Analysis of spatio-temporal changes in Arctic Ocean ecosystem using machine learning and its impact on marine transportation system

Abstract

Arctic Ocean is large and important ecosystem. Growth of human impact in last decades of XXI century leads to significant changes in different environmental variables. Main goal of this research was analysis of spatio-temporal structure of Arctic Ocean ecosystem to find main causes of its changing and impact on Arctic transportation system. For analysis of Arctic Ocean ecosystem structure, we used Bayesian network, while for spatio-temporal analysis – k-means clustering. This analysis was produced for 3 timestamps – 2007, 2012 and 2017. Main drivers of this ecosystem structure changes generally occur from human activity in this region – shipping and other marine operations. In last years, level of anthropogenic impact significantly increased, large number of territories was turned into agricultural lands, this leads to increase in number of biogenic substances falls into Arctic waters. Spatio-temporal analysis shows very significant changes in last 10 years. Despite the fact, that spatially areas of different zones in Arctic Ocean doesn’t change, qualitative changes here look major. The same as ecosystem structure, the main cause of spatio-temporal changes mainly caused by marine shipping impact.