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

Few studies have effectively shown how to use satellites that gather optical data to monitor plastic debris in the marine environment. For the first time, floating macro-plastics distinguishable from seaweed are identified in optical data from the European Space Agency's Sentinel-2 satellites. Case studies from three Brazilian areas, selected for suspected macro-plastics in Sentinel-2 data, utilized a unique Floating Debris Index (FDI) for the Sentinel-2 Multi-Spectral Instrument (MSI) to detect surface material patches. Sub-pixel-scale detection revealed macro-plastics mixed with seaweed and sea foam. Using a Machine Learning-based Naive Bayes algorithm, we classified materials and identified macro-plastics, achieving an 87.25 % accuracy in identifying suspected plastics. Temporal analysis tracked plastic debris movement and accumulation. This methodology is scalable and transferable, with potential applications for monitoring marine plastic pollution in other coastal regions globally.