How Satellites And Machine Learning Are Being Used To Identify Plastic In The Ocean
While we know plastic is awful for marine life, detecting plastic contamination in the sea is notoriously challenging. Plastics come in many colors, break down to microscopic sizes, and are created from a variety of chemicals. Adding to the dilemma is the huge size of the sea, to which millions of tons of plastic are added every year.
It’s essential to identify which parts of the ocean accumulate the most plastic to efficiently target cleanup and pollution prevention efforts. Might satellite fortified with machine learning be up to your oceanic job of monitoring plastic pollution?
A group of scientists at the Plymouth Marine Laboratory in the United Kingdom tested whether data from two satellites run by the European Space Agency could be examined using a machine-learning algorithm trained to discover plastic.
The two Sentinel-2 satellites used in this research are each equipped with 12-band Multi-Spectral instrument (MSI) sensors that allow for 10-meter resolution in the information they collect. Together with the efforts of the two satellites united, data is repeatedly collected from all coastal places around the globe every 2 to 5 days.
Satellites collect data on light signals, among other things. Materials could be differentiated using light signals based on what wavelengths of light they reflect. While clear water efficiently absorbs light from the near-infrared (NIR) to shortwave infrared (SWIR) mild selection, floating materials like vinyl and organic debris signify NIR instead. These differences in light absorption allow satellites to detect floating objects in distance.
The NIR signs of various floating objects vary. Using the satellite data, researchers trained a machine-learning algorithm to identify the light sign of drifting plastic by releasing a plastic float off the coast of Greece and getting the associated light sign data in the satellites.
The researched used this light information to teach the algorithm to connect certain NIR light signs with floating plastic debris. Similarly, they also instructed the algorithm to distinguish between natural and plastic materials such as seaweed, driftwood, and seafoam.
The satellites re-visit every coastal region in the world each 2-5 days, collecting a Great Deal of data which…
Overall, the algorithm found plastic with 86% accuracy. Better yet, the algorithm was 100% accurate in its investigation of the information in the Gulf Islands. Not too bad for information collected from thousands of miles over!
Significantly, this algorithm is only equipped to find large floating plastic bits. But it’s from those drifting’macroplastics’ that many harmful micro plastics form. These results show that satellite data along with machine-learning algorithms can help in the tracking, and subsequent clean-up, of plastic pollution around the globe.