Artificial Intelligence for Improved Baltic Sea Protection

– Joint Press Release of the GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel University, and the State Chancellery of Schleswig-Holstein –

The bays of Schleswig-Holstein's Baltic coast are not in good ecological condition. Seasonal oxygen-depleted zones are expanding, leading to more frequent mass fish kills. The data used to assess the state of the environment is currently often based on so-called "discrete" measurements, i.e. water samples are taken on site and then taken to the laboratory for analysis. Results can take days or even months to become available.

In the future, artificial intelligence will be used to collect, assess and develop measures to improve environmental conditions in near-real time: The pilot project INSYST (INtelligent SYSTem for Coastal Water Monitoring using Artificial Intelligence), jointly developed by GEOMAR Helmholtz Centre for Ocean Research Kiel and Kiel University, aims to implement this innovative monitoring concept specifically for the Eckernförde Bay. Dirk Schrödter, Schleswig-Holstein's Minister for Digitalisation, has now awarded the project a grant of €750,000 from the state's AI strategy.

"Environmental protection, digitalisation and artificial intelligence are inextricably linked," says Digitalisation Minister Dirk Schrödter. "AI and data are at the centre of our efforts to make Schleswig-Holstein a digital flagship region. To achieve this goal, we are promoting AI projects in business, science and administration. We are focusing on the maritime sector. The INSYST AI project enables efficient assessment of the current water quality on our doorstep in the Bay of Eckernförde and can be used as an early warning system if the condition deteriorates. The project thus impressively demonstrates the direct benefits for biodiversity and the people of Schleswig-Holstein.

Professor Dr Katja Matthes, Director of GEOMAR, emphasises the groundbreaking importance of the project: "The funding of this project is an important step towards more effective protection of the Baltic Sea. The integration of artificial intelligence into marine monitoring concepts is a promising approach, but one that has hardly been used so far. With INSYST, we can make a decisive and innovative contribution to the development of such concepts, which will help to protect and preserve our marine ecosystems in the future".

"We are very pleased about the successful funding decision by the state. It is a special recognition of the fact that we at Kiel University can use our innovative methods and research in artificial intelligence to contribute to solving societal challenges such as improving the health of the oceans and better assessing risks in the future," said Professor Dr Ralph Schneider, Vice-President of Kiel University.

“For the project we will use the data from the Boknis Eck time series station,” says project leader Dr Helmke Hepach, environmental scientist at GEOMAR, who is responsible for the data collection together with Professor Anja Engel and Professor Hermann Bange. Since 1957, Boknis Eck has been regularly sampled monthly for a variety of physical, chemical and biological parameters. For several years, these series of measurements have been supplemented by sensor data from a permanently installed underwater observatory, which can be accessed live via a data cable. After the loss of the sensor structure in summer 2019, it will now be reinstalled at Boknis Eck. In addition, a small-scale computer model of the Baltic Sea is available at GEOMAR, which provides a good representation of the physical parameters in Eckernförde Bay.

“Conventional statistical methods cannot be applied to these large data sets,” says Professor Olaf Landsiedel from Kiel University, who is responsible for the AI methodology in the project, "this is where artificial intelligence comes in." Not only can AI link and make use of these large data sets, but it also becomes more accurate as the data sets grow. By combining the data with various AI methods, an assessment of the state of the environment can be made almost in real time. This also allows hypothetical scenarios and measures for environmental improvement to be tested in advance in order to act more effectively and sustainably. “We hope that this will enable us to make increasingly accurate predictions and warnings,” says Dr Helmke Hepach.

The project could also support the monitoring efforts of the relevant authorities in Schleswig-Holstein. For example, it will help to identify suitable locations for assessing the environmental status of the Baltic Sea ecosystem or to improve the protection of marine habitats.

The data and results will also be made available to all interested parties through a communication concept. An app will allow individuals to participate in the data collection with their own observations.

Project funding:

AI funding guideline of the state of Schleswig-Holstein