Classification of hydrochemical samples with use of Fuzzy logic

Abstract

Geological and hydrochemical systems are sometimes too complex to analyze easily using conventional graphical or statistical methods. Often the chemical and physical properties of the natural system vary continuously, rather than abruptly. In other words, these underlying physical and chemical processes do not always produce discrete outcomes. Because of this continuity, statistical clusters may not be well separated and instead may form a sequence of overlapping clusters. Therefore, methods related to “fuzzy logic” may be useful for modeling and classification purposes. In this research by using Fuzzy partitioning (FP) method 61 groundwater samples of Zeviercherry and Kherran plains were classified into four groups. At the end results of this method were compared with hierarchical cluster analysis (HCA).