Statistical and spatial assessment of soil heavy metal contamination in areas of poorly recorded, complex sources of pollution

Abstract

In recent years, soil contamination assessment studies have aimed to quantify the extent and identify its sources. However, the quantification of the contributions to pollution from different sources still represents a serious challenge to the environmental scientists and engineers. Statistical and spatial analysis tools have been used successfully to assess the volume and spread of soil contamination. The techniques suggested so far require good historical record of the study area in addition to the large amounts of soil quality data that need to be collected. Furthermore, they are only able to identify/provide a qualitative description of the pollution sources and do not guaranty convincing results in cases of complex or poorly recorded contamination. Research described in this paper has developed a methodology that combines statistical and geostatistical analysis tools with geographic information systems for the quantitative and spatial assessment of contamination sources.

In a previous paper, principal component and factor analysis tools where shown to successfully explore the physical processes behind pollution. This information is proven invaluable in the design of canonical correlation analysis for the soil contamination data, so that distinctive pollution sources are separated quantitatively. The geostatistical analysis tools in conjunction with geographic information systems address the spatial dimension of the data and the pollution sources. The practical aspects of the methodology are illustrated using soil contamination data from Lavrio old mine site in Greece.