Data-driven monitoring and alerts to protect the Arctic
Arctic sea ice is shrinking by 74 000 square kilometres per year, averaging a loss of 13 % per decade. As the region becomes more accessible, there has been an increase in maritime activities such as shipping and fishing. While this economic activity can benefit local inhabitants, it brings with it risks. Uncontrolled fishing can lead to the decimation of stocks, while the extraction of natural resources such as fossil fuels can have significant environmental impacts. It is critical that such activities are carefully monitored, to ensure the protection of this fragile ecosystem, but this is challenging. “Extreme weather and rapidly changing land and sea conditions have often made it difficult to detect changes in the Arctic,” explains ARCOS project coordinator, Massimo Sernicola from e-GEOS, a joint venture between the Italian Space Agency and spaceflight services company Telespazio. He adds: “The opening up of shipping routes and access to natural resources has also raised new security concerns.”
Detecting suspicious behaviour
The ARCOS project sought to strengthen EU situational awareness of the Arctic region by making better use of satellite data. The project created a comprehensive monitoring system that combines Earth observation (EO) data from Copernicus satellites with non-EO data, AI and human expertise. This system offers a global picture of Arctic trends, as well as early warnings. “The multi-component platform processes EO data, and then combines it with other sources such as sensory data and human intelligence,” explains Sernicola. Users can activate specific services based on their knowledge and needs, and can design reports and dashboards based on their analyses. The platform can be used to detect suspicious behaviour, such as unusual vessel movements, and alert users immediately when such anomalies are detected.
Surveillance in the Arctic
Several pilots were implemented to test and validate the system, including maritime surveillance and land activity monitoring. “We used the system to identify ‘dark vessels’ that don’t broadcast their position,” notes Sernicola. “We were also able to detect changes in specific sites of interest on land, such as the size of industrial plants.” These pilots demonstrated how the ARCOS system can provide analysts and regulatory bodies with new insights into unusual vessel behaviour and other activities. The detection of icebergs, for example, can help to minimise the risk to ship traffic. Such insights are critical for maintaining security and environmental protection in the Arctic. “We were able to show how EO data, non-EO data and AI can be successfully integrated, to produce useful spatial information,” says Sernicola. “The platform we developed can be customised to meet various user needs, be it for research or operational purposes. This adaptability makes ARCOS a valuable tool for continuous monitoring in the Arctic.”
Improved AI models
The ARCOS system is an important development in enriching the Copernicus Service on Support to EU External and Security Actions (SESA). This European geospatial information service aims to assist the EU and its Member States by enhancing situational awareness in security-related domains. Moving forward, Sernicola identifies the development of advanced AI models that can handle the complex environmental conditions of the Arctic as a key challenge. The lack of training data that specifically targets the Arctic is an important topic that still needs to be addressed. “Looking ahead, we need to further refine our AI models and include more data sources,” he says. “We need this to improve our capabilities in vessel detection, to discriminate vessel from rocks or icebergs, for example. This will provide for even greater accuracy and predictive capabilities.”
Keywords
ARCOS, Copernicus, Arctic, fishing, shipping, ecosystem, weather, satellite, Earth observation, iceberg, AI