Decoding the Plastic Patch: Exploring the Global Microplastic Distribution in the Surface Layers of Marine Regions...

Area of Science:

• Environmental Science
• Marine Biology
• Data Science

Background:

• Marine environments face significant microplastic (MP) pollution challenges.
• Comprehensive ocean surveys for MP assessment are logistically difficult and costly.
• Machine learning (ML) offers a viable approach to model MP distribution and impact.

Purpose of the Study:

• To develop and interpret a predictive ML model for marine MP distribution.
• To identify key factors influencing global MP pollution patterns.
• To assess the ecological risk associated with specific MP size ranges.

Main Methods:

• Utilized four ML algorithms with MP data (20-5000 μm size range).
• Incorporated biogeochemical, anthropogenic, atmospheric, and physical factors into the model.
• Employed an interpretable ML framework for data preprocessing, prediction, and factor analysis.

Main Results:

• Biogeochemical and anthropogenic factors were identified as major drivers of marine MP pollution.
• Atmospheric and physical factors showed lesser influence on global MP distribution.
• Predicted global marine MP concentrations ranged from 0.176 to 27.055 particles/m³.
• Microplastics within the 20-5000 μm size range were found to pose no significant ecological risk.

Conclusions:

• The interpretable ML framework provides a robust method for understanding and managing marine MP pollution.
• Factor influence on MP distribution can be region-specific, requiring localized management strategies.
• This study offers critical insights for policymakers and researchers involved in marine conservation and pollution control.