Mapping of groundwater vulnerability using optimized AHP-GOD Models and geospatial factors (Case study: littoral sedimentary aquifer of Ain Temouchent – NW Algeria

Abstract

Abstract The monitoring and protection of coastal aquifers has become a primary priority and a subject of research with the integration of Vertical Electrical Sounding (VES) in order to map groundwater vulnerability to contamination, and to evaluate their aquifer protective capacity. A study aimed to to identify an exposure level of contamination for a sustainable use of the water resources in the aquifer of Ouled Taoui - Boujema aquifer situated in Ain Temouchent – Algeria. For this purpose, a dataset was established including 170 VES soundings, topography ASTER GDEM, geological maps and finally boreholes were acquired to develop the AHP (Analytical Hierarchy Process) by ordering the weightage of geospatial factors in vulnerability mapping. However, a GOD-LT model (G: groundwater type; O: aquifer overlying strata; D: aquifer depth and L: Longitudinal conductance; T: Topography slope and Th: Thickness) has been evaluated using optimized AHP as predictive model abbreviated GOD-L, GOD-LT and GOD-LTh. Furthermore, these factors are considered as the causal factors of aquifer vulnerability (AVCF’s). The geospatial factors were derived on AHP-WLA (Weight Linear Average) for rating optimization. The pairwise comparison matrix was developed on the AVCF’s factors for AHP-GOD. The validation result showcases that 72% of prediction consistency is optimized for GOD-L model compared to 64% for the standard GOD. The GOD-LT and GOD-Th demonstrated the best reliability under consistency index. However, the vulnerability to pollution is validated by using nitrate concentration, water corrosivity and the degree of aquifer capacity protection and can be efficient mapping for water supply and environmental assessment.