Natural radioactivity and metals in pit lakes in Sweden analyzed by principal component and cluster analysis

Abstract

The aim of this work was to determine which parameters are sufficient to measure in order to describe the water quality of a pit lake and to identify patterns in the data among different kind of pit lakes. The data consisted of ambient dose equivalent rate, elemental and radionuclide concentration, pH, and specific conductance in surface water and sediment samples collected from different types of mines. Data were tested for normality and log-normality and used in principal component analysis (PCA) and hierarchical cluster analysis (HCA). The normality tests indicated that only 40K was normally distributed, while only the 234,238U isotopes were log-normally distributed. HCA performed on parameters measured in surface water provided clusters that in most cases separated the elements according to their chemical groups. However, when HCA was performed on pit lakes, the clustering seemed to indicate that surface water might not be the preferred sample to differentiate between different types of pit lakes. PCA of surface water data resulted in three components that explained 72% of the variance when pH, SC, concentration of the elements Mg, K, Ca, Cu, Zn, Sr, Pb, activity concentration of 234,238U and 210Po, and ambient dose equivalent rate were included. For surface sediment data, the PCA resulted in three components explaining 83% of the variance when the concentration of Na, Mg, Al, P, K, Ca, Rb, Sr, Y, Tl, activity concentration of 234Th, 226Ra, 210Pb, 232Th (series average), and 40K, and ambient dose equivalent rate were included.