Estimation of measurement uncertainty including surrogate recoveries in the study of polycyclic aromatic hydrocarbons in marine sediments

We examine the use of surrogates in the measurement of concentrations of polycyclic aromatic hydrocarbons (PAH) in marine sediment, consider shortcomings of current interpretations of the results from this widely used approach, and propose an alternative data treatment. We focus on the current common practice of accepting all data within a predefined range of acceptable recoveries as equally valid and propose treating surrogate recovery data as one component of measurement uncertainty. Our analysis leads us to conclude that the more uniform surrogate recoveries are, the more justified the assumption that they reveal measurement bias and appropriately can be used as corrections; but that the less uniform recoveries are, the greater the need to treat them as a source of measurement uncertainty. We recognize that acceptance of the ideas presented here will require much thought and refinement by the environmental analytical community. We welcome a thorough discussion.