CONICET | Buscador de Institutos y Recursos Humanos

This work aimed to analyze a hydrochemical dataset of Langueyú creek basin (Buenos Aires, Argentina) applying two multivariate statistical analysis methodologies: cluster analysis and main component analysis. In the basin is settled the Tandil city, which relies on groundwater for different purposes (drinking, industry, agriculture, etc.). This is why it is considered essential to a thorough understanding of hydrochemical characteristics. Phreatic levels were measured and groundwater samples were taken with a quaterly frecuency in a monitoring net of thirty wells. Physicochemical parameters, conductivity, pH, major ions and nitrate, were determined. The results of cluster and main component analysis showed that it is possible to separate two groups of samples: One group is composed of samples located in the recharge zone and the proximal foothill (elevation 150 m.a.s.l); and a second group is composed of samples that are located to the NE. It was concluded that two processes are determinants to define hydrochemical characteristics of the study area: a) the salts acquisition during the way through the porous medium, which begins with the dissolution of carbonates, due to the acidity acquired by the reaction with CO2 during its infiltration through the unsaturated zone, and b) calcium-sodium exchange combined with the retention of potassium in clays. Multivariate statistical analysis has proved to be a useful tool for interpreting hydrochemical results and the validation of the hydrogeological conceptual model.