CONTRIBUIÇÃO DE TÉCNICAS MULTIVARIADAS E GEOESTATISTICAS PARA O ESTUDO DA DISTRIBUIÇÃO ESPACIAL DE METAIS EM ÁGUAS SUPERFICIAIS DO PÓLO INDUSTRIAL DE MANAUS

An investigation on spatial distribution and possible pollution sources in the Manaus Industrial
District (MID) was conducted using geographic information system (GIS) technique and multivariate
statistics. The box plot results show a high scattering of the temperature, pH values and organic matter.
Principal components (PCA) and hierarchical analysis (HCA), two multivariate statistics used, showed a
separation between dissolved and total metal fractions. The spatial distribution of the concentrations of Fe,
Mn and Zn were determined by kriging method. The semivariogram maps, obtained by ordinary kriging
interpolation, indicate that the TPM has high mobility according to season period. In rainy season, heavy
metal has higher mobility that dry season.