Heterogeneity and discrete samples

Small- and Micro-Scale Heterogeneity. Discrete samples typically consist of about 200–300 g of soil, of which perhaps 1–30 g is processed and analyzed in the laboratory. Therefore, a discrete sample contains between about 7 and 300 possible analytical subsamples, only one of which is actually analyzed. The assumption is that every subsample taken from the discrete sample will result in the same concentration estimate if analyzed. As discussed in Section, this is often a poor assumption.

Large-Scale Heterogeneity. As discussed in Section 2.2.3, variation in contaminant concentration is expected at relatively large scales (i.e., on the order of residential yards and larger). This is the scale at which concentration trends, hot spots, and clean volumes of media are often of most interest. However, when low numbers of discrete samples are used and microscale and short-scale heterogeneity are present, data from discrete samples can miss the presence of large-scale contaminant trends. In other cases, they can misidentify the effects of microscale and short-scale heterogeneity as contaminant trends or hot spots that are not actually present. When a single discrete sample is found to have a concentration that is “hot,” it may mean that some meaningfully sized volume of contaminated soil is actually present at a site. But it may just as easily simply reflect the reality that a few “hot” samples are to be expected when collecting discrete samples from heterogeneous particulate materials like soil. Without additional corroborating evidence or additional discrete samples, these two situations are indistinguishable.