New paper on a machine learning model to predict microplastic concentration in urban stormwater

In a recent article in Scientific Reports, Amir Reshadi and coauthors present a machine learning model for the prediction of microplastic (MP) concentration in urban stormwater using an original dataset with a broad array of environmental and socioeconomic variables. They demonstrate that MP concentration in stormwater is governed by a combination of hydrological and meteorological conditions, human actions, land use patterns, and particle size definition, a key factor specific to MP research. The study employs an optimized CatBoost model, and interpretation techniques provide insights into the relative importance of input variables, informing pollution reduction policies. The results indicate the significant effect of inconsistent MP size definitions and highlight the necessity of including data-driven models in urban water management to achieve maximum pollution control measures.

Coauthors include Fereidoun Rezanezhad, Ali Reza Shahvaran, Sarah Kaykhosravi, Stephanie Slowinski, and Philippe Van Cappellen from ERG, and Amirhossein Ghajari from North Carolina State University.

For further details, visit the open-access publication at https://doi.org/10.1038/s41598-025-90612-0