Análise multivariada e modelagem espacial de parâmetros hidrogeoquímicos dos sistemas aquíferos médio e inferior da bacia sedimentar do Cariri Cearense

Abstract

The growing demand for water combined with the droughts and the decreasing levels in the surface reservoirs, has turned groundwater into a strategic reserve, decisively influence the political-socio-economic development of a region. THE The present research proposes a methodology capable of modeling the spatial distribution in a environment with scarcity of physical-chemical parameters of groundwater. The parameters were analyzed according to the interpretation of hydrochemical facies, identification of dominant processes, the chlor-alkaline indices, the evaluation of the possible types of the weathering and the determination of the degree of water saturation for halite, gypsum, dolomite and calcite. Principal component analysis (PCA) was applied to identify the parameters that best explain their relationship with the waters of the wells under study. The analysis clustering was developed using dendrograms, the best Pearson correlation defined the best agglomeration and dissimilarity matrix techniques. For the scarcity of physicochemical parameters, the multiple data imputation technique was applied. At geostatistical modeling addressed the combination of 38 directions, 11 correlation functions and 3 types of weights to minimize the loss function. To the case study carried out in municipalities that exploit groundwater in the Cariri Cearense Sedimentary Basin, there was a higher incidence of bicarbonate waters, with prominent contact with the rock, with wells tending to saturation in the dry period for calcite and dolomite. The main components formed showed distinct characteristics, Dim1 - salinity, Dim2 - activities anthropic and Dim3 - responsible for the content of dissolved substances, according to the percolation of precipitated water on its way through the soil profile. For the groups it was possible to verify in the rainy season distinct groups for salinity and alkalinity, already in the rainy season dried by anthropic activities. By spatial modeling, the tendency of anisotropy was verified for most parameters with the exception of total dissolved solids, hardness and Electric conductivity. The model proved to be valid, the analysis of correlations, hydrochemistry, principal components and clusters showed congruence among themselves, an important factor in the inferential process of spatial modeling in data scarcity.