2.6.3 Incremental Sampling Methodology
Although composite samples are not typically considered to be ISM samples, by definition, all ISM samples are considered to be composite samples. It should be noted that a number of organizations, including regulatory agencies, are still in the process of defining what characteristics must be present to be considered an incremental sample vs. a traditional composite sample. However, ISM is a specialized type of composite sampling with specific structure and requirements that stand apart from common compositing practices. ISM is designed to provide more precise and less biased estimates of the mean concentration in soil by addressing specific sampling errors. Consequently, ISM can result in better performance in terms of decision error reduction than other sampling methodologies. The following are primary advantages to the use of ISM sampling approaches:
- requires designation of a targeted population (the DU) prior to sampling
- provides less biased and more precise estimates of the mean than low-density discrete sampling plans
- is more cost-effective than moderate- to high-density discrete sampling plans with a comparable level of decision quality
- tends to produce normal rather than lognormal or nonparametric data distributions
- specifies protocols for laboratory and field procedures to control sampling error
Gy theory is designed to minimize sources of error in the sampling and subsampling of heterogeneous bulk volumes of particulate material. ISM is consistent with the principles of Gy theory and provides a structured sampling protocol intended to reduce the sampling error associated with heterogeneity through the implementation of the following steps:
- collection of a large number of increments
- reduction of particle size
- collection of a large bulk sample mass
- implementation of field and laboratory subsampling techniques
These steps control the FE and the GSE. Control of the long-range and periodic fluctuation heterogeneity errors is addressed in project planning, during which appropriately sized DUs are identified. The increment DE, the increment EE, and the PE can be controlled through correct sampling and subsampling, aspects of which are discussed in Section 5 and Section 6.
ISM sampling produces an estimate of the mean contaminant concentration in soil within a specified volume (i.e., a DU). As with any estimate derived from sampling, ISM results are subject to errors, the components of which were described in Section 2.5. Statistical analysis can provide an understanding of error introduced by sampling. Rigorous statistical analysis regarding the extent to which various ISM sampling strategies provide accurate estimates of the mean contaminant concentration have not yet been published, but Section 4 includes an in-depth discussion of the statistics for ISM, and Appendix A includes relevant simulation studies. This information is necessary to understand how factors such as number of increments, number of replicates, and contaminant distributions across the site influence the reliability of ISM estimates of mean contaminant concentration. The reliability of ISM based on statistical principles is vital to widespread acceptance of this sampling method for regulatory purposes.