EVALUATION OF EFFICIENCY SELECTED SECONDARY SAMPLING METHODS IN SOIL STUDIES

The purpose of the work was to compare effectiveness of common secondary sampling methods for assessing the distribution of soil pollution. The study case is based on an example of assessing the spatial distribution of soil contamination with lead in Slawkow area (Upper Silesian Industrial Region). This comparison was made in regard to both precision of the spatial estimation and minimization the cost of measuring campaign. The special attention was given to the often applied secondary sampling designs such as threshold radial (also known as adaptive cluster sampling) and adaptive fill sampling. These two methods were tested in typical municipal and suburban environment in Slawkow area. The work contains also detailed statistical and geostatistical analysis of above-mentioned contamination, and elaboration of series of its spatial distributions using numerous alternative sampling designs. The determined sampling plans make it possible to find compromise between ecological and financial aspects. A combination of the obtained results with the legal regulations in force concerning concentrations of heavy metals in soils are the basis for reliably estimation the ecological hazard arising from the soil contamination with lead in the Slawkow area. The results of performed analyses show that better efficiency in terms of cost and precision of measuring campaign gives rather coarser preliminary sampling design followed by appropriate secondary sampling then use the one-stage very dense measuring grid. However, the effectiveness of both threshold radial and adaptive fill secondary sampling designs is much worse than secondary sampling designs based on geostatistical methods using eg minimization of maximum or mean kriging variance criterion. However, it was also found that the effectiveness of both threshold radial and adaptive fill secondary sampling designs is significantly worse than secondary sampling designs based on geostatistical methods. Therefore, when a larger environmental research is envisaged the collaboration with experienced geostatisticians is always the right choice.