Description
Leader of the group Multiscale Dynamics: Ute Ebert.
Nature and technology are full of dynamics, often involving multiple scales in length, time and energy. To model these processes, we combine scientific computing with model reduction and machine learning, with particular focus on plasma dynamics in lightning and space weather, and in high voltage and plasma technology.
Our research addresses questions in nature such as start and propagation of lightning strokes, terrestrial gamma-ray flashes and space weather, and closely related technological problems such as switch gear for long-distance electricity nets, air purification and disinfection with corona reactors, and protection of satellites and electricity nets from space weather.
Within national and European projects, we collaborate with colleagues in applied plasma physics, electrical and mechanical engineering, atmospheric electricity, and cosmic particle and space science, and with non-academic partners such as ABB, DNV-GL, ESA-ESTEC and NASA.
Find more about our work (including publications) on the personal homepages of the staff scientists:
- Ute Ebert (group leader)
- Enrico Camporeale
- Jannis Teunissen (starting 1 October 2018)
and on the page with our numerical codes for Multiscale Plasma Dynamics.
View a photo of the Multiscale Dynamics group.
Vacancies
PhD student on computational models for lightning physics
The Multiscale Dynamics group has a vacancy for a PhD student on computational models for lightning physics.
Postdoc on the subject of Machine Learning and Space Weather
The position involves research in machine learning techniques and Bayesian inference, applied to real-time forecasting of energetic electrons in the Earth radiation belts. These electrons can be harmful to satellites, whose disruption can potentially lead to catastrophic societal and economic events. The emerging field of Space Weather is concerned with making accurate predictions of such dangerous events, sufficiently in advance so that countermeasures can be taken. The aim of this project is to advance our space weather prediction capability by enhancing physics based models with a new data-driven probabilistic framework. The project will involve state-of-the-art numerical simulations, Bayesian parameters estimation, uncertainty quantification and machine learning techniques.
News
Machine Learning Techniques for Space Weather book published
Recently, the book 'Machine Learning Techniques for Space Weather', edited by CWI researcher Enrico Camporeale and others, was published by Elsevier. The book provides “a thorough and accessible presentation of machine learning techniques that can be employed by space weather professionals”.
Technology grant awarded to optimize plasma-assisted combustion
Incomplete combustion in engines and central heating systems can generate toxic exhaust gases. An electric discharge or plasma can preprocess the mixture of air and fuel, which could largely improve the combustion process. Researchers from CWI and TU/e were awarded a technology grant from NWO to understand and improve this.
Efficient models for high voltage switching
In the search for better environmentally friendly high-voltage switches, it is important to understand electrical breakdowns in various media. PhD student Ashutosh Agnihotri (CWI) developed efficient computational models to simulate these phenomena in different gas mixtures. He defends his PhD thesis at Eindhoven University of Technology on 25 June.
AIDA: Exploring space with Artificial Intelligence
The EU H2020 project 'AIDA' on using artificial intelligence in space physics was signed in May. The Multiscale Dynamics Group at CWI is one of the eight research partners involved. The ultimate goal is to create an intelligent system that enables new scientific discoveries in space.
Members
Associated Members
Publications
-
Teunissen, H.J, & Ebert, U. M. (2018). Afivo: A framework for quadtree/octree AMR with shared-memory parallelization and geometric multigrid methods. Computer Physics Communications. doi:10.1016/j.cpc.2018.06.018
-
Camporeale, E, Sorriso-Valvo, L, Califano, F, & Retinò, A. (2018). Coherent structures and spectral energy transfer in turbulent plasma: A space-filter approach. Physical Review Letters, 120. doi:10.1103/PhysRevLett.120.125101
-
Hare, B.M, Scholten, O, Bonardi, A, Buitink, S, Corstanje, A, Ebert, U. M, … Winchen, T. (2018). LOFAR lightning imaging: Mapping lightning with nanosecond precision. Journal of Geophysical Research - Atmospheres. doi:10.1002/2017JD028132
-
Bliksemschicht veroorzaakt gammaflitsen - New Scientist - 31-01-18. (2018). Bliksemschicht veroorzaakt gammaflitsen - New Scientist - 31-01-18.
-
Elkholy, A, Shoshyn, Y, Nijdam, S, van Oijen, J, van Veldhuizen, E.M, Ebert, U. M, & de Goey, L.P.H. (2018). Burning velocity measurement of lean methane-air flames in a new nanosecond DBD microplasma burner platform. Experimental Thermal and Fluid Science, 95, 18–26.
-
Camporeale, E, Wing, S.P, Johnson, J.R, Jackman, C.M, & McGranaghan, R. (2018). Space weather in the machine learning era: A multidisciplinary approach. Space Weather (Vol. 16, pp. 2–4). doi:10.1002/2017SW001775
-
Rutjes, C, Diniz, G, Ferreira, I.S, & Ebert, U. M. (2017). TGF afterglows: A new radiation mechanism from thunderstorms. Geophysical Research Letters, 44(20), 10702–10712. doi:10.1002/2017GL075552
-
Bliksem produceert nagloed van gammastraling - Engineersonline.nl - 23-10-2017. (2017). Bliksem produceert nagloed van gammastraling - Engineersonline.nl - 23-10-2017.
-
A natural neutron source - Physics World - 01-10-2017. (2017). A natural neutron source - Physics World - 01-10-2017. Physics World, October 2017, 13–15.
-
Het weer donder en bliksem. Koning der wolken krijgt hulp uit de ruimte - Trouw - 20-07-2017. (2017). Het weer donder en bliksem. Koning der wolken krijgt hulp uit de ruimte - Trouw - 20-07-2017.
Software
Afivo: A framework for adaptive mesh refinement with geometric multigrid methods
Afivo, short for Adaptive Finite Volume Octree, is a framework for adaptive mesh refinement on quadtree (2D) and octree (3D) meshes with shared memory parallelization (OpenMP) and geometric multigrid methods.
PumpKin: Analyzing complex chemistry models
PumpKin, short for Pathway Reduction Method for Plasma Kinetic Models, is a tool for post-processing results from zero-dimensional plasma kinetics solvers.