Geostatistical approach for validating contaminated soil measurements

Soil and top soil contaminations are generally produced by air deposition and subterranean leaks due to diverse factors, namely, industrial activities and natural phenomena (e.g. aerosols). However, a continuous monitoring is needed to assess eventual contaminants that can be on top soil. When the soil extension is very high, it is very difficult to perform accurate analysis because of excessive cost of characterization and successive analytical measurements. But in some cases, analytical data could not be available for all co-ordinates located in the area under test. Geostatistical approach could help in solving the missing data problem or helping in finding data in case of large meshes applied on the area under test. The research illustrates the opportunity of recovering data making a prediction by means of Kriging techniques. The application has been performed on data coming from deposimeters used for collecting atmospheric deposition on soil. PCBs (polychlorinated byphenils) are pollutants that are necessary to determine on soil, especially where industrial activities are carried out. The paper also illustrates the optimal conditions for increasing accuracy in recovering data thanks to fact that once a few numbers of point are known, it is possible to predict the trend of values of PCBs in unknown locations of the considered area.