Microplastic monitoring and data quality assessment for effective model error evaluation

<p>The assessment of microplastic pollution in the marine environment is a requirement for the evaluation of the present situation and development of efficient combatting strategies. In combination with model based transport assessments, the monitoring of microplastics provides a tool for the assessment of the overall budget for the entire Baltic Sea and sub-basins. Furthermore, monitoring data sets can be used to evaluate the model performance in capturing spatial and seasonal pattern of marine pollution. Currently, the scope of microplastic monitoring is quite limited, which is why it is an important issue to control the quality of the available datasets and to derive useful indicators from the available data. Our study of single source data sets, with improved error statistics compared to multi source data sets, shows that the mean sampling error is still relatively high, about 40%&#8211;56%, which has been estimated using replicate samples. The lack of surface flow correction when using mantra net or trawl methods introduces and additional 12% uncertainty. When compared to model data, additional uncertainties come into play, related to the model characterization of microplastics as a set of spherical particles with a given density and diameter, which differs fundamentally from the broad range of values occurring in nature. It is therefore important to derive useful indicators from the measured and monitored data sets before attempting to validate the model. In our presentation, we will detail the assessment of sampling errors and provide an overview over the extent of microplastic monitoring assessed in the CLAIM project for the Baltic Sea. The collected data set was used to evaluate the quality of DMI&#8217;s microplastic transport model in reproducing spatial and seasonal patterns. The database of the model-observation assessment covers the 6 years period 2014-2019, with regular monitoring data sets in the eastern Baltic Sea being available since 2016. Finally, we discuss recommendations that could help to reduce sampling errors and derive indicators that are useful for a quality assessment of microplastic models. Aim is the development of operational modelling and monitoring capabilities for marine microplastic pollution. &#160;</p>