Multivariate statistical techniques were employed in order to identify the similarity in the salinity levels of the water reservoirs of Ceará State, Brazil. Also, to identify the factors that explain the variability of water quality. The data set used in this investigation came from the Company of Water Management (COGERH). The studied period extended from 1998 to 2009, in a total of 807 samples from 48 different sources. The investigated parameters were: Cl-, Ca2+, Mg2+,Na+, HCO3 -, EC and SAR. The Hierarchical Cluster Analysis identified four groups, being the types and concentrations of the analyzed salts the determining factors. In groups 1 and 2, the salinity level of the water was, mainly, due to the geology and the climate of the region. While for group 3 and 4, besides these factors, the water salinity was influenced by human activities. According to the results, the first and the second components (independent factors) explaining, 57.28 and 26.77%, respectively, the total variance of seven variables. The most important parameters related to the variability of the water salinity level were Cl-, CE, Na+ and Mg2+, which were related to salt solubility.
water quality; salts; reservoirs; semiarid
