Application of multivariate statistics in the diagnosis of salinization process in dams of northeast semiarid, Ceará/Brasil

Abstract

In Ceará state of there are several dams used to mitigate the impacts of droughts to the populations. However the waters of these dams are being affected by different salinization processes, affecting their quality and consequent use. This work refers to the Favelas and Trici basins, a sub-basin of the Alto Jaguaribe/CE, located southeast of the Municipality of Tauá. In order to establish a diagnosis of the cause or causes of salinization in the watersheds of the Favelas and Trici reservoirs in the Ceará state, a set of information was generated for the two basins. Among them, water samples were analyzed, in which the variables water temperature (TH₂O), electrical conductivity (CE μS / cm), pH (hydrogen ion concentration of medium), oxidation and reduction potential (ORP), total dissolved solids (TDS). Statistical analysis univariate and multivariate of principal component analysis and discriminante, using the PAST 3 software were used to analyze the results obtained. The statistical results and the multivariate statistical analysis of the variables TH₂O, CE μS / cm, pH, ORP and TDS, both in Favelas and Trici, it is verified that they have different behaviors. Therefore justifying, the influence of the different processes of salinization that are occurring in the two basins.