- DOE unveils the Exascale Earth Modeling System
Michael Feldman |
April 24, 2018 11:11 CEST
After four years of development, the US Department of Energy (DOE) releases the Energy Exascale Earth System Model (E3SM), a computer-aided platform for high-resolution simulations of weather and other Earth systems.
ES3M is described as "the first end-to-end multiscale Earth system model" that couples simulations of the atmosphere, the land and the ocean (including sea ice). As such, there will be predictions about how these different elements interact. The software is also used to study the interaction between human activities and natural processes. In particular, ES3M can model how human-induced global climate change will affect the various Earth systems.
Simulation of the sea surface temperature during the hurricane in the Atlantic.
Of course, the DOE had such models in the past, but E3SM was developed to take advantage of the greater computational power of exascale systems, and thus more detailed questions on things like drought, sea-level rise, hurricane intensity, and ocean circulation. The reason why all of this is important to the DOE is that all of these processes will have a profound impact on the energy needs and production of solar, wind and hydro power.
"This model adds a more complete picture of the interactions between the energy system and the Earth system," said David Bader, a scientist at the Lawrence Livermore National Laboratory (LLNL) and leader of the E3SM project. "The increase in computational power allows us to add more details to the processes and interactions that lead to more accurate and useful simulations than previous models."
At this point ES3M runs on pre-exascale DOE supercomputers, but the software promises to show its true value as soon as exascale machines become available. The first such system, Aurora, is to be used in 2021
in the Argonne National Lab. Between 2022 and 2023, at least two more DOE exascale supercomputers will go online.
"Today, one of our biggest accomplishments is the model's performance on prototype exascale computers, with the model running faster per node and using less energy than traditional architectures," said Mark Taylor, chief computational scientist for ES3M at Sandia National Laboratories. "This is due to our work in scalability, vectorization, threading and software engineering."
More than 100 scientists and software developers participated in the development of E3SM at several DOE laboratories and a number of universities. Energy laboratories include Argonne, Brookhaven, Lawrence Livermore, Lawrence Berkeley, Los Alamos, Oak Ridge, Pacific Northwest and Sandia. E3SM also relied on collaborations with the Exascale Computing project, SciDAC (Scientific Discovery through Advanced Computing) and Climate Model Development and Validation (CMDV), as well as a number of other DOE funded Earth science programs.
The software, which is managed as a GitHub project, contains code, documentation, and output from an initial set of benchmarks. Detailed information on the model and its simulation tools as well as future research directions can be found on the E3SM website.