3.2.2 Use of Sampling Units

SUs are subdivisions of DUs from which separate ISM samples are collected. The boundaries of an SU indicate the coverage of a single ISM sample; therefore, SUs define the scale of the ISM sampling, while DUs define the scale of the decision(s) based on that sampling. These definitions allow for the possibility that ISM samples from several SUs composing a DU can be used collectively to make the decision on that DU. It is possible to later redefine SUs as DUs (if the resulting scale meets the project objectives) to use the ISM sample results in appropriate decision mechanisms.

Figure 3-2 illustrates the relationship between SUs and DUs; a DU consists of one or more SUs. In the simplest situation shown in Figure 3-2a, the DU consists of a single SU (in which case the term "SU" is not necessary—the DU and the SU are one in the same). In this case, results from one or more ISM samples (e.g., replicates) collected from within the DU are used to make a decision. Figure 3-2b shows the DU divided into four SUs, each of which is separately sampled with one or more ISM samples. When SUs are used, the SU sample results can be used in an appropriate decision mechanism to support a decision for the DU. Valid estimates of the mean or 95% UCL for the DU can be derived from the SU replicates, while they also provide some information on the spatial distribution of contaminants within the DU.

Figure 3-2a. DU = SU (SU concept not needed)

Figure 3-2a. DU = SU (SU concept not needed)

Figure 3-2b. DU is subdivided into 4 SUs

Figure 3-2b. DU is subdivided into 4 SUs

Figure 3-2. Decision units and sampling units.


The mean concentration over the entire DU should be the basis for decision making if the DU is properly sized; that is, the DU passes or fails based on the SU sample results over the entire DU in an appropriate decision mechanism. Results for individual SUs should not be used to make decisions on SUs because by definition SUs are smaller than appropriate for a decision or because they have been insufficiently sampled. This is especially true if sampling involves only one ISM sample per SU (e.g., no replicates were collected) as discussed further in Section 4. Estimation of the mean DU concentration from individual and replicate SU samples is discussed in more detail in Sections 4.3.4.3 and 7.2.3.

In some cases, the planning team may determine that information about spatial variability within the DU is needed in addition to an estimate of the mean concentration of the DU. Sampling at the SU scale can provide such information, which may be valuable if the DU fails the decision and further investigation or remedial activities are indicated. In this case, additional systematic planning, designation of smaller DUs, and resampling may be necessary (see also Section 7.2.3). However, as discussed in Section 4.3.4.3, there are advantages as well as disadvantages to using SUs in this manner. Therefore, this approach should not be considered the default ISM strategy but merely an option that can be useful depending on the project objectives.

SUs may also be used when there are multiple sampling objectives or sampling scales for a given volume of soil, for instance, when areas must be assessed for multiple receptors with exposure areas of different scales. In this situation, the mean of smaller volumes of soil may be estimated through one or more ISM samples on the SU scale, while the sample results of the SUs can also be combined to estimate the mean concentration at the larger DU scale.

SUs may also be advantageous when very large volumes of soil cannot be sampled at the desired scale or sample density due to practical limitations (e.g., costs) or multiple receptors (e.g., human and ecological receptors). In this situation, sampling of SUs and some form of extrapolation to infer the DU mean from the subset of SUs actually sampled may be a feasible, if imperfect, alternative. However, there are a number of assumptions and cautions associated with this approach, as discussed in Sections 4.4.2 and 7.2.6. Decision mechanisms involving extrapolation are not acceptable to regulators in some states.

Finally, SUs may also be useful to characterize smaller source areas encompassed by large exposure areas to determine whether or not they require separate investigation and additional action. More detailed discussions on the use of SU sample data are offered in Sections 4.4.1 and 7.2.5.