David Ingelfinger
· 18.12.2025
The research centres on the automation of scientific analysis methods for environmental protection. The aim is to use AI to overcome delays and establish a reliable data basis in real time.
In the German-Danish joint project "RECOVER" scientists at the GEOMAR Helmholtz Centre for Ocean Research Kiel are developing a virtual image of the south-western Baltic Sea together with partners from southern Denmark and the Christian-Albrechts University.
This so-called "digital twin" makes it possible to record environmental changes and simulate different scenarios. The research centres on the prediction of potential climate change and the effects of agricultural nutrient inputs on ecosystems. Microalgae serve as biological measuring instruments. According to the researchers, their reaction to light, temperature and salinity allows direct conclusions to be drawn about the ecological health of the sea.
A combination of high-resolution camera systems and DNA analysis methods are used to provide the "digital twin" with reliable data. While sophisticated cameras on the research vessels identify microalgae species in real time, so-called "environmental DNA analysis" makes it possible to detect even the smallest traces of biological material in the water.
All information from the sea flows directly into the virtual image of the Baltic Sea. The AI-supported platform is designed to recognise patterns and report changes at an early stage. GEOMAR hopes that this will significantly increase the efficiency of its research.
The project involves numerous network partners from local authorities, tourism and industry in order to ensure that water quality is recorded across the board. The researchers' aim is to sensitise the public to the state of the oceans.
The nationwide project "AI National Park" uses artificial intelligence in 13 national parks to automatically record ecological data. Special sensors have also been in use in the Lower Saxony Wadden Sea National Park since September 2025. Camera traps and audio and climate sensors continuously collect information about biodiversity and the impact of human activities in the region.
The evaluation is carried out by an artificial intelligence system that analyses large volumes of image and sound material. First, the system's sensors automatically recognise their surroundings. The AI then identifies specific sounds, such as bird calls, bat calls or human disturbances. The aim is to ultimately obtain a reliable picture of the actual situation on site.
The researchers hope to collect precise and up-to-date data in this way without placing additional pressure on the animal world through human presence. The project provides a new, reliable data basis for environmental protection on land.