Application of statistical methods to study seasonal variation in the mine contaminants in soil and groundwater of Goa, India

Abstract

Soil and groundwater samples were collected during two seasons, premonsoon (April 1990) and postmonsoon (December 1990), and analyzed for major elements (Na, Mg, Al, Si, P, K, Ca, Ti, Mn and Fe), trace elements (Ni, Pb, Co, Cr and Zn) and water parameters (pH, conductivity, acidity, alkalinity, hardness, Cl and SO4). All the data were subjected to linear discriminant analysis and partial correlation analysis in order to understand the seasonal variation in the data. It was observed from the Mahalanobis generalized distance that in both soil and groundwater samples there was a large difference in the concentration level of premonsoon and postmonsoon data. Linear discriminant functions were calculated to distinguish between premonsoon and postmonsoon samples. From the partial correlation coefficient analysis of soil samples, dominance of chemical weathering and precipitation of atmospheric fallout during monsoon were inferred. In the case of the water samples, high conductivity and high hardness in the postmonsoon samples as well as atmospheric fallout of Pb and Ni during the premonsoon period was suggested from partial correlation of water samples.