The use of prevalent technologies and crowdsourced data may benefit weather forecasting and atmospheric research, according to a new paper authored by Dr. Noam David, a Visiting Scientist at the Laboratory of Associate Professor Yoshihide Sekimoto at the Institute of Industrial Science, The University of Tokyo, Japan. The paper, published in Advances in Atmospheric Sciences, reviews a number of research works on the subject and points to the potential of this innovative approach.

Specialized instruments for environmental monitoring are often limited as a result of technical and practical constraints. Existing technologies, including remote sensing systems and ground-level tools, may suffer from obstacles such as low spatial representativity (in situ sensors, for example) or lack of accuracy when measuring near the Earth's surface (satellites). These constraints often limit the ability to carry out representative observations and, as a result, the capacity to deepen our existing understanding of atmospheric processes. Multi-systems and IoT (Internet of Things) technologies have become increasingly distributed as they are embedded into our environment. As they become more widely deployed, these technologies generate unprecedented data volumes with immense coverage, immediacy and availability. As a result, a growing opportunity is emerging to complement state-of-the-art monitoring techniques with the large streams of data produced. Notably, these resources were originally designed for purposes other than environmental monitoring and are naturally not as precise as dedicated sensors. Therefore, they should be treated as complementary tools and not as a substitute. However, in the many cases where dedicated instruments are not deployed in the field, these newly available 'environmental sensors' can provide some response which is often invaluable. {module In-article}

Smartphones, for example, contain weather-sensitive sensors and recent works indicate the ability to use the data collected by these devices on a multisource basis to monitor atmospheric pressure and temperature. Data shared as an open source in social networks can provide vital environmental information reported by thousands of 'human observers' directly from an area of interest. Wireless communication links that form the basis for transmitting data between cellular communication base stations serve as an additional example. Weather conditions affect the signal strength on these links and this effect can be measured. As a result the links can be utilized as an environmental monitoring facility. A variety of studies on the subject point to the ability to monitor rainfall and other hydrometeors including fog, water vapor, dew and even the precursors of air pollution using the data generated by these systems.

Notably, the data from these new 'sensors' could be assimilated into high-resolution numerical prediction models, and thus may lead to improvements in forecasting capabilities. Put to use, this novel approach could provide the groundwork for developing new early-warning systems against natural hazards, and generate a variety of products necessary for a wide range of fields. The contribution to public health and safety as a result of these could potentially be of significant value.

Peter Dayan to strengthen partnership between University and Max Planck Institute for Biological Cybernetics

The University of Tübingen is to host another Humboldt Professorship. Peter Dayan, one of the world's leading experts in the field of theoretical neuroscience, will soon be conducting research in the Department of Informatics. Dayan recently became director of the Max Planck Institute for Biological Cybernetics in Tübingen, where he is establishing the Department of Computational Neuroscience and is involved in the restructuring of the institute. Potential cooperation with the University, the hospitals and other research institutions in Tübingen were decisive factors in his decision to come to the institute. Now he will also receive the Alexander von Humboldt Professorship to further intensify this cooperation. Previously, he worked at University College London (UCL). The Humboldt Professorship comes with five million euros over five years. It is Germany’s richest research prize.

Dayan’s research takes in the overlapping fields of neuroscience, medicine and machine learning. His research interests include the question of how the brain makes decisions. Using theoretical models, he has investigated various forms of learning, including reinforcement learning, in which the brain exploits information about previous positive and negative experiences to decide what to do next. Among other things, he analyzed how neuromodulators – chemical messengers such as dopamine, serotonin and acetylcholine – are involved in the decision-making process. {module In-article}

Dayan also investigates how different forms of dysfunctional decision-making are associated with conditions such as depression, addiction, anxiety and personality disorders. He combines the psychological and neural view of such diseases and hopes to gain insights into their causes and possible treatment.

Dayan has developed statistical and programming methods to simulate the brain’s decision-making processes. He has thus helped lay the foundations for the development of artificial neural networks. The 53-year-old will also contribute his knowledge and experience to Tübingen's research on artificial intelligence and machine learning.

"Peter Dayan's fields fit perfectly into Tübingen's research landscape," says Professor Bernd Engler, President of the University of Tübingen. "He combines several of our existing research priorities: neurosciences, clinical research and research on machine learning. His professorship in informatics strengthens the University's cooperation with the Max Planck Campus and other non-university research institutions."

"I am profoundly honoured to receive this professorship. One cannot but be humbled by Alexander von Humboldt's polymathic achievements, and it is a particular delight to be able to help celebrate the 250th year of his birth,” says Dayan. ”I am thrilled to get this opportunity to study both normal and dysfunctional learning and decision-making in a broad and deep way; and am delighted to do so within the rich intellectual environments of the University of Tübingen and the Max Planck Institute for Biological Cybernetics."

The Humboldt Professorship is a bridge to Germany for top international researchers. The award is presented to academics who have already established themselves abroad in their field of research and who declare their willingness to conduct research in Germany for at least five years. Universities nominate the candidates, who are then selected by the Alexander von Humboldt Foundation. Dayan is the fourth Humboldt Professor at the University of Tübingen. Other winners of the award are linguist Professor Rolf Harald Baayen, plant geneticist Professor Marja Timmermans and Geo- and Environmental researcher Professor Lars T. Angenent.

The official award ceremony will take place in May 2020, along with the other prize winners of 2019.