Regional Climate and Weather Modeling on the Next Generations High-Performance Computers: Towards Cloud-Resolving Simulations

Climate change is one of the most pressing social and economic issues currently facing humankind, but also one of the most complex scientific challenges. Over the last decades, numerical models have become a central methodological element of climate research. They are increasingly used for simulations of higher complexity and at higher resolution, which require enhanced computational resources. The rapidly growing computational power of modern HPC systems allows addressing this challenge. However, as the underlying hardware architectures are changing dramatically, existing numerical weather prediction (NWP) and climate codes need to be adapted for efficient usage of these new architectures.

This proposal aims at improving the COSMO and CCLM models for next-generation high-performance computers. The COSMO model is a research and NWP model that is applied and further developed by several national weather services (including MeteoSwiss). The COSMO-CLM (CCLM) model corresponds to the COSMO model in CLimate Mode and is used at research institutions (e.g., ETH, Empa) for climate research. In this project we propose three interrelated tasks:

Task (1) aims at the near-term development of high-resolution cloud-resolving climate modeling capability by extending the versatility of CCLM, and to apply this new tool for the generation of climate change scenarios.

Task (2) addresses the refactoring of the current version of the CCLM model to better exploit current hardware architectures with a much larger number of processors. The current code will be enhanced with a hybrid MPI/Open MP parallelization and parallel data input/output strategies.

Task (3) involves a more aggressive shift toward future technologies. We will rewrite the CCLM dynamical core to improve the data layout for emerging architectures and employ state-of-the-art software engineering to ensure that the resulting code can be easily ported to emerging hardware.

• Dr. Isabelle Bey, ETH Zurich
  
• Dr. Dominik Brunner, Empa
• Dr. Oliver Fuhrer, MeteoSwiss
• Prof. Ulrike Lohmann, ETH Zurich
  
• Prof. Christoph Schär, ETH Zurich
  
• Dr. Ulrich Schättler, Deutscher Wetterdienst
• Prof. Thomas Schulthess, ETH Zurich
  
• Prof. Sonia Seneviratne, ETH Zurich
  

Not applicable yet