Recent trends in marine microplastic modeling and machine learning tools: Potential for long-term microplastic monitoring

Abstract

The increase in the global demand for plastics, and more recently during the pandemic, is a major concern for the future of plastic waste pollution and microplastics. Efficient microplastic monitoring is imperative to understanding the long-term effects and progression of microplastic effects in the environment. Numerical models are valuable in studying microplastic transport as they can be used to examine the effects of different parameters systematically to help elucidate the fate and transport processes of microplastics, thus providing a holistic view of microplastics in the ocean environment. By incorporating physical parameters (such as size, shape, density, and identity of microplastics), numerical models have gained better understanding of the physics of microplastic transport, predicted sinking velocities more accurately, and estimated microplastic pathways in marine environments. However, availability of large amounts of information about microplastic physical and chemical parameters is sparse. Machine learning and computer-vision tools can aid in acquiring environmental information and provide input to develop more accurate models and verify their predictions. More accurate models can further the understanding of microplastic transport, facilitate monitoring efforts, and thus optimize where more data collection can take place to ultimately improve machine learning tools. This review offers a perspective on how image-based machine learning can be exploited to help uncover the physics of microplastic transport behaviors. Additionally, the authors hope the review inspires studies that can bridge the gap between numerical modeling and machine learning for microplastic analysis to exploit their joined potential.