Integration of shannon entropy (SE), frequency ratio (FR) and analytical hierarchy process (AHP) in GIS for suitable groundwater potential zones targeting in the Yoyo river basin, Méiganga area, Adamawa Cameroon
Bidichael Wahile Wassouo Elvis, Arsène Meying, Ndougsa Mbarga Théophile, Kabe Moukété Eric Bruno, Owono Amougou Olivier
Abstract
Yoyo river basin, Méiganga, Adamawa region, Cameroon. The study aimed firstly to delineate and classify groundwater prospective zones in granito-gneissic terrain using advanced applications of remote sensing and geographical information systems through the integration of analytical hierarchy process (AHP), frequency ratio (FR) and Shannon entropy (SE). Secondly to compare the performance of these models to ensure effectiveness in computation. Thirteen groundwater influencing factors were adopted and weights were assigned to each thematic layer based on their relationship and characteristic with groundwater occurrence. Overlay weighted linear combination analysis tool in Arcgis environment was applied for groundwater prospect maps constructed. Groundwater maps was classified into four categories:I poor, moderate, good and excellent. These maps analysis revealed that AHP have performed well for good and excellent classes with percentage of coved area of 23.73% and 23.75% respectively. These areas are localised in central, south-eastern, north-western and southern parts of the basin. Validation of outputs maps using wells location showed that AHP have higher accuracy (Kappa= 79.53%, AUC (predicted) = 82.40%, AUC (success) = 87.09%), compared to FR (Kappa= 74.62%, AUC (predicted) = 76.68%, AUC (success) = 77.21%) and to SE (Kappa= 77.75%, AUC (predicted) = 70.61.40%, AUC (success) = 75.78%). These results are reliable and useful for various decision-making processes and sustainable management of groundwater resources in the Yoyo river basin.