Project 04: Replacement of physical PSC simulation by machine learning in an Earth system model

Author

Ole Kirner, Jörg Meyer, Achim Streit

Editor

Uwe Ehret, Martin Frank, KIT-Zentrum MathSEE

Participating institute

KIT-Zentrum Mathematik in den Natur-, Ingenieur- und Wirtschaftswissenschaften (KIT-Zentrum MathSEE)
Scientific Computing Center (SCC)
Institut für Wasser und Gewässerentwicklung (IWG)

Genre

Veranstaltung

Description

04 Replacement of physical PSC simulation by machine learning in an Earth system model
MATH PI: Dr. Ole Kirner, Steinbuch Centre for Computing (SCC), Scientific Computing & Mathematics (SCC-SCM)
SEE PI: Dr. Jörg Meyer, Steinbuch Centre for Computing (SCC), Data Analytics, Access and Applications (SCC-D3A)
Department(s): Informatics (Computer Science) or Physics
Type of position: 75% FTE, E13 TV-L
Polar stratospheric clouds (PSCs) exist in winter in the lower/middle atmosphere and are responsible for ozone depletion in the polar spring and the resulting ozone hole.
The goal of the doctoral research is to show that the physical simulation of PSCs within an earth system model can be replaced by an Al model. It will be investigated and evaluated if this enables a realistic simulation of the PSCs and thus improves the performance of the PSC modul as part of the earth system model ICON-ART.
Tasks of the thesis include:
• Performance analysis of the ICON-ART model code at different resolutions on High Performance Computing (HPC) systems
• Creation of a concept for replacing the PSC simulation with a suitable Al model (such as Transformer, LSTM, CNN) including the identification of suitable features dimension reduction, and hyperparameter tuning
• Implementation and evaluation of the procedure and investigation of suitable metrics
• Evaluation of the Al model integrated into ICON-ART (including parallelization)
Requirements for this position:
- Completed studies (master) in computer science, mathematics or physics
- Programming skills (e.g. Fortran, C++, Python)
- Ability to work and publish in a targeted and scientific manner.
- Good communication and presentation skills and willingness and ability to work a team
- Good writing and oral communication skills in English
Optional requirements:
- Knowledge of current deep learning frameworks (e.g. PyTorch or Tensorflow)
- Experience in working with climate/earth system models

Duration (hh:mm:ss)

00:07:18

Series

KCDS Virtual Open House 2023 - Fall

Published on

23.10.2023

Subject area

Computer science

License

KITopen Licence