RESUMEN
In order to assess the capabilities of a remediation technology, and to judge of its efficacy, it is necessary to evaluate the initial average contamination level of the soil, an operation that can be difficult because of the inhomogeneity of the contamination itself. The goal is even more challenging when different contaminants are present, greatly differing both in terms of nature and of concentration. By referring to an industrial site contaminated mainly by As, Cd, Cu, Pb, Sb, Tl and Zn, we present a new approach for the necessary processing of sampling data, in order to establish the pre-intervention baseline: an estimate of the average contamination has been obtained through a suitable integration of the volume underlying the distribution curve of each contaminating species. This information, otherwise not accessible by means of sampling of discrete points, is useful in evaluating the effectiveness of the remediation technology under investigation, and can also be considered for other reclamation approaches as well. Since "chemometrically acceptable" results are typically achieved by increasing the number of samples (with related analytical investments), the proposed approach can help keep low these ancillary costs, while providing results that are more reliable.