Humboldt Research Award Winner: Renowned US scientist comes to TU Ilmenau

The renowned US physicist Professor Michael Chertkov from the University of Arizona Tucson will be working at Technische Universität Ilmenau for the next two years as part of a Humboldt Research Award. Together with Prof. Jörg Schumacher from TU Ilmenau and Prof. Markus Reichstein from the Max Planck Institute for Biogeochemistry in Jena, the scientist has been researching artificial intelligence for calculating turbulent flows, which are required for modeling weather and climate phenomena, since mid-July. The Humboldt Research Award is endowed with prize money of 60,000 euros.

Every year, the Alexander von Humboldt Foundation awards up to 100 Humboldt Research Awards to leading international scientists from all scientific disciplines. With the award money of 60,000 euros, they can conduct research at German universities for up to 12 months together with their hosts. One of the 2025 award winners, Prof. Michael Chertkov from the University of Arizona Tucson , is currently conducting research on Artificial Intelligence together with Prof. Jörg Schumacher, head of the Fluid Mechanics Group at TU Ilmenau, and Prof. Markus Reichstein from the Max Planck Institute for Biogeochemistry in Jena next year.

During Prof. Chertkov’s research stay, the three scientists will develop and apply AI algorithms that can be used to model the atmospheric transport of carbon dioxide and other chemical compounds in global Earth system models or to investigate turbulent flows using high-resolution numerical simulations. These so-called Lagrangian algorithms follow the paths of small, almost massless particles through the observed flow and describe the mixing and material transport in the turbulence - for example, the countless soot particles of a plume of smoke that emerges from a chimney, spreads out by following the flow and finally dissolves seemingly "into nothing". These generative algorithms - computational processes based on a specific resit pattern - can generate such flow paths synthetically, even if only a few flow parameters are given to them, without having to start up large, data-intensive simulations.

Prof. Michael Chertkov was honored with the Humboldt Research Award for his outstanding work on fundamental questions of machine learning for artificial intelligence and the application of Lagrangian techniques in turbulence research: graphical methods for modeling electrical networks, for example. Chertkov is one of the first researchers to incorporate physical laws into the largely data-driven algorithms of machine learning with deep neural networks, a method now known as physics-informed machine learning. He has received numerous awards for his scientific achievements since his doctoral studies, including the prestigious R. H. Dicke Postdoctoral Fellowship from Princeton University and a Robert Oppenheimer Fellowship from the Los Alamos National Laboratory.

Prof. Michael Chertkov received his doctorate in theoretical physics in 1996 from the Weizmann Institute of Science Rehovot in Israel. His further scientific career took him to the USA via the prestigious Princeton University to the Los Alamos National Laboratory and finally to the University of Arizona Tucson, where he has held a professorship at the Department of Applied Mathematics since 2019 and heads the doctoral program. Prof. Chertkov is a passionate cross-country and marathon runner and, during his stay in Ilmenau, enjoys morning runs of several kilometers through the hills of the Thuringian Forest in his free time.