Spatial scale of DUs

Different study objectives pose very different kinds of assessment questions, and, most importantly for sampling considerations, different objectives require different spatial scales. Some objectives call for characterizing contaminant concentrations over a relatively large scale (up to acres), while others are contingent on distinguishing concentration differences at a much smaller scale (within tens of feet). The scale of a DU can depend upon whether the primary objective is to inform a risk assessment (i.e., DU is an exposure area) or to select a remedy (i.e., DU is a remediation unit).

DUs applicable to human health may not be readily applicable to ecological receptors. Therefore, when sites are evaluated for both human health and ecological receptors, multiple scales and project objectives are likely. For example, DUs for human health evaluations may correspond with individual residential properties, while DUs established to address ecological receptors may consist of much larger or much smaller volumes of soil, depending on the nature of the specified ecological receptor. Thus, the designation of new DUs and additional data collection efforts after initial sampling are sometimes unavoidable.

The need to characterize concentration variations at different spatial scales poses significant planning challenges. While the ISM approach has gained some acceptance for estimating mean concentrations in large areas, there are lingering concerns about its application where small areas with high contaminant concentrations are hypothesized. These concerns notwithstanding, variations in contaminant concentrations over small spatial scales can be addressed using ISM or a combination of ISM and discrete sampling. However, project planning always requires compromises and decisions, for example:

  • Smaller DUs are more likely to indicate the possible presence of areas with higher contaminant levels. However, sampling smaller DUs requires more resources, and project planners must always strike a balance between cost and certainty.
  • High-density discrete sampling may offer a reasonable alternative for assessing small-scale variations but may not be practical for large areas.
  • While low-density discrete sampling plans may be more feasible, they suffer from the inadequacies discussed in Section