Postdoctoral Researcher in Analysis of Space Weather using AI/ML

Catholic University of Leuven

The Centre for mathematical Plasma-Astrophysics (CmPA) of the KULeuven has received funding from the Belgian federal government (DEFRA project AIDefSpace) to study and develop modern AI/ML techniques to detect and forecast dangerous activity in space weather, considering ionosphere, maghnetosphere and solar active regions. We are processing different data sources relative to the ionosphere and the magnetosphere and high-resolution magnetograms and multi-spectral images of the Sun. Using advanced machine learning techniques, we want to make nowcasts and forecasts of space weather and discover signatures of flaring activity and other potentially dangerous conditions. The tools developed will be compared to existing models and proposed as new services for the Belgian and European space weather centres.

Responsibilities

The postdoctoral researcher will work on the development of machine learning techniques for safe weather applications based on neural networks and unsupervised learning. The researcher will have the opportunity to explore and test multiple modern machine learning techniques, in particular the use of supervised deep neural networks, of autoencoders, and unsupervised techniques like Self-Organizing Maps, and Gaussian Mixture Models. The work will also enable the researcher to explore the physics of the datasets we use for solar images, ionospheric and magnetospheric conditions. The researcher will take a leadership role in this project and will guide, with their expertise, the selection of the AI/ML algorithms used in the project. The researcher will also have the freedom within the scope of the project to direct the research towards the methods she or he will find most interesting and promising.

Profile

Required:

- PhD in Space Science, Physics, Engineering or in AI/ML with a background in computer sciences applied to physics.

- Bases on machine learning, including neural networks, data processing and unsupervised learning.

- Knowledge on Python programming

- Knowledge of a neural network framework (PyTorch, Tensorflow)

- Knowledge of a domain of physics related to space sciences: e.g. particle physics, plasma physics, aerospace engineering, fluid dynamics.

Desirable:

- Experience in HPC

- Experience in applied mathematics

- Experience in solar physics, space plasma physics, and/or space weather

- Experience in image analysis and the respective Python tools

- Expereince in parallel programming using Python (or alternatively a parallel framework for ML, like Horovod or PyTorch Lightning)

Offer

We offer a 2 year contract in a full-time position with a competitive salary. The contract can be further extended depending on performance evaluation.

Interested?

For more information please contact Prof. dr. ir. Giovanni Lapenta, tel.: +32 16 32 79 65, mail: giovanni.lapenta@kuleuven.be.

You can apply for this job no later than March 31, 2023 via the
KU Leuven seeks to foster an environment where all talents can flourish, regardless of gender, age, cultural background, nationality or impairments. If you have any questions relating to accessibility or support, please contact us at diversiteit.HR@kuleuven.be.
  • Employment percentage: Voltijds
  • Location: Leuven
  • Apply before: March 31, 2023
  • Tags: Wetenschappen
  • Download: file