Applying multivariate statistics for identification of groundwater resources and qualities in NW Turkey


EVEREST T., ÖZCAN H.

ENVIRONMENTAL MONITORING AND ASSESSMENT, vol.191, no.2, 2019 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 191 Issue: 2
  • Publication Date: 2019
  • Doi Number: 10.1007/s10661-018-7165-6
  • Journal Name: ENVIRONMENTAL MONITORING AND ASSESSMENT
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Keywords: Hierarchical cluster analyses, Lithology, Water origins, Interpretation, Quality, WATER-QUALITY, HYDROCHEMICAL EVOLUTION, CLUSTER-ANALYSIS, DISTRICT, CHEMISTRY, BASIN, PLAIN, GEOCHEMISTRY, TAMILNADU, SURFACE
  • Çanakkale Onsekiz Mart University Affiliated: Yes

Abstract

This study, performed in Canakkale-Ezine in NW of Turkey, analyzes the physicochemical properties of 37 groundwater wells. These 37 wells were chosen to represent each geological unit in the study area. The main purpose of the study and its contribution to the literature is to produce information about the resources and availability of groundwater by using multivariate statistical methods and lithology. For determination hydrochemical facies of groundwater, Piper trilinear diagram was used. Gibbs diagram was applied for determining the mechanism of groundwater chemistry and diagram showed that the interaction of rock-water is more dominant in the study area. Multivariate statistics were applied to physicochemical properties for identification origins of waters. According to the Piper diagram, 16 of the wells were identified as Ca-HCO3 type, 13 of them as Ca-Cl type, 5 of them as mixed Ca-Mg-Cl type, 2 of them as Na-Cl type, and 1 as Ca-Na-HCO3 type. In the study with the purpose of determining the resources of groundwater, the physicochemical properties of the wells are analyzed with hierarchical cluster (HCA) and non-hierarchical cluster (K-means) methods, and the resources are associated with the lithology based on these methods. A total of 37 wells are divided into five different clusters through the HCA method. Further, for the interpretation of the resources of the groundwater, the facies of the waters on the Piper diagram are evaluated based on the five clusters generated through the HCA method and on the lithology. In the study, the results obtained from the K-means method are not significant and in line with the lithology for the interpretation of the resources of the groundwater. In conclusion, this study with limited dataset reveals that using HCA method is very effective to identify the origins of groundwater and present the association with lithology.