Plastic debris detection along coastal waters using Sentinel-2 satellite data and machine learning techniques.

Area of Science:

• Remote Sensing
• Marine Pollution Monitoring
• Oceanography

Background:

• Limited success in using optical satellite data for marine plastic debris detection.
• Need for effective methods to monitor widespread plastic pollution.

Purpose of the Study:

• To identify and map floating macro-plastics using optical satellite data.
• To develop and validate a novel methodology for marine plastic detection.

Main Methods:

• Utilized Sentinel-2 satellite optical data and a Floating Debris Index (FDI).
• Applied a Machine Learning-based Naive Bayes algorithm for material classification.
• Conducted sub-pixel-scale detection and temporal analysis of debris.

Main Results:

• Successfully distinguished floating macro-plastics from seaweed and sea foam in optical data.
• Achieved 87.25% accuracy in identifying suspected plastic debris.
• Tracked the movement and accumulation patterns of plastic pollution.

Conclusions:

• The developed methodology is effective for detecting marine macro-plastics using Sentinel-2 data.
• The approach is scalable and transferable for global coastal pollution monitoring.
• Demonstrated a novel application of remote sensing for environmental surveillance.