Analysis of Groundwater Quality using Mamdani Fuzzy Inference System (MFIS) in Yazd province, Iran

Abstract

Precise classification and identification of groundwater quality is an essential task for meeting the goals of environmental management. Traditional classification methods of the water quality parameters use crisp set with prescribed limits of various organization. One of the decision making problems about water quality using methods is facing various uncertainties. Recent years have proven fuzzy-logic-based methods capability controlling uncertainties in different environmental problems. The present study utilized a newly devised Mamdani fuzzy inference system to assess groundwater quality in Yazd province. This method made use of 10 measured chemical parameters in 60 samples of groundwater. The samples were collected from wells, springs and kanats. The results showed that 20 groundwater samples were in the "Desirable" class with a certainty level of 32.29-100%, and 20 samples were in the "Acceptable" group with a certainty level of 37.07-92%, and 20 samples were in the "Non-acceptable" category with a certainty level of 43.33-88.78% for potable purposes.