A study by German scientists from Jena and Hamburg, published today in the journal Nature, shows that artificial intelligence (AI) can substantially improve our understanding of the climate and the Earth system. Especially the potential of deep learning has only partially been exhausted so far. In particular, complex dynamic processes such as hurricanes, fire propagation, and vegetation dynamics can be better described with the help of AI. As a result, climate and Earth system models will be improved, with new models combining artificial intelligence and physical modeling.
In the past decades mainly static attributes have been investigated using machine learning approaches, such as the distribution of soil properties from the local to the global scale.
Anomalies in the CO2 exchange on land during El Niño years, based on AI-upscaled FLUXNET data. Radiation anomalies are shown in red, temperature anomalies in green and water anomalies in blue.
Graphik: Martin Jung, MPI-BGC
For some time now, it has been possible to tackle more dynamic processes by using more sophisticated deep learning techniques. This allows for example to quantify the global photosynthesis on land with simultaneous consideration of seasonal and short term variations.
Deducing underlying laws from observation data
“From a plethora of sensors, a deluge of Earth system data has become available, but so far we've been lagging behind in analysis and interpretation,” explains Markus Reichstein, managing director of the Max Planck Institute for Biogeochemistry in Jena, directory board member of the Michael-Stifel-Center Jena (MSCJ) and first author of the publication.
“This is where deep learning techniques become a promising tool, beyond the classical machine learning applications such as image recognition, natural language processing or AlphaGo” adds co-author Joachim Denzler from the Computer Vision Group of the Friedrich Schiller University Jena and member of MSCJ.
Examples for application are extreme events such as fire spreads or hurricanes, which are very complex processes influenced by local conditions but also by their temporal and spatial context. This also applies to atmospheric and ocean transport, soil movement and vegetation dynamics, some of the classic topics of Earth system science.
Artificial intelligence to improve climate and Earth system models
However, deep learning approaches are difficult. All data-driven and statistical approaches do not guarantee physical consistency per se, are highly dependent on data quality, and may experience difficulties with extrapolations. Besides, the requirement for data processing and storage capacity is very high.
The publication discusses all these requirements and obstacles and develops a strategy to efficiently combine machine learning with physical modeling. If both techniques are brought together, so-called hybrid models are created. They can for example be used for modeling the motion of ocean water to predict sea surface temperature.
While the temperatures are modelled physically, the ocean water movement is represented by a machine learning approach. “The idea is to combine the best of two worlds, the consistency of physical models with the versatility of machine learning, to obtain greatly improved models,” Markus Reichstein further explains.
The scientists contend that detection and early warning of extreme events as well as seasonal and long-term prediction and projection of weather and climate will strongly benefit from the discussed deep-learning and hybrid modelling approaches.
Prof. Dr. Markus Reichstein
Tel: +49 (0)3641 57 6200
Deep learning and process understanding for data-driven Earth system science
Reichstein M., Camps-Valls G., Stevens B., Jung M., Denzler J. Cavalhais N., Prabhat (2019)
Nature xx, doi: 10.1038/s41586-019-0912-1
https://www.bgc-jena.mpg.de/bgi/index.php/Main/HomePage Webpage of the Reichstein department at MPI-BGC
http://www.inf-cv.uni-jena.de/denzler Webpage of Prof. Denzler at FSU Jena
Dr. Eberhard Fritz | Max-Planck-Institut für Biogeochemie
Atmospheric scientists reveal the effect of sea-ice loss on Arctic warming
11.03.2019 | Institute of Atmospheric Physics, Chinese Academy of Sciences
11.03.2019 | University of Tokyo
Nano- and microtechnology are promising candidates not only for medical applications such as drug delivery but also for the creation of little robots or flexible integrated sensors. Scientists from the Max Planck Institute for Polymer Research (MPI-P) have created magnetic microparticles, with a newly developed method, that could pave the way for building micro-motors or guiding drugs in the human body to a target, like a tumor. The preparation of such structures as well as their remote-control can be regulated using magnetic fields and therefore can find application in an array of domains.
The magnetic properties of a material control how this material responds to the presence of a magnetic field. Iron oxide is the main component of rust but also...
Due to the special arrangement of its molecules, a new coating made of corn starch is able to repair small scratches by itself through heat: The cross-linking via ring-shaped molecules makes the material mobile, so that it compensates for the scratches and these disappear again.
Superficial micro-scratches on the car body or on other high-gloss surfaces are harmless, but annoying. Especially in the luxury segment such surfaces are...
The Potsdam Echelle Polarimetric and Spectroscopic Instrument (PEPSI) at the Large Binocular Telescope (LBT) in Arizona released its first image of the surface magnetic field of another star. In a paper in the European journal Astronomy & Astrophysics, the PEPSI team presents a Zeeman- Doppler-Image of the surface of the magnetically active star II Pegasi.
A special technique allows astronomers to resolve the surfaces of faraway stars. Those are otherwise only seen as point sources, even in the largest telescopes...
Researchers at Chalmers University of Technology and the University of Gothenburg, Sweden, have proposed a way to create a completely new source of radiation. Ultra-intense light pulses consist of the motion of a single wave and can be described as a tsunami of light. The strong wave can be used to study interactions between matter and light in a unique way. Their research is now published in the scientific journal Physical Review Letters.
"This source of radiation lets us look at reality through a new angle - it is like twisting a mirror and discovering something completely different," says...
New research group at the University of Jena combines theory and experiment to demonstrate for the first time certain physical processes in a quantum vacuum
For most people, a vacuum is an empty space. Quantum physics, on the other hand, assumes that even in this lowest-energy state, particles and antiparticles...
11.03.2019 | Event News
01.03.2019 | Event News
28.02.2019 | Event News
21.03.2019 | Life Sciences
21.03.2019 | Physics and Astronomy
21.03.2019 | HANNOVER MESSE