Various inference systems for classification of water quality status: A case study

Water quality is considered one of the main factors controlling health and the disease state of humans and animals. Four assessment methods (two pollution indexes and two fuzzy mathematical models) were used to understand the water quality parameter levels and characteristic accurately. Several physico-chemical parameters such as dissolved oxygen, nitrate, nitrite, phosphate, chlor, sulphure and total organic carbon were measured in Karasu River, Turkey. Water quality was assessed as class IV (heavy polluted) in A, B, C and D stations and class III (polluted) in station E with single-factor index method. It was also identified as class III (polluted) for waters in A, B, C and D stations and class II (slightly polluted) for water in station E with the comprehensive index model. Using the two fuzzy mathematical methods (single-factor deciding and weighted average models), the water quality was determined as II, II, III, III and II classes for waters in A, B, C, D and E stations, respectively. In conclusion, it can be proposed that fuzzy logic assessment methods may also be used as an alternative tool for decision-making in environmental management.