Z04
Central Supercomputing Support
Dr. Klaus Görgen
Forschungszentrum Jülich | +49 2461 61-6456 |
Prof. Dr. Petra Mutzel
University of Bonn | +49 228 73-69917 |
Dr. Daniel Caviedes-Voullième
Forschungszentrum Jülich | +49 2461 61-4070 |
Dr. Stefan Poll
Forschungszentrum Jülich | - |
Dr. Johannes Keller
Forschungszentrum Jülich | - |
Dr. Marco van Hulten
University of Bonn | +49 228 73-60590 |
Summary
The key objectives of the CRC rely heavily on numerical modelling, data processing and analysis. Project Z04 will be responsible for: (i) Providing basic support for a multitude of HPC-related issues, such as the porting and tuning of scientific software applications (simulation, processing, analysis, and visualization codes) on the latest supercomputers (at Jülich Supercomputing Centre, JSC, and University of Bonn HPC/A-Lab) throughout the entire CRC; (ii) helping with model development through contributions to the developments in the CRC projects and support the performance optimization of workflows towards big data readiness, linked with Z03 tasks; (iii) model simulation support as a service to the CRC, including provenance tracking and configuration management for reproducibility; and (iv) training of early-career scientists at the interface of geoscience, HPC, and data science in collaboration with Z02 and Z03. Hence, Z04 will help the CRC efficiently conduct its demanding numerical model and simulation experiments, as well as processing and analysis tasks, in a big data context.
Contribution to the CRC
Z04 contributes to the central hypothesis by enabling the CRC to efficiently conduct its demanding simulation experiments, as well as processing and analysis tasks in a big data context, and by helping to ensure that setups and configurations are linked with each other and aligned to each other, which is crucial to determine cause and effect relationships in the assessment of the impact of human water use and LULCC vs GHG-driven regional climate change.
Approach
Z04 essentially collaborates through its service and support tasks with all projects of clusters A, B, C, and D, that have any modelling or data analytics task, always in close collaboration with the data project Z03 / INF. Information management, sharing of common configurations, setups, work environments, and workflow, as well as the joint development, discussions, and testing, and finally the deployment of software tools and simulation components is entirely through git repositories, which helps to ensure transparency and reproducibility.
Main results in 2022
- HPC training day for early career DETECT staff
- HPSC Fall School, week long training on HPC and terrestrial system modelling
- Deployment of central GitLab DETECT infrastructure
- Deployment of wiki pages, grid definitions and standards for the entire CRC
- Contribution to DETECT reference model implementations (grids, configurations, etc.)
- Support for JSC compute time applications
- Fortnightly meetings, supporting all HPC matters (compilation, scaling, software environments, etc.)
Main results in 2023
- Initial 3km pan-European model setup and test simulations with TSMP1 (Goergen et al., 2023, AMS)
- Z04 supported frontier simulations with TSMP1 ongoing; linking to international initiatives, such as EURO-CORDEX CMIP6 downscaling (Goergen et al., 2023, ICRC)
- Initial successful MSA (modular supercomputer architecture) simulations on heterogeneous systems across CPUs and GPUs with TSMP1 (Caviedes-Voullieme, et al., 2025)
- In-depth coupling refinements of TSMP2 model system (Poll et al., 2023, CLM-GA and CW2023)
- Scientific computing and HPC service and support, information management, communication
- Support for JSC compute time applications
- HPSC Fall School 2023 contribution
Main results in 2024
- Finalization of major TSMP2 improvements (Poll et al., 2024, AGU; Ho-Hagemann et al., 2024, GMD)
- Development of water budget and analysis tools on the icosahedral ICON grid
- Contribution to the finalized reference configuration for ICON 12km convection-parametrized climate simulations configuration with the CLM-Community (Climate Limited-Area Modelling, WG EVAL, COPAT2 initiative)
- Extension of TSMP2-PDAF to ingest new observational inputs
- Scientific computing and HPC service and support, information management, communication (Figure 1)
- Support for JSC compute time applications
- Contribution to HPSC Fall School 2024
- Contribution to 6th ICTP water resources workshop in Trieste, Italy
Main results in 2025
- TSMP2 simulation and development platform extensions, improvements, and testing towards publication preparation (TSMP2 interface, complete with built system, workflow engine, external / static file generators, auxiliary tools) (Figure 2)
- Contribution to the finalized reference configuration for ICON convection-permitting climate simulations configuration with the CLM-Community (Climate Limited-Area Modelling, WG CRCS, CECPI-Ger initiative)
- Frontier simulations of DETECT/D02 finished with Z04 advanced support; simulations are part of the EURO-CORDEX CMIP6 RCM climate change ensemble with the expanded GCM-RCM matrix (Goergen et al., 2024, EGU; Zhang et al., 2024, Clim. Dynam.)
