How close are we to global km-scale weather and climate simulations?
A recent study investigates how close we are towards achieving global km-scale weather and climate simulations. A baseline of what is achievable today is established using a fully refactored model code on Europe’s largest supercomputer Piz Daint at the Swiss National Supercomputing Centre CSCS in Lugano.
The best hope for reducing long-standing global weather and climate model biases is by increasing resolution to the kilometer scale. But how far are we away from achieving such simulations? A external pagerecent collaborative studycall_made of a team of MeteoSwiss, ETH and C2SM researchers, led by Dr. Oliver Fuhrer, has addressed this question by performing simulations using the non-hydrostatic regional climate model external pageCOSMOcall_made 5.0 for a near-global setup running on the full external pagePiz Daint supercomputercall_made on 4’888 GPUs (graphics processing units). The model code has been refactored for performance portability across different hardware architectures. Results at a grid spacing of 1.9 km indicate a baseline of what is already achievable today on Europe’s largest supercomputer
Reference
Fuhrer, O., Chadha, T., Hoefler, T., Kwasniewski, G., Lapillonne, X., Leutwyler, D., Lüthi, D., Osuna, C., Schär, C., Schulthess, T. C., and Vogt, H.: Near-global climate simulation at 1 km resolution: establishing a performance baseline on 4888 GPUs with COSMO 5.0, Geosci. Model Dev., 11, 1665-1681, external pagehttps://doi.org/10.5194/gmd-11-1665-2018call_made, 2018