1 Northern Region Persistent Organic Pollutant Control (NRPOP) Laboratory, Faculty of Engineering and Applied Science, Memorial University of Newfoundland, St. John's, NL, A1B 3X5, Canada. zhengwangdk@outlook.com.
2 Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, QC, H3G 1M8, Canada. chunjiang.an@concordia.ca.
3 Department of Chemical Engineering, McGill University, Montreal, QC, H3A 0G4, Canada.
4 Department of Civil Engineering, Queen's University, Kingston, ON, K7L 3N6, Canada.
5 Kenneth Lee Research Inc, Halifax, NS, B3H 4H4, Canada.

The widespread use of plastics and their improper disposal have released a large number of micro- and nano-plastics (MNPs) into various environmental media. Although the release of MNPs from individual plastic products has been widely reported, there is a lack of a holistic assessment framework to determine the overall release of plastic products to soil, water, and air during their life cycle. Therefore, based on big data, neural network algorithms, and material flows, a new open platform for the comprehensive assessment of the release of MNPs from plastic products will be developed. The proposed emission inventory platform consists of three main modules: a global polymer product production dataset, an assessment of the emission processes, influencing factors, and emission factors of MNPs, and an emission inventory of MNP releases to the environment. The global data on polymer production, use, and waste disposal, and collate data on the degradation behavior of different plastic types under various environmental conditions will be collected. Next, big data analysis will be applied to train the patterns of MNP production and emissions, and algorithms such as neural networks will be used to simulate the complex processes and mechanisms of MNP emissions. Finally, a comprehensive emission inventory model will be established. The proposed dynamic MNPs emission assessment platform integrates material flow analysis and experimentally validated release kinetics. Utilizing machine learning techniques and laboratory and field datasets, the platform can derive dynamic, environment-specific emission factors to support specific emission estimates, source prioritization, and targeted emission reduction strategies.