Taking control of microplastics data: A comparison of control/blank data correction methods

Establishing a robust strategy to account for unintended processing contamination in microplastics research is of interest to the microplastic community who are currently focussed on developing harmonised methods, and to environmental managers who are calling for accurate risk assessment wrt microplastics.Six commonly used 'core' data correction methods were assessed for their suitablity to microplastics research: a) No correction; b) Subtraction; c) Subtraction of mean; d) Spectral Similarity; e) Limits of detection/ limits of quantification (LOD/LOQ) and f) Statistical analysis. An additional 45 variant methods based on these 6 core methods (n=51 in total) were used to correct a dummy microplastics dataset using control data. The dummy microplastics dataset (n=10 identical samples) was created in the laboratory to mimic the laboratory contamination which may arise throughout sample processing and handling. These were free from ‘sample’ matrix but contained processing solutions and MilliQ water as the surrogate sample matrix. Microplastics processing was conducted following AIMS Microplastics SOPs, and polymer type identified by FTIR and confirmed against the Nicodom Polymer Library.Data was analysed in Excel and R. Bayesian analysis was also assessed for suitablity.This work informs work practices for the IMOS long-term microplastic monitoring project, and for all projects conducted by the AIMS microplastics group. This work will also inform the wider microplastics community by starting the conversation towards harmonisation of microplastics data analysis and reporting.