Managed aquifer recharge site suitability analysis in the Nabogo Basin of Ghana using GIS and Monte Carlo simulation-aided analytic hierarchy process

• Published in: Arabian journal of geosciences 2024, Vol.17 (2), Article 76
• Main Authors: Adam, Ayuba B., Appiah-Adjei, Emmanuel K., Adjei, Kwaku A., Gyampo, Maxwell Anim, Owusu-Ansah, Emmanuel De-Graft Johnson
• Format: Article
• Language: English
• Subjects: Analytic hierarchy process; Aquifer management; Aquifers; Climate change; Confidence intervals; Drainage density; Dry season; Earth and Environmental Science; Earth Sciences; Geographical information systems; Groundwater; Groundwater recharge; Hierarchies; Land use; Monte Carlo simulation; Original Paper; Overexploitation; Soil properties; Soil texture; Texture; Transmissivity; Uncertainty; Water management; Water resources
• ISSN: 1866-7511; 1866-7538
• DOI: 10.1007/s12517-024-11865-z

Groundwater in the Nabogo Basin is threatened by increasing domestic and irrigation demands during the dry season, owing to climate change and overexploitation. Managed aquifer recharge (MAR) has therefore been proposed as a water management technique to augment groundwater resources. However, the success of MAR schemes depends on the selection of sites with suitable surface and subsurface properties to enhance the recharge rates. Thus, this study created spatial maps of the slope, drainage density, land use/cover, soil texture, transmissivity, groundwater depth, and aquifer thickness in the basin to create a suitability map. Analytical hierarchy process (AHP) with and without Monte Carlo (MC) simulation was used to create weights for each criterion based on seven expert opinions to assess their performance in dealing with uncertainties. The spatial maps and weights were then integrated to create a MAR suitability map. The results showed similar mean weights for AHP- and MC-aided simulations. However, MC-aided simulations narrowed the confidence interval and reduced the standard deviation of the AHP weights by 50–89%, thereby reducing the uncertainty and sensitivity of the weights. Furthermore, the suitability map showed that 38.8% (1124 km 2 ), 39.2% (1136 km 2 ), and 22% (637 km 2 ) of the basin was suitable, less suitable, and unsuitable, respectively, for MAR applications. A receiver operating curve (ROC) value of 77.8% from the validation process indicated good model performance. This proves the effectiveness of MC simulations in reducing errors and uncertainty in the AHP. Therefore, it is a useful tool for mapping MAR suitability.