Photo credit: phys.org
Advancements in Climate Modeling with SCREAM
The Simple Cloud-Resolving E3SM Atmosphere Model (SCREAM) represents a significant leap in climate modeling by offering enhanced resolution of cloud dynamics. Results from SCREAM have been matched against visible satellite imagery from the Himawari-8 satellite to analyze events such as cold-air outbreaks in Siberia and cyclones south of Australia.
Climate models play a crucial role in understanding Earth’s climate system by simulating a wide range of phenomena, from minute water droplets to expansive weather systems. By accurately balancing Earth’s energy across the atmosphere, land, oceans, and ice, scientists can determine the driving factors behind climate change and make informed predictions for the future.
The complexity of climate modeling necessitates the division of the Earth’s atmosphere into a three-dimensional grid comprising numerous cells. Modern general circulation models (GCMs) utilize a grid system dividing the atmosphere into roughly 1° squares across the horizontal dimension, extended through about 60 vertical levels. While this is an advancement over earlier models, such resolution can still introduce significant uncertainties in predictions.
A newer generation of models known as global storm-resolving models (GSRMs) utilizes grids with cells measuring only 5 kilometers across. This finer resolution facilitates a deeper understanding of extreme weather events, particularly tropical cyclones, by providing greater detail. However, GSRMs still face challenges, particularly in accurately portraying cloud characteristics close to the Earth’s surface and capturing the vertical movements of heat and moisture within the atmosphere. Additionally, the complexity of these models often results in slower processing times and higher computational costs when using conventional hardware.
In a recent study published in the Journal of Advances in Modeling Earth Systems, A. S. Donahue and colleagues examined SCREAM, a GSRM that balances high resolution with high computational speed, designed to operate on exascale computing systems capable of performing over 10^18 computations per second. This level of performance is currently achieved by only the world’s fastest supercomputers. The study involved running four 40-day seasonal simulations on a grid with a resolution of 3 kilometers, comparing the outcomes with data gathered from both satellites and ground observations.
The initial version of SCREAM successfully re-established the global energy balance to within an impressive accuracy of 1.2 watts per square meter. It also effectively modeled midlatitude jet streams and atmospheric rivers that transport moisture to polar regions. Although SCREAM demonstrated the ability to capture diurnal variations in boundary clouds across all seasons, it encountered difficulties in resolving midlevel cloud formations, particularly in tropical areas. Despite being in the developmental phase, SCREAM presents a promising avenue for more efficient high-resolution climate simulations.
Further Reading
More information:
A. S. Donahue et al, “To Exascale and Beyond—The Simple Cloud‐Resolving E3SM Atmosphere Model (SCREAM), a Performance Portable Global Atmosphere Model for Cloud‐Resolving Scales,” Journal of Advances in Modeling Earth Systems (2024). DOI: 10.1029/2024MS004314
This article is republished courtesy of Eos, hosted by the American Geophysical Union. Read the original story here.
Citation:
Modeling Earth systems at a quintillion calculations per second (2024, August 7) retrieved 7 August 2024 from Phys.org.
This document is provided for information purposes only and is subject to copyright restrictions. No reproduction is allowed without permission, except for fair dealing for private study or research.
Source
phys.org