Assessment of groundwater quality with analytic hierarchy process, Boolean logic and clustering analysis using GIS platform in the Kebili’s complex terminal groundwater, SW Tunisia

• Published in: Environmental earth sciences 2022-09, Vol.81 (17), Article 419
• Main Authors: Ben Brahim, Fatma, Boughariou, Emna, Hajji, Soumaya, Bouri, Salem
• Format: Article
• Language: English
• Subjects: Agricultural development; Analytic hierarchy process; Aquifers; Arid environments; Arid zones; Aridity; Biogeosciences; Boolean; Boolean algebra; Chemical elements; Cluster analysis; Clustering; Data analysis; Data processing; Decision analysis; Earth and Environmental Science; Earth Sciences; Elements; Environmental Science and Engineering; Evaporation; Geochemistry; Geographic information systems; Geographical information systems; Geology; Graphics; Groundwater; Groundwater irrigation; Groundwater management; Groundwater quality; Groundwater resources; Hydrochemicals; Hydrology/Water Resources; Ion exchange; Irrigation; Irrigation water; Irrigation water quality; Logic; Mathematical analysis; Mineralization; Modelling; Multiple criterion; Original Article; Performance evaluation; Quality assessment; Semi arid environments; Semiarid environments; Soil quality; Soil water; Soils; Spatial analysis; Spatial distribution; Surveys; Sustainability; Terrestrial Pollution; Water quality; Water resources
• ISSN: 1866-6280; 1866-6299
• DOI: 10.1007/s12665-022-10541-3

Groundwater resources in arid and semi-arid environments are of limited availability and variable quality, impacting hence their agricultural development. In this work, an attempt to provide a groundwater management context focusing on irrigation groundwater sustainability and surveys was conducted in Kebili’s complex terminal aquifer. The analytical physicochemical data and several hazard groups of irrigation water quality that have negative impacts on the soil quality and cropping performance were evaluated and then processed with the spatial multi-criteria decision analysis (MCDA) models, Hierarchical Classification, and simulations graphics. The evaluation of the hydrochemical characteristics indicated that the increase of groundwater mineralization is governed by the processes of evaporation and the reverse ion exchange effect throughout the water–rock interactions. The final irrigation water quality index (IWQI) maps, obtained from the GIS-based on weighted overlay index (WO v I), analytical hierarchy process (AHP) and Boolean logic (BL) models, produce different findings in both mapping and percentages of the qualitative classes. Indeed, it reveals three classes of suitability (high, medium, and low) with WO v I and AHP. Whereas, Boolean Logic model shows four classes (good, permissible, doubtful, and unsuitable). Additionally, the spatial distribution of the cluster C2 obtained from the hierarchical cluster analysis (HCA) characterized by high levels in all chemical elements globally coincides with the unsuitable areas for irrigation purposes. Moreover, the validation using the receiver operating characteristic (ROC) curve analysis shows an area under the ROC curve value of 81.6% for BL, which is very close to AHP model (81.1%), and both of these are more accurate than WOvI model (AUC = 79.8%). This approves the prediction accuracy of the Boolean integration data processing in assessing and modeling suitability of irrigation water in the study area. In this regard, obtained results may be considered as an appropriate tool for agricultural planning as well as to guide decision‐makers to choose areas where irrigation can be suitable without damaging both soil and water quality.