Comparing Analytic Hierarchy Process and Frequency Ratio Models for Delineating Groundwater Potential Zones in Northern Mozambique

This study investigated the effectiveness and accuracy of the analytic hierarchy process (AHP) and frequency ratio (FR) models, which integrated remote sensing (RS) and geographic information systems (GIS), for delineating and validating groundwater potential zones (GWPZs) in Northern Mozambique. For both methods, six key factors influencing groundwater potential were selected from which thematic maps were generated for each. In the AHP method, the influencing factors (IFs) were subjected to a hierarchy of criteria and sub-criteria. Then, using a pairwise comparison matrix based on professional assessment and literature review, relative weights were assigned to each factor. Analysis of the AHP ultimately resulted in the integration of the various thematic layers using the weighted sum tool in ArcGIS 10.8 to produce a composite GWPZ map. In the FR method, 1026 borehole locations for the area were randomly divided into two sets: 718 boreholes (70%) were used as a training dataset, and the remaining 308 boreholes (30%) were kept as a testing dataset for validation purposes. From the training dataset, the ratio of the probability of a groundwater event (wells) occurring in a class of influencing factors to the overall probability of that event happening in the study area was calculated as the FR for that class, representing the weight assigned to each factor. The overall FR was also computed using the weighted sum tool to produce a GWPZ map. The map produced from each model delineated the area into five zones: “very low”, “low”, “moderate”, “high” and “very high”. The testing dataset was then used to validate the GWPZ maps using the field data overlay technique. The respective validation results revealed the AHP model outperformed the FR model in terms of accuracy for delineating GWPZs.