Bottom-Up Evaluation of the Uncertainty of the Quantification of Microplastics Contamination in Sediment Samples

The quantification and comparison of microplastic contamination
of sediments are affected by sample heterogeneity and the systematic
and random effects affecting sample analysis. The quantification and
combination of these components in the measurement uncertainty allows
the objective interpretation of analysis results. This work presents
the first detailed evaluation of the uncertainty of microplastic contamination
quantification in sediments. The random and systematic effects affecting
microplastic counts are modeled by the Poisson-lognormal distribution
with inputs estimated from duplicate sediment analysis and the analysis
of sediments spiked with microparticles. The uncertainty from particle
counting was combined with the uncertainty from the determination
of the dry mass of the analytical portion by the Monte Carlo method.
The developed methodology was implemented in a user-friendly spreadsheet
made available as the Supporting Information. The contamination of
sediment samples collected in various inland Portuguese waters was
determined, ranging from [0; 160] to [361; 2932] kg–1 for a 99% confidence level, and compared by assessing if the difference
between contamination levels is equivalent to zero for the same confidence
level. Several samples proved to have metrologically different microplastic
contamination. This work represents a contribution to the objectivity
of the assessment of environmental contamination with microplastics.