The use of multivariate statistical methods for the classification of groundwater quality: a case study of aqueducts in the east of Tehran, Iran

Natural and human factors have always threatened the health of Qanats, valuable water sources for arid and semi-arid regions. The present study decided to qualitatively classify eight selected Qanats of East Tehran, Iran, using two multivariate statistical methods, cluster analysis (CA) and principal component analysis (PCA) based on parameters includind pH, TDS, EC, Na2+, Ca2+, Mg2+, K+, Cl-, NO3 and SO42- according to standard methods during the summer of 2020. Data were analyzed by CA and PCA methods, the results of which based on the degree of pollution divided the studied stations into three groups, high pollution (anthropogenic origin), moderate pollution (natural and anthropogenic origin) and low pollution (natural origin). The stations close to each other for quality status were placed in the same group. The eigenvalues obtained from PCA based on the evaluated parameters showed that the first and second components explained more than 58% of changes between the stations. Analyzing the coefficients of each parameter (eigenvectors) for the first and second components revealed that the main causes for the difference between the stations were Cl-, Na2+, Mg2+ and SO42-, TDS and NO3. The two-dimensional display of the stations based on the first two main components confirmed the grouping resulting from the cluster analysis and was able to separate the investigated stations from each other like cluster analysis. The findings of this research highlighted the usefulness and efficiency of two multivariate statistical techniques, CA and PCA, to effectively manage Qanat water quality.