Beyond advection: A Lagrangian framework for microplastic transport in depth-averaged shallow water flows

Microplastics are an emerging contaminant in aquatic environments, posing risks to ecosystems and human health. Reliable computational tools capable of predicting microplastic transport are therefore essential for environmental assessment and mitigation strategies. This work presents a Lagrangian Particle Tracking (LPT-MP) model for microplastic transport under depth-averaged shallow flow conditions, coupled online with a two-dimensional Eulerian shallow water solver. The LPT-MP model incorporates the main processes governing microplastic dynamics, including horizontal advection and turbulent dispersion, settling velocity, degradation and biofouling, as well as a vertical flow velocity approximation. The model is implemented within the open-source SERGHEI framework using a Kokkos-based architecture, ensuring performance portability across CPU and GPU platforms and enabling large-scale simulations. Model performance and computational efficiency are evaluated through two synthetic benchmarks and one experimental validation case. Results demonstrate that settling and biofouling significantly influence the particle trajectories, while degradation produces negligible effects within the simulated time scales. The additional MP physical processes increase the total computational cost by approximately 6% with respect to the baseline advection-dispersion implementation, demonstrating the computational efficiency of the proposed extensions. The predictive capability of the model is assessed against an experimental benchmark, showing reasonable agreement in global deposition metrics while revealing limitations in the representation of turbulent lateral mixing. In summary, the presented framework provides an extensible and performance-portable tool for microplastic transport simulations in shallow water environments.