Application of Dempster-Shafer theory and fuzzy analytic hierarchy process for evaluating the effects of geological formation units on groundwater quality.

Abstract

This study investigates the impacts of different geological units on groundwater quality of an aquifer in southern Iran. The Kriging interpolation technique with a Gaussian semivariogram model was employed to prepare groundwater maps for different water quality constituents. In the next stage, two different models based on fuzzy analytic hierarchy process (AHP) and Dempster-Shafer theory (DST) were used to evaluate the overall water quality index based on the World Health Organization's drinking water standard in different parts of the aquifer. The DST model was able to generate water quality maps with 99.5%, 99%, and 95% confidence levels. The water quality maps were subsequently compared with the geology map of the area to determine the effects of different soil types on the water quality of the aquifer. Both methods showed poor water quality indices in the areas with an Asmari formation containing elevated levels of chloride and sodium ions. Comparison of water quality maps generated by the fuzzy-AHP and DST model revealed that the DST could more reliably handle the uncertainty in the water quality data, and thus was able to generate more accurate water quality maps. Increasing the confidence level in the DST model yielded water quality maps with a decreased overall water quality index. Results of this study could assist water management practices to generate water quality maps for their groundwater resources with confidence levels commensurate socio-economic importance of the region.