Auteur :
Amar
Hicham,
Rahali
Hassane
Date de publication : 23/11/2016
Type : Actes de congrès / Séminaire / Atelier
Thème : Information écologique
Couverture : Maroc
Groundwater contamination has become a serious problem in the world, and particularly in arid and semi arid area. This water is threatened by the contamination of different kinds of pollutants (biological, physical or chemical). So the protection of that source has become an obligation to the global policy by making the best decision to resolve this issue. In this study, a DRASTIC (D) and also Pesticide DRASTIC (PD) approaches have been performed within Geographic Information System (GIS) to evaluate the groundwater vulnerability in the Gharb Plain (Morocco). In addition, sensitivity analysis (SA) was applied to highlight the importance of model parameters for the aquifer vulnerability assessment. The result o the SA was used with the analytic Hierarchy process AHP to optimize the weight of the sensitive parameters by revising their weights. Finally, nitrates concentrations are used to evaluate the performance of the D and PD in the first step; and between AHP-D and AHP-PD in the second part with the AUC curve (Area Under the Receiver Operating Characteristic Curve (AUCROC)). It turned out that the highly vulnerable areas are located in the flat terrain and in the coastal areas where agricultural activities are practiced. The evaluation of the performance of the fourth models with the application of the AUC curve (Specificity Vs sensitivity) demonstrates that the PD and the AHP-PD method are efficient than D and the AHP-D. The optimization of intrinsic methods (AHP and AS) did not improve their performance, since they were unable to obtain better correlations with nitrates than conventional methods.