The Energy Exascale Earth System Model (E3SM) is an ongoing Department of Energy (DOE) earth system modeling, simulation and prediction project aiming to “assert and maintain an international scientific leadership position” in those areas. The project, launched in 2014, released E3SMv1 in 2018. Now, three years later, E3SMv2 is here, promising dramatic speed and accuracy improvements.
“E3SMv2 is faster and better than E3SMv1,” said Chris Golaz, an atmospheric scientist at Lawrence Livermore National Laboratory (LLNL), in an interview with LLNL’s Anne M. Stark. “It’s approximately twice as fast on identical machines, four times as fast on machines we have now, compared to what we had for E3SMv1. From one generation to another, earth system models typically become better but also quite a bit slower, so faster and better is significant.”
One of the jobs that E3SMv1 – and, in general, major weather and climate models – have struggled with is the representation of precipitation and clouds, which often operate at resolutions below the multi-kilometer grids that make these models computationally manageable. E3SMv2 handles both precipitation and clouds much better than the first version – specifically, Golaz said, “how clouds change in a warmer climate is much more realistic.”
E3SMv2 comes with two fully coupled configurations: a 100km-grid global atmospheric model and an “RRM” (regionally refined model) with a 25km grid over North America. “Thanks to the performance improvements, the RRM configuration of E3SMv2 runs as fast as E3SMv1 did in its standard resolution configuration (100km) a few years ago,” Golaz said. “We are essentially getting the much higher resolution for ‘free.’”
“E3SMv2 allows us to more realistically simulate the present, which gives us more confidence to simulate the future,” added David Bader, lead of the E3SM project, a scientist at LLNL. “The increase in computing power allows us to add more detail to processes and interactions that results in more accurate and useful simulations than the previous version.”
Lawrence Livermore is just one of many partners on the project, which incorporates over 100 scientists and engineers, multiple universities and a series of other DOE laboratories, including Argonne, Brookhaven, Lawrence Berkeley, Los Alamos, Oak Ridge, Pacific Northwest and Sandia. E3SM also has links with the Exascale Computing Project and other DOE programs.
Now, the researchers are working on conducting a simulation campaign with E3SMv2, with several thousand years already simulated and thousands more to go. The E3SM project is asking affiliated researchers to delay submitting papers pending publication of the core E3SMv2 research, but updates about the project and the ensuing research can be found at the E3SM website here.