Early warning system for critical changes in ecosystems
June 27, 2016
International experts met at the Max Planck Institute for Biogeochemistry in Jena, Germany, to discuss the state of terrestrial ecosystems. Their research project focusses on describing and understanding the state of ecosystems and how ecosystems have changed over the recent past. The scientists rely on a combination of novel satellite data, ground observations and innovative methods from informatics. The ultimate goal is to develop an early warning system for detecting critical environmental transitions.
Copernicus Sentinel data (2015): The image shows the area east of Lago di Bolsena in Italy acquired by the European Radar Remote Sensing Satellite “Sentinel-1a”. Radar measures tell about the physical properties of the Earth surface by transmitting and receiving electromagnetic waves. With long wavelengths we are able to detect and analyze different surface properties, such as the moisture content, surface roughness, etc., without any restrictions due to cloud cover or daytime. The picture shows the amount of the received backscatter (echo reflected from the object) for different land cover types. As water and unmanaged agriculture areas (dark areas) has very low backscatter intensities, cities, settlements and forests (bright areas) are resulting in higher backscatter of the radar signal. (MU)
Ongoing climate change and continuous land use changes are intimately related to societal, political and economic processes. Jointly, these processes lead to changes of ecosystem functioning at the local, regional and global scale. Such ecosystem transformations are a matter of concern since the 1980s and have led to integrated research programs often considering ecosystem research, climate sciences and social sciences. Today, however, new opportunities for research emerge in view of the dramatic progress in the quality of Earth Observations with respect to spatial, temporal and spectral resolution (in particular from ESA’s sentinel satellites). Likewise, informatics increasingly develops means to deal with large amounts of data and effective processing tools. Thus, old questions regain relevance: Can we automatically detect critical changes of ecosystems by combining novel satellite data, ground observations and innovative methods of machine learning? How can we use scientific findings resulting from these innovations to better inform society?
Such questions were addressed in a meeting of 35 scientists of the EU-funded "BACI" project, which were coming from Germany, Great Britain, Spain, Italy, Denmark, Canada and Switzerland. The meeting took place at the Max Planck Institute for Biogeochemistry in Jena, Germany, from June 7th to 9th. The aims were to synchronize efforts to develop and implement an early warning system for changes and disturbances at the interface between terrestrial ecosystems and the atmosphere.
Initial results of the research project suggest that new satellite data, local ecosystem observations and machine learning can be combined as a promising methodology. “Compared to earlier times, we are now well prepared to achieve a more comprehensive detection and understanding of ecological extreme events, such as e.g. droughts or heat waves,” says Dr. Miguel Mahecha, leader of the BACI project, “and we put a special focus on biosphere - atmosphere interactions and feedbacks in the Earth System.”
The project’s most ambitious goal is to develop an early warning index to detect sudden changes and unusual occurrences in the data streams. “We especially want to find those anomalies, which are of high relevance for the functioning of terrestrial ecosystems, such as changes in the CO2 and water fluxes between the biosphere and the atmosphere”, Dr. Mahecha explains. Such an early warning system for ecosystem changes should allow scientists and users to develop specific management and adaptation strategies at an early stage of changes.
The BACI project “Detecting changes in essential ecosystem and biodiversity properties – towards a Biosphere Atmosphere Change Index” is financed by the European Commission´s Horizon 2020 research and innovation program. The project is coordinated by the Max Planck Institute for Biogeochemistry in Jena, Germany. In concert with the 10 European project partners, the University of Jena substantially contributes to the success of the project with their research on analysis of novel satellite data and on innovative methods of machine learning. The project exemplifies that today research in this area needs to embrace multiple branches of science across disciplinary boundaries. In Jena this has been conceptually realized with the foundation of the Michael Stifel Center for Data-driven and Simulation Sciences where the University and the Max Planck institutes join their expertise. [MM/EF/DF]
Webpage BACI project
Webpage Michael Stifel Center
Webpage Friedrich Schiller University Jena
Press Release (pdf)