Scientists Create World’s Largest Database of Arctic Microplastic Spectra
The new Microscan database will help neural networks distinguish polyethylene from polypropylene directly in the ocean with more than 96% accuracy.
Scientists at Novgorod State University and the P.P. Shirshov Institute of Oceanology of the Russian Academy of Sciences have compiled what is now the world’s largest open database of infrared microplastic spectra and the first database of its kind focused on the Arctic.
The database is based on particles found in surface waters of the White, Barents, Kara, East Siberian and Laptev seas in 2020–2021.
Researchers previously faced a problem because most available sources described only “ideal” granules of pure plastic. Real marine microplastics age, become covered with bacteria and crack under sunlight, making them harder to identify. Each particle was examined under infrared light without chemical cleaning, and the results were verified through spectral matching, artificial intelligence and manual review by experts. The database will allow neural networks to learn from particles collected “in the wild” and speed up the identification of plastic waste during expeditions.
