| Autor: |
Glass, H. J., Camm, G. S., Glass, H. J., Camm, G. S. |
| Zdroj: |
Journal de Physique IV - Proceedings; May 2003, Vol. 107 Issue: 1 p549-552, 4p |
| Abstrakt: |
Data analysis techniques arc applied to characterize 1. correlations between inorganic elements occuring in contaminated soil and 2. correlations between areas on the basis of similar chemical footprints. A dataset of 78 points was formed by measuring the concentration of 33 inorganic elements in soil samples from a site where arsenic was formerly processed. Using principal component analysis, it is shown that the concentrations of a series of heavy metals and arsenic are strongly correlated. Factor analysis suggests that only these correlations are significant. Elemental cluster analysis is used to obtain a definitive characterization of the correlations. The main elements of the first principal component are firmly clustered. Elemental clustering may be complicated by the complex nature of the area in terms of topology, vegetation/land use, hydrology, and history. In attempt to discern the explanatory variables, spatial cluster analysis was performed. This technique proved very suitable for identifying areas with high chemical similarity as well as outliers in the data. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
