Assessment of groundwater potential zones using remote sensing and GIS-based fuzzy analytical hierarchy process (F-AHP) in Mpwapwa District, Dodoma, Tanzania

• Published in: Geosystems and geoenvironment 2024-02, Vol.3 (1), p.100232, Article 100232
• Main Authors: Ally, Ally Mgelwa, Yan, Jianguo, Bennett, George, Lyimo, Neema Nicodemus, Mayunga, Selassie David
• Format: Article
• Language: English
• Subjects: Fuzzy analytic hierarchy process; Geographic information system; Groundwater potential zones; Mpwapwa District; Remote sensing; Semi-arid region
• ISSN: 2772-8838; 2772-8838
• DOI: 10.1016/j.geogeo.2023.100232

•Fuzzy-AHP was used to delineate groundwater potential zones.•Structure and lithology control groundwater potentiality.•Existing boreholes were used for validation. Groundwater is a very important resource for socio-economic development. The uncertainty of where potential groundwater resources is located often causes some groundwater development projects to fail. It is common for water resources development projects hitting dry wells after heavy investments of resources. In Mpwapwa District, borehole drilling locations are uncertain, determined by trial-and-error techniques based on geophysical survey methods that involve the study of the behaviour of rock and soil types in specific geological locations. To reduce such uncertainty, this study used remote sensing and GIS-based Fuzzy Analytical Hierarchical Process (F-AHP) to simulate groundwater potential zones (GWPZ) in Mpwapwa District, Dodoma region, Tanzania. The F-AHP model was used to reclassify, weight, and rank various thematic maps, including lithology, soil types, drainage density, lineament, magnetic intensity, slope and elevation. The overall GWPZ map was created by combining the seven (7) ranking thematic map layers in a GIS environment. The resulting GWPZ map that was then validated using two methods: overlaying method and area under the curve (AUC) method. The resulting GWPZ map shows that 19%, 31%, 28% and 22% of the area are classified as very good, good, moderate, poor and very poor zones, respectively. The accuracy of the generated map is 72% using the overlaying method and 93% using the AUC method. [Display omitted]