihood algorithm was applied to Landsat 7 satellite images from 1997, 2007 and Landsat 8 satellite images from 2017 in the Baya watershed. The cartographic precision obtained is 98.6% (1997); 99.12% (2007) and 98.47% (2017) with Kappa Coefficients (CK) of 0.981, 0.988 and 0.980 respectively. These values suggest that the classifications performed are reliable and acceptable. Biga et al., (2020) Dialo et al., (2025) interprets the values of the kappa coefficient from 1 to 0.81 as excellent and good. Moreover, these cartographic details are in line with the range of values obtained by some authors, notably Aké (2010) and N’go (2015) with cartographic details of 90% and 83.24% respectively. These results revealed an increase in cultivation areas and a decrease in natural plant formations. Cette régression de la végétation aux profits des zones anthropisées est devenue de plus en plus récurrente. These results are consistent with observations by Koffi (2016) in d'Azagny, southern Côte d'Ivoire, Kpedenou et al. (2016) in southeastern Togo, Millogo et al. (2017) in the province of Bam in Burkina Faso, Mamane et al. (2018) in the total wildlife reserve of Tillabéry in Niger of Biga et al. (2020). The regressive trends of land use units in the 2007-2017 chronicle concern forest and fallow land with a respective proportion of -51%, and -33% to the detriment of crop classes (314%) and urban bare soil (16%). The extension of cultivation areas synonymous with regression of plant formations and fallow land is also supported by the work of certain authors such as Adjonou et al. (2010), Koumoi et al. (2013) and Djohy et al. (2016). For these authors, agricultural activities could be the major causes of vegetation degradation. The same observation was made by Sudhakar et al. (2017) respectively in Benin and Sri Lanka. For the latter, this deforestation is linked to the marked increase in agricultural land and urbanization. In addition, the introduction of arboriculture into the Ivorian economic system, such as rubber, cashew and Tecks, has helped regenerate vegetation in some localities (Djohy et al., 2016). Also, the widespread poultry areas in the area have contributed to a significant reduction in forest areas. Woodland is the most represented vegetation in the basin with an increase of nearly 314% in their area in 10 years. Then, urbanization with its corollary of electrification and subdivision has led to the loss of agricultural land and an increase in bare areas or 16% of the total area of the Baya basin as indicated by Dialo et al. (2025) in the Djiby basin. Water surfaces in the Baya catchment area have increased slightly during this decade. They increased from 0.02% in 2007 to 4.78% in 2017, a rate of change of 205%. This gradual trend in water surfaces would indicate a recovery in rainfall at the basin level. Indeed, a study (Ouattara, 2018) on the regression of the number of rainy days between 1970 and 2000 indicates a positive slope of about 0.38 and 0.18 respectively for the Ferkessédougou and Bondoukou stations. GSC Advanced Research and Reviews, 2025, 25(02), 016-024 23 5. Conclusion At the end of this study, it appears that image processing techniques and GIS have made it possible to characterize the dynamics of land occupation and use between 1997 and 2017. The changes in land use in the Baya basin were highlighted using Landsat TM (2007) and Land SAT 8 (2017) satellite image processing combined with field observations. The results of the dynamics of land use show a regression of the forest (-51%) and fallow land (-33%) on the one hand and an increase in cropping areas (315%), bare habitats and soils (105%) and water surfaces (16%) on the other hand. The decline in plant formations is due to the expansion of agricultural areas, particularly the search for fertile land and still traditional farming techniques, urbanization and demographic expansion. Compliance with ethical standards Disclosure of conflict of interest No conflict of interest to be disclosed. References [1] Adjonou K., Djiwa O., Kombate Y., Kokutse A. D., Kokou K. (2010). Study of the spatial dynamics and structure of Togo's relict dense dry forests: implications for sustainable management of protected areas. International Journal of Biological and Chemical Sciences, 4(1): 168-183. [2] Ake G.E., Kouadio H.B., Dongo K., Dibi B., Kouame F.K., Biemi J. (2010). Application of DRASTIC and SI methods for the study of vulnerability to nitrate pollution (N03-) of the Bonoua slick (South-East Côte d'Ivoire), International Journal of Biological and Chemical Sciences, 4(5) : 1676-1692. [3] Belghiti, L., Chahlaoui, Abdelkader, Bengoumi, D., & El Moustaine, R. (2014). Effect of anthropic activities on the quality of subsoil waters in rural medium in the area of Meknes (Morocco). LARHYSS