Convection-resolving climate modeling on future supercomputing platforms (crCLIM)

The crCLIM project aims to develop a European-scale convection-resolving climate modeling capability at a horizontal resolution of about 2 km using the next generation of supercomputing platforms.

Enlarged view: crCLMI animation — Animation of European-scale simulations for a day in July 2006, with a visualization of the cloud cover and precipitation rates (in mm/h). The panels show a 12-km simulation with parameterized convection (left panel) and a 2-km simulation with explicit convection (right panel). Note the more realistic representation of short-term precipitation and the organization of convection in the 2 km simulation (from PhD of David Leutwyler, ETH Zurich, project crCLIM).

Please follow this link to the external page crCLIM webpage.

Major uncertainties remain in our understanding of the processes that govern the water cycle in a changing climate and their representation in weather and climate models. Of particular concern are heavy precipitation events of convective origin (thunderstorms and rain showers).

The crCLIM project attempts to better represent these processes by developing a European-scale climate modeling capability at a horizontal resolution of about 2 km. This resolution is about 10 to 100 times higher than in conventional climate models. With such resolution, it becomes possible to explicitly represent relevant processes (for example the dynamics of deep convective and thunderstorm clouds) without using semi-empirical parameterizations. To achieve its goals, the project will exploit the next generation of high-performance computing architectures.

From a computer science perspective, this development poses major challenges. First, the need for increasing computer power requires the application of codes on emerging hardware architectures that includes heterogeneous many-core architectures consisting of both "traditional" central processing units (CPUs) and accelerators (e.g., GPUs). Second, with increasing computational resolution, the model output becomes unbearably voluminous, which requires new approaches to perform the analysis online rather than storing the model output.

The crCLIM project is highly interdisciplinary as it combines the expertise of climate and computational scientists. Ultimately, crCLIM will lead to a substantial reduction of some of the key uncertainties in the current generation of climate models, yield an improved representation of the water cycle including the drivers of extreme events (heavy precipitation events, floods, droughts, etc.), and enable more sophisticated climate change scenarios. This, in turn, will provide better guidance for impact assessment and climate change adaptation measures.

Partners

The project crCLIM is funded by the external page Sinergia program of the external page Swiss National Science Foundation and coordinated by Christoph Schär. It includes the following partners:

ETH Research Group on Climate and Water Cycle, PI: Christoph Schär
ETH Research Group on Atmospheric Dynamics, PI: Heini Wernli
external page MeteoSwiss, PIs: Oliver Fuhrer, Philippe Steiner
ETH Research Group on Scalable Parallel Computing Laboratory, PI: Torsten Hoefler
external page Swiss National Super Computing Center, PI: Thomas Schulthess

People

The project crCLIM is organised in 4 subprojects which are all intimely connected. In addition, C2SM provides support for the COSMO model development and maintenance.

Subproject A assesses how to efficiently exploit heterogeneous many-core computing architectures for weather and climate models. This involves addressing key questions such as the relative costs and benefits of higher-order numerical accuracy versus higher spatial resolution.
People involved:
    - external page Oliver Fuhrer (lead)
    - Andrea Arteaga
    - external page Christoph Charpilloz
    - Torsten Höfler
    - Thomas Schulthess
Subproject B explores the virtualization of climate simulations from a computer science -perspective. It addresses the critical question of computational versus mass-storage loads, and develops an online analysis platform for high-resolution models.
People involved:
    - Torsten Höfler (lead)
    - Salvatore Di Girolamo
    - Thomas Schulthess
Subproject C performs, validates and analyzes convection-resolving climate simulations over Europe.
People involved:
    - Christoph Schär (lead)
    - Linda Schlemmer
    - Nikolina Ban
    - David Leutwyler
Subproject D exploits the new level of virtualization using Eulerian and Lagrangian perspectives for the online analysis of synoptic features and water transport.
People involved:
    - Heini Wernli (lead)
    - Michael Sprenger
    - Nicolas Piaget
    - Stefan Rüdisühli

Animations

Leutwyler, D., O. Fuhrer, X. Lapillonne, D. Lüthi, C. Schär, 2015: Continental-Scale Climate Simulation at Kilometer-Resolution. ETH Zurich e-collection, DOI: external page http://dx.doi.org/10.3929/ethz-a-010483656 (short description and animation), external page online video on Vimeo.
Leutwyler, D., O. Fuhrer, X. Lapillonne, D. Lüthi, C. Schär, 2015: Winter storm Kyrill in a Continental-Scale Convection-Resolving Climate Simulation. ETH Zurich e-collection, DOI: external page http://dx.doi.org/10.3929/ethz-a-010483662 (short description and animation), external page online video on Vimeo.

Publications

Fuhrer, O., Chadha, T., Hoefler, T., Kwasniewski, G., Lapillonne, X., Leutwyler, D., Lüthi, D., Osuna, C., Schär, C., Schulthess, T.C., and Vogt, H., 2017: Near-global climate simulation at 1 km resolution: establishing a performance baseline on 4888 GPUs with COSMO 5.0. Geosci. Model Develop., Submitted, external page doi.org/10.5194/gmd-2017-230
Belusic, A., Telisman Prtenjak, M., Guettler, I., Ban, N., Leutwyler, D. and Schär, C., 2017: Near-surface wind variability over the broader Adriatic region: Insights from an ensemble of regional climate models. Clim. Dyn., external page doi:10.1007/s00382-017-3885-5
Leutwyler, D., Lüthi, D., Ban, N., Fuhrer, O., and Schär, C., 2017: Evaluation of the Convection-Resolving Climate Modeling Approach on Continental Scales , J. Geophys. Res. Atmos., 122, 52375258, external page doi:10.1002/2016JD026013.
Leutwyler, D., Fuhrer, O., Lapillonne, X., Lüthi, D., and Schär, C., 2016: Towards European-scale convection-resolving climate simulations with GPUs: a study with COSMO 4.19, Geosci. Model Dev., 9, 3393-3412, external page doi:10.5194/gmd-9-3393-2016.