Optimizing Subsampling Methods for Microplastic Analysis
Penn collection
Degree type
Discipline
Data Science
Subject
Funder
Grant number
Copyright date
Distributor
Related resources
Author
Contributor
Abstract
Microplastics (MPs) are emerging contaminants in aquatic, terrestrial, and biological matrices whose degradation and bioaccumulation in food chains due to water and air pollution harm plants, animals, and humans alike. Common MP detection involves FTIR & Raman spectroscopy, or fluorescence microscopy. All these methods can be labor-intensive, especially since MP sizes often follow a power law with excessive smaller plastics. Thus, subsampling methods are often employed, which examine smaller regions of interest to enable extrapolating of data to save time. This project analyzes several geometric subsampling layouts for MPs captured on a filter, specifically using fluorescence imaging, for which there has been limited work comparing. Across these layouts, the most accurate layouts tended to be those with radial orientations, specifically wedges that may be able to capture the differences in MP concentration along the radius. With more samples and samples with higher MP counts, more precise conclusions can be made in comparing layouts.