Stellenangebot / Job offer
Postdoc in Hybrid ML/Physical Modeling for Earth Science
The newly created Model-driven Machine Learning group combines ML- and physics-based approaches to Earth system modeling. We use deep learning to build faster and more accurate components for climate and weather models, generate forecasts and quantify uncertainties. We make use of algorithms and architectures from the cutting edge of ML research, while adapting them to the challenges of large-scale computation and data in Earth science.
This position is designed for a curious and driven Postdoc to focus on challenging and impactful research, supported by extensive on-site expertise and top-notch computational resources, with minimal administrative or teaching duties. The ideal candidate would be an ML expert first and foremost, with knowledge of numerical integration and geoscientific modeling welcome but not essential. However, geoscientists with experience and a publication record in ML are also encouraged to apply. The position is initially limited to three years.
Your tasks:
- develop hybrid methods that combine deep learning with physical models for describing the ocean and atmospheric physics (fluid dynamics, radiation, turbulence, etc.)
- integrate ML-based model components and algorithms in PyTorch with existing climate and weather models in FORTRAN
- validate new algorithms at scale using large-scale observational datasets
- absorb and synthesize knowledge from machine learning, numerical modeling and Earth science
- write, publish and present your research in scientific journals and at conferences
Your profile:
- PhD in mathematics, physics, computer science or the geosciences
- previous ML publications
- solid background in the mathematical foundations of ML (linear algebra, probability, etc.)
- comfortable programming in Python and PyTorch
- experience with HPC/cluster computing (SLURM)
- skills in numerical PDE integration, fluid dynamics, probabilistic ML and/or FORTRAN are welcome
- can carry out research independently with curiosity, creativity, initiative and persistence
- works well in a small group with diverse skill sets and backgrounds
- learns quickly on the job and from research articles on ML and the geosciences
- communicates well in written and spoken English
