A.6.4.1 Compositional heterogeneity (CH)
CH is present when particles differ in their size, shape, density, and contaminant loadings (i.e., mass per volume). Different particles consist of different amounts of mass and concentrations. For example, finer particles can be expected to contain greater contaminant masses and concentrations. Presence of CH gives rise to FE (i.e., bias or relative bias in the mean). For a heterogeneous DU with CH, mean estimates are typically biased, and FE (relative bias) cannot be completely eliminated through improvements in sampling design. Presence of CH introduces DH; therefore, as CH increases DH also increases. Gy’s sampling theory suggests that FE can be addressed (i.e., reduced but not eliminated) by collecting and combining several increments, each of the same mass (Pitard 1993). This is a fundamental motivation for applying ISM.