Journal PISSN 11123680/E-ISSN 2521-9782, (17). [4] Diallo S., Bakayoko S., Kone S.1, Noufe Do Dao A., Tra Bi Z., Kamagate B. (2025). Modeling of landscape dynamics in the Djibi watershed by. J. Mater. Approximately. Sci., 2025, 16(3) : 495-508. [5] Diallo, S., Bi, ZAT, Noufé, D., Dao, A., Kamagaté, B., Effebi, RK, & Seguis, L. (2019). Land use dynamics of the Ivorian basin of the Aghien lagoon. International Journal of Innovation and Applied Studies, 26 (1), 203-217. [6] Biga I., Amani A., Soumana I., Bachir M., Mahamane A. (2020). Spatio-temporal dynamics of land use in the communes of Torodi, Gothèye and Tagazar in the Tillabéry region of Niger. International Journal of Biological and Chemical Sciences 14(3): 949-965. DOI: https://doi.org/10.4314/ijbcs.v14i3.24. [7] Djohy G.L., Totin V.H.S., Kinzo N.E. (2016). Dynamics of land use and evolution of agricultural land in the municipality of Sinende in North Benin. Benin Center for Scientific and Technical Research CBRST Notebooks, Letters, Humanities and Social Sciences ISSN : 1840-703X, Cotonou (Benin), 9:101-121. [8] INS (2014). Institut National de la Statistique, Recensement général de la Population et de l’Habitat. www. ins.ci., : 25-47. [9] Judge, V. (2019). Contribution of machine learning to the modeling and analysis of land use changes, PhD thesis from the University of Bourgogne Franche-Comté, 365 p [10] Koffi K.A.D., 2016. Vegetation dynamics and conservation value of formerly cultivated areas of Azagny National Park (southern Côte d'Ivoire). PhD thesis, UFR Biosciences, Félix Houphouët-Boigny University, Abidjan, 214 p. [11] Koné M, Aman A., Yao ACY (2007). Diachronic monitoring by spatial remote sensing of woody cover in the Sudanian savannah environment in Côte d'Ivoire. Rev. Télédét. 7(1-2-3-4) : 433-446. [12] Koné J., Yéo P.A., Koné Y.N. (2018). Water is a major concern in Côte d'Ivoire. Dispatch Afrobarometer, 218 : 1-8. [13] Koumoi Z., Alassane A., Djangbedja M., Boukpessi T., Kouya A.E. (2013). Spatio-temporal dynamics of land use in central Togo. AHOHO, Revue de Géographie du LARDYMES, 7 (10): 163-172. [14] Ouattara I. (2018). Stochastic modeling of hydroclimatic drought and characterization of the interaction between surface water and groundwater in the Comoé watershed. Doctor's Thesis, University NANGUI ABROGOUA, 262p. GSC Advanced Research and Reviews, 2025, 25(02), 016-024 24 [15] Kpedenou K.D., Boukpessi T., Tanzidani T., Tchamie K. (2016). Quantification des chan gements de l’occupation du sol dans la préfecture de Yoto (Sud Est Togo) à l’aide de l’imagerie satellitaire Landsat. Revue Sciences de l'Environnement, 13 : 137-156. [16] Mangoua M J. (2013). Assessment of the potential and vulnerability of groundwater resources in the aquifers fissures of the Baya watershed (Eastern Côte d'Ivoire). Doctoral thesis. Nangui Abrogoua University, Abidjan (Ivory Coast), 171p. [17] Mamane B., Amadou G., Baragé M., Comby J., Ambouta J.M.K. (2018). Timotemporal spa dynamics of land use in the Tamou Total Wildlife Reserve in a context of climate variability (western Niger). International Journal of Biological and Chemical Sciences, 12(4): 1667-1687.DOI : https://dx.doi.org/10.4314/ijbcs.v12i4.13. [18] Millogo D., Nikiéma A.A., Koulibaly B., Zombré N.P. (2017). Analysis of the evolution of land occupation from aerial photographs of the locality of Loaga in the province of Bam, Burkina Faso. International Journal of Biological and Chemical Sciences, 11(5) : 2133-2143. [19] N’Guettia (2023). Dynamique de l’occupation du sol et qualité physico-chimique des ressources en eau souterraine du bassin versant de la Baya (Est Côte d’Ivoire), 150P. [20] N’go Y.A. (2015). Hydrology and dynamics of land surface conditions in south-western Côte d'Ivoire: impacts and drivers of degradation. State Doctorate in Natural Sciences. Nangui Abrogoua University in Environmental Science and Management, 125 p. [21] Sudhakar R.C., Manaswini G., Jha C.S. Diwakar P.G., Dadhwal V.K. (2017). Development of National Database on Long-term Deforestation in Sri Lanka, Journal of the Indian Society of Remote Sensing, 45:825-836. DOI 10.1007/s12524-016-0636-8. [22] Youan Ta M., Yao K.A.F., Bakar D., De Lasmz. O., Lasm T., Adja M.G., Kouakou S., Onetie Z.O., Jourda J.P.R., Biemi J. (2015). Mapping of potential areas for the implantation of large-flow boreholes in a crack environment by multicriteria analysis : Case of the department of Oumé (west-central Côte d'Ivoire), Larhyss Journal, 23 :155181. [23] Yves S., Delor C., Zeade Z., Kone Y., Yao B.D., Vidal M., Diaby I., Konan G., Irié D.B., N’da D., Dommanget A., Cautru J.P., Guerrot C., Chiron J.C. (1995). Geological Map of Ivory Coast at 1/200 000 ; Agnibilekrou leaf. Memorandum from the Directorate of Mines and Geology, n°8, Abidjan, Ivory Coast, 19p.