Machine Learning Approach for Detection of Water Overgrowth in Azov Sea with Sentinel-2 Data

The Azov Sea estuaries play an important role in the reproduction of semi-anadromous fish species. Spawning efficiency is closely connected with overgrowing those species spawning grounds; thus, objective water vegetation research has vital fisheries importance. Thus, main goal was to develop a machine learning algorithm for detection overgrowth Phragmites australis based on Sentinel-2 data. conducted field botanical and investigations 2020–2021 Soleniy Chumyanniy firths. Collected remote sensing data were processed semi-automatic classification plugin QGIS. For estuaries, random forest used. obtained results showed that 2020 areas occupied by reeds reached 0.37 km2, while 2021, they increased 0.51 km2. There high level growth rapid period 2020–2021, where area covered reed doubled, primarily attributed eutrophication. This due nutrient enrichment from agricultural lands located northern part near Novonekrasovskiy village. Additionally, changes flows hydrological conditions can also contribute favorable reed. result rate australis, which reach up 2 m per year propagate both through vegetative sexual means, leading formation large dense clusters.