Utilizzo della spettroscopia vis-NIR per la caratterizzazione di suoli potenzialmente contaminati da metalli pesanti

In recent years, soil reclamation and its sustainable management have become important environmental issues. In fact, soil resource is increasingly degraded both directly and indirectly by human activities, therefore making it necessary activities of reclamation, requalification and monitoring. The degree of contamination and soil degradation varies from country to country but it is generally more severe in more developed countries, where it represents an ever-growing concern. Therefore, remediation of contaminated soils should be a priority in these countries. Usually, the study of soil, the assessment of its properties and its contamination status are carried out by conventional laboratory analyses which, although useful and practically irreplaceable for detailed investigations, involve long times and high costs and they often require the use of environmentally damaging chemicals. Therefore, analytical determinations are unsustainable where a large number of soil samples need to be analysed, such as in the case of particularly large areas and/or for the determination of many environmental parameters and many soil properties. Hence the need to use faster and cheaper techniques for soil analysis. In this regard, in recent decades reflectance spectroscopy in the visible-Near-Infrared domain (visNIR) has proven to be a viable alternative to conventional soil analysis. Compared with traditional methods, vis-NIR spectroscopy is faster, cheaper and non-destructive. It allows qualitative and quantitative assessments of the main soil constituents, as well as of organic and inorganic contaminants that may be present in the soil. Therefore, this technique can be useful both in the pre-screening of the possible soil contaminants and in the activities of characterization and monitoring of the environmental quality status of soils. Inorganic soil contaminants and, in particular, heavy metals, do not show characteristic absorption features in the vis-NIR region but they may be indirectly detected because they are absorbed or complexed by some spectrally active soil constituents, such as organic matter, oxides, carbonates and clay minerals. In view of this, the aim of the present work is to highlight the potential of using the vis-NIR spectroscopy, combined with multivariate statistical methods, for the evaluation and prediction of soil heavy metal contents, reporting a recent application of this technique in the investigation of potentially contaminated soils of the so-called "Area Vasta" of Taranto (Apulia Region, Southern Italy), which is one of the most investigated and monitored areas of Italy, due to the possible risk of contamination. The attention was focused on three of the investigated heavy metals (Beryllium, Be; Vanadium, V; Thallium, Tl) which exceed the contamination threshold concentrations values, defined by the Legislative Decree 152/2006 and by the recent Ministerial Decree 46/2019 for the agricultural areas, with more frequency than other heavy metals detected in the study area. Partial Least Square Regression (PLSR) analysis was applied to reflectance spectra (predictive variables) and to contaminant contents measured in the laboratory (response variable) to calibrate and validate prediction models of these three contaminants in the specific pedo-environmental conditions of the study area. Very good models were calibrated for Tl and Be and a good model was calibrated for V. This is the first study regarding the application of vis-NIR spectroscopy for the evaluation of soil contamination by Be, V and Tl in the specific pedo-environment of southern Italy.