Skip to main content

Kilometer-Scale Cess-Potter Experiments with SCREAMv1

Walter
Hannah
LLNL
Christopher Ryutaro Terai, Lawrence Livermore National Laboratory
Peter Martin Caldwell, Lawrence Livermore National Laboratory
Luca Bertagna, Sandia National Laboratories
Hassan Beydoun, Lawrence Livermore National Laboratory
Peter A Bogenschutz, Lawrence Livermore National Laboratory
Andrew Bradley, Sandia National Laboratories
Thomas C Clevenger, Sandia National Laboratories
James G Foucar, Sandia National Laboratories
Wuyin Lin, Brookhaven National Laboratory
Oksana Guba, Sandia National Laboratories
Benjamin R Hillman, Sandia National Laboratories
Jeffrey Johnson, Cohere Consulting LLC
Aaron Sheffield Donahue, Lawrence Livermore National Laboratory
Noel D. Keen, Lawrence Berkeley National Laboratory
Balwinder Singh, Pacific Northwest National Laboratory
Mark A Taylor, Sandia National Laboratories
Jingjing Tian, Lawrence Livermore National Laboratory
Paul Ullrich, Lawrence Livermore National Laboratory
Xingqiu Yuan, Argonne National Lab
Yunyan Zhang, Lawrence Livermore National Laboratory
Jean-Christophe Golaz, Lawrence Livermore National Laboratory
Poster
The Simple Cloud-Resolving Energy Exascale Earth System (E3SM) Atmosphere Model (SCREAM) was created to take full advantage of GPU parallelism and facilitate longer global kilometer-scale experiments with minimal subgrid parameterizations. An outstanding question is whether this type of model will simulate climate feedbacks consistent with 100-km scale climate models. Building on a recent study that found one-year of Cess-Potter type simulations can produce cloud feedbacks similar to those from multi-decade simulations, we will discuss what we have learned from 13-month SCREAM simulations forced with either present-day sea-surface temperatures (SSTs) or with a +4K SST globally uniform perturbation.

The large-scale response generally matches previously reported results with 100-km scale models, suggesting that the lack of a deep convective scheme at the storm-resolving scale does not dramatically alter the response. On the other hand, SCREAM's 3-km configuration shows one of the most positive cloud feedbacks, and hence one of the most positive overall feedbacks, when compared to CMIP6 models. SCREAM simulations run at 12-km resolution also reveal a large resolution sensitivity of the cloud feedbacks. We will discuss the physical underpinnings of the model’s resolution sensitivity and what can be gained from examining climate feedbacks in kilometer-scale modes.
Poster Thumbnail