Microplastic pollution is a global concern, affecting various water bodies. Rivers act as major transport pathways, making it essential to understand microplastic transport across different landscapes. While field monitoring provides snapshots of microplastic concentrations, achieving high spatial and temporal resolution remains challenging due to factors such as inconsistent measurement quality and resource limitations. Existing (micro)plastic transport models often focus on smaller scales (single catchments or few large rivers), requiring highly accurate input data which limits their broad applicability. This paper aims to develop a broadly applicable model for predicting microplastic fate and transport at high spatial resolution for large geographical areas, such as countries. The model integrates existing fate modeling approaches and a spatially resolved release model with a large-scale hydrological model. This approach will provide a more comprehensive understanding of the factors influencing microplastic transport and accumulation across diverse river systems, ultimately contributing to improved management strategies and mitigation efforts.

Several models of (micro)plastic transport in water have been developed. However, high-spatial-resolution models for large areas like countries or continents are limited, especially for freshwater systems. Existing freshwater models cover single catchments with a single river, a few large rivers, or small catchments without detailed river considerations. These models demand highly accurate hydrological and particle-specific input data, restricting their application to areas with readily available data. Furthermore, these models haven’t been widely applied across diverse river systems to investigate microplastic transport in various conditions. This lack of comprehensive, large-scale modeling hinders a complete understanding of microplastic transport dynamics and distribution.

The researchers developed a model integrating existing fate modeling approaches (similar to nanoDUFLOW or Full Multi models) and a spatially resolved release model with a large-scale hydrological model. This integrated model was applied to Switzerland as a case study, predicting the transported masses of seven different polymers (EPS, PP, LDPE, HDPE, PS, PVC, and PET) across all rivers and lakes. The entire Swiss river network and three focal catchments (Rhine, Rhône, and Doubs) were modeled, connecting river segments and lakes. Different scenarios were assessed: no retention (S₀), retention only in lakes (Slake), retention in the 15 largest lakes (S₁₅lakes), and retention in lakes and rivers (Sall). Microplastic masses were considered in three states: suspension, sediment, and deep sediment (accumulation). Input emissions were allocated to suspension, then subject to sedimentation and accumulation factors (fsed and facc), derived from literature data and adjusted for each polymer and location (river/lake). Transport in suspension was based on river flow velocity, while a steady-state assumption was used for lakes. Sedimentation and resuspension factors were applied for each river segment and lake, varying according to polymer type and location. The model was parameterized for each river segment and lake based on data including river segment length, flow velocity, lake surface area, and polymer-specific sedimentation and accumulation rates from literature. For a more detailed explanation, refer to Supplementary Information Sections 2, 5, 7, and 8.

The model highlights regions with high microplastic pollution and numerous rivers without significant pollution. Microplastic masses increase downstream, with the Rhine showing the highest levels. Smaller rivers in mountainous areas have less pollution. Polymer distribution, when normalized, is very similar across different polymers (correlation coefficient of 0.9). Approximately half of all input emissions are retained in Switzerland, with lakes retaining 33% (99% in the 15 largest lakes) and rivers retaining about 17%. Retention varies widely among polymers, due to differences in sedimentation and accumulation factors. The three focal catchments (Rhine, Rhône, and Doubs) show significant variation in outflowing microplastic masses (4565 kg/year, 300 kg/year, and 61 kg/year respectively). These catchments account for 88% of total Swiss microplastic outflows. The model estimated that 2–9% of microplastic masses leaving Switzerland are transported via sediment transport, with higher percentages for rivers with greater total microplastic transport (e.g., the Rhine). Lakes are crucial for retention, and their location influences the total microplastic mass. The largest lakes, Lake Geneva and Lake Constance, significantly reduce microplastic masses through sedimentation, up to a third of inflow mass. However, for the Rhine catchment, lake retention is less dominant compared to the Rhône due to the Aare's flow into the Rhine downstream of Lake Constance. The City of Geneva's emissions roughly equal the microplastic mass retained in Lake Geneva, meaning pollution at the Swiss-French border in the Rhône is primarily influenced by downstream sources, especially for denser polymers. For less dense polymers, lake retention is much more significant than river retention, particularly in lake-dominated catchments like the Rhône. A comparison with existing measurement data for the Rhine and Rhône rivers showed varying levels of agreement, with some underestimations and overestimations. The model successfully captured the dominance of polyethylene (HDPE and LDPE) and PP, aligning with previous measurements. However, comparisons for other polymers remained challenging due to inconsistencies in data reporting. The analysis of 502 randomly selected river segments revealed a non-linear relationship between microplastic retention and river length, with significant variation between catchments. The Rhône catchment more closely follows a logarithmic function, driven by Lake Geneva's influence. The Rhine catchment exhibits more complex retention patterns due to tributary contributions and variations within sub-catchments.

The model's comparison with existing measurement data reveals varying degrees of agreement, highlighting the need for improved data quality and more process-based understanding for better model validation. The model's limitations, including the lack of fragmentation and atmospheric deposition considerations, may lead to underestimation of total microplastic masses. Future improvements could integrate fragmentation, emissions into soil, and transport from soil to water, as well as coupled macroplastic modeling. The model's high spatial resolution reveals the importance of considering localized emissions, as demonstrated by the contrasting effects of upstream versus downstream pollution reduction scenarios in the Rhône catchment. The model's focus on masses, rather than particle numbers, necessitates future research to align with toxicity-related concerns. Size-specific release data would allow for incorporating multiple size classes within the model. The model's steady-state approach doesn’t capture the dynamics of flooding events, which could be addressed with temporal resolution improvements. Highlighting the non-linearity of microplastic retention underscores the importance of high-resolution modeling and spatially distributed data. Catchment-based approaches may overlook critical spatial variations.

This study provides a novel country-wide high-resolution model for predicting microplastic masses in river networks, emphasizing the significant role of lakes as retention sites. The model's findings highlight the importance of spatially resolved data and high-resolution modeling for accurate microplastic assessment in large areas. Future research should focus on improving data quality and process understanding, particularly in relation to lakes, dams, and flooding events, along with further model development to include factors currently not considered (e.g., soil-water transfer). The model serves as a valuable tool for policymakers to evaluate the impact of different policies and explore various scenarios.

The model's accuracy is limited by the availability and quality of input data, particularly regarding microplastic release and fate processes. The model currently simplifies microplastic size to a single class, neglecting size-dependent processes. The steady-state approach may not fully capture the temporal dynamics of microplastic transport, especially during extreme hydrological events like floods. Further model development is needed to integrate fragmentation of macroplastics, atmospheric deposition, and transport from soil to water bodies for a more comprehensive and accurate assessment of microplastic fate and transport.