- Scientific computing and HPC service and support, information management, communication
- Support for JSC compute time applications
- HPSC Fall School 2025 contribution
Selected References
Peer-reviewed journal publications
- Ho-Hagemann, H. T. M., Maurer, V., Poll, S., & Fast, I. (2024). Coupling the regional climate model ICON-CLM v2.6.6 to the Earth system model GCOAST-AHOI v2.0 using OASIS3-MCT v4.0. Geoscientific Model Development, 17(21), 7815–7834. https://doi.org/10.5194/gmd-17-7815-2024
- Zhang, Y., Wagner, N., Goergen, K., & Kollet, S. (2024). Summer evapotranspiration-cloud feedbacks in land-atmosphere interactions over Europe. Climate Dynamics, 1–17. https://doi.org/10.1007/s00382-024-07475-w
Conference Presentations
- Caviedes-Voullième, D., Koh, S.-R., Poll, S., Suarez, E., Arakawa, T., Nakajima, K., & Sumimoto, S. (2025). Portability of Multiphysics Applications on Heterogeneous Modular Supercomputers. In R. Wyrzykowski, J. Dongarra, E. Deelman, & K. Karczewski (Eds.), Parallel Processing and Applied Mathematics (Vol. 15581, pp. 47–62). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-85703-4_4
- Goergen, K., Wagner, N., Poll, S., & Kollet, S. (2023). A km-Scale Pan-European Groundwater-to-Atmosphere Regional Climate System Model Setup to Study the Impact of Human Interventions on the Terrestrial Water Cycle. AMS 103rd Annual Meeting, Denver, USA, 8 Jan 2023 - 12 Jan 2023.
- Goergen, K., Wagner, N., Poll, S., Hartick, C., Zhang, Y., Caviedes Voullieme, D. & Kollet, S. (2023) Coupled groundwater-to-atmosphere simulations with the regional climate system model TSMP as a contribution to the new European CORDEX-CMIP6 ensemble. International Conference on Regional Climate, ICRC-CORDEX 2023, Trieste, Italy, 25 Sep 2023 - 29 Sep 2023.
- Goergen, K., Poll, S., Zhang, Y., & Kollet, S. (2025). Future evolution of European terrestrial water resources in the groundwater-to-atmosphere regional climate system model TSMP1. EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17892. https://doi.org/10.5194/egusphere-egu25-17892, 2025.
- Poll, S., Caviedes Voullieme, D., Goergen, K., and Kollet, S. (2023). Flux exchange over heterogeneous land surfaces. CLM-Community assembly, CLMcom2023, Leuven, Belgium, 18-22 Sep 2023.
- Poll, S., Rigor, P. K., Caviedes Voullieme, D., Goergen, K., Kollet, S. (2023): Land-atmosphere coupling with ICON and CLM5.0 in TSMP using OASIS3-MCT. Workshop on Coupling Technologies for Earth System Models, CW2023, Toulouse, France, 18-20 Jan 2023.
- Poll, S., Rigor, P. K., van Hulten, M., Patakchi Yousefi, K., Caviedes Voullieme, D., Goergen, K., & Kollet, S. (2024). The new version of the Terrestrial Systems Modeling Platform (TSMP2) based on ICON, eCLM, and ParFlow. American Geophysical Union Annual Meeting, AGU24, Washington D.C., USA, 9 Dec 2024 - 13 Dec 2024.
Poster and Presentation
