Variability in Contamination by Heavy Metals: Sampling Implications

The choice of a sampling design forassessing soil contamination is a key issue. The leveland spatial pattern of contaminants needs to be takeninto account by those buying, selling or developingsuch land. This paper evaluates sampling design andsize in order to provide satisfactory and efficientestimates of contamination levels. Geostatisticalmethods are used to quantify the spatial dependence ofheavy metals. Semivariance is calculated forlog-transformed data to minimise the effects ofextreme outliers. Several models at different samplingstages are presented for characterising spatialpattern of heavy metals at the investigated site.Variogram models are calculated for a range of samplesizes. The results are assessed by a jack-knifingtechnique and then used for block kriging. The overallconclusion is that a sample size of 50 to 60 isoptimal for characterising variability incontamination levels at an old industrial site of 6.25 ha.