A study published today in Nature Scientific Reports reports a new method for detecting large plastic (> 5 mm) floating garbage patches in the Marine environment. Using data from the European Space Agency’s Sentry 2 satellite, the method was able to distinguish plastic from other materials with 86 percent accuracy.
Based on the wavelengths of visible and infrared light absorbed and reflected by the floating objects, the scientists identified the floating belt in the Sentry 2 data. They then developed a machine-learning algorithm that classifies the individual materials that make up these floating belts based on specific spectral characteristics of different plastics and natural materials.
These features are derived from satellite data on plastic waste that washed up in the Port of Durban, South Africa, on April 24, 2019, and floating plastic that the authors deployed off the coast of Mytilini, Greece, in 2018 and 2019. They also used previously available satellite data on natural materials such as algae, wood, foam and volcanic rock that may have been found with Marine plastics.
The authors tested the method using Sentry 2 data from four different coastal areas. The method was able to distinguish plastic from other floating materials or seawater with an average accuracy of 86 percent, in some cases 100 percent. The authors hope this approach could be used in conjunction with drones or high-resolution satellites to improve global monitoring of plastic waste in the oceans while helping with clean-up operations.