ASSESSING SOIL POLLUTION USING A GENETIC ALGORITHM

In this article, we present a new approach for the classification of soils in different groups based on the proportion of chemical elements contained by them. Firstly, the k-means algorithm is used for clustering purposes, and then the soil samples are redistributed in groups using a genetic algorithm. The application is important for selecting soils of the same type for agricultural use. Heavy metals are elements found in the Earth's crust, in variable concentrations [1]. They also appear in contaminated sites as a result of human activity, the most toxic elements being arsenic, cadmium, chromium, copper, lead, manganese, nickel, and zinc [2]. Even most of the heavy metals are not soluble in water, their availability is influenced by different factors, as biological, chemical, and physical factors [3-5]. In the presence of ligands in aqueous solutions, the sorption of metals competes with their complexation with humic substances [4, 5]. The infiltration process due to the precipitation leads to the contamination of groundwater whose cleaning is very difficult or, many times, impossible [6-8]. Soil contamination has as a secondary effect the groundwater contamination, so that the groundwater vulnerability is assessed in different studies in correlation with the land use and human-built environment [9-13].