1.1 Summary of ISM as an Environmental Sampling Approach
Like all sampling approaches, ISM should be applied within a systematic planning framework. Figure 1-1 shows a general ISM flow process. One of the first steps in such a framework is to have the investigation project team establish a working conceptual site model (CSM). Once the CSM has been agreed to, the project team defines the data quality objectives (DQOs) and determines the appropriate decision unit (DU) size(s) and location(s). DUs are based on project-specific needs and site-specific DQOs; both considerations specify and constrain the appropriate end use of the data. The size of a DU is site-specific and represents the smallest volume of soil about which a decision is to be made (USEPA 1999, Ramsey and Hewitt 2005, HDOH 2008a, ADEC 2009). In some cases a DU is composed of smaller units known as sampling units (SUs), as discussed in Section 3. The requirement to explicitly and appropriately define the DU that each incremental sample represents is a key component of ISM and is discussed in detail in Section 3.
* The statistical performance of the 95% UCL calculation depends on the properties of the data set and the sampling design. Note that ProUCL or FLUCL does not currently include the statistical algorithms for handling ISM data (see Section 4.0 and Appendix A). ** See Section 7.
ISM planning includes the development of an ISM protocol for the number of increments and replicates to be collected for each ISM sample. An incremental sample is created by collecting many (usually 30–100) equal-volume increments in an unbiased manner from throughout the entire DU. The combined increments (frequently totaling a kilogram or more) are typicallyprocessed at the laboratory and subsampled to provide an analytical aliquot of only a few grams that is used for analysis. The final analytical aliquot is the target sample.
ISM addresses major sources of sampling error and increasing sample representativeness.
ISM is designed to provide an unbiased, statistically valid estimate of the mean value of an analyte within the DU. Through adequate spatial coverage of the DU as well as disciplined handling, processing, and subsampling of the single sample formed from the increments collected, ISM works to overcome major sources of error in both sampling and subsampling of soils that have often been apparent with current sampling practices. By design, ISM provides complete spatial coverage within the DU; however, ISM does not provide information on the spatial distribution of contaminants within the DU. Should this spatial variability be important to the decisions being made, a smaller DU should be used. ISM may not be appropriate in certain situations (see Section 8 for further information on the limitations on ISM).