Working on an experimental dataset to derive descriptive parameters with well-known procedures appears to be a straightforward process. However, due to the complexity of environmental systems, the procedures cannot be completely defined, and commonly hypothesis are required to obtain final results. These degrees of freedom cause a spread in results that is often wider than the experimental uncertainties and that should be taken into account in the measurement accuracy estimation. As an example of this issue, the sources of uncertainty, including reproducibility, that occur in dating marine and lacustrine sediment cores with the 210Pb methodology are discussed in this paper. The IAEA Coordinated Research Project (CRP) “Study of Temporal Trends of Pollution in Selected Coastal Areas by the application of Isotopic and Nuclear Tools” promoted this discussion, its main objective being the determination of a streamlined, harmonised and validated methodology for marine sediment dating. The uncertainties estimated from experimental data and mathematical propagated must be complemented with reproducibility contribution: some elements are specific for the 210Pb method (mixed layer, finite thickness of layers and identification of the equilibrium depth), while others are more general (exponential fitting and data interpolation). The final spreading of results can be mainly controlled by reproducibility, and efforts are needed to standardise procedures to reduce the effect of reproducibility and to better estimate the overall uncertainty of measurements. It is necessary to underline the importance of a thorough (a-priori) cost-effective analysis of the uncertainty evaluation of the measurement results in relation to the aim of the study, taking primarily into account the overall accuracy required case by case. © 2022 Elsevier B.V.

cited By 0