Better Clouds Than Ever With New Exascale Compute-Ready Atmosphere Model


Researchers working on the Energy Exascale Earth System Model (E3SM) have developed a completely new global atmosphere model. The model has a resolution 30 times finer than global climate models. This resolution allows scientists to model and study the atmosphere in much greater detail than before. A new paper describes the equations that govern the new global atmosphere model and evaluates the first simulation of the model.

The impact

This new model is a cornerstone of E3SM exascale computing strategy. This strategy is to take full advantage of future powerful computers to improve climate models. The proof of the new model is therefore an important step of the project. Evaluation of model results validates the E3SM strategy by showing how modeling of storms and topography at the global storm resolution scale corrects many long-standing biases in climate modeling.


Upcoming exascale computers will allow scientists to explicitly resolve deep convection and regional topography in multidecadal simulations. This capability will greatly increase the power of climate predictions. Seizing this upcoming opportunity, E3SM has created a new global atmospheric model designed for the exascale machines planned for the Department of Energy’s (DOE) Leadership Computing Facilities in 2023-2024. This new research presents the new model, describes its governing equations and demonstrates that a prototype version faithfully reproduces the current climate. The researchers simulated the period from January 20 to February 28, 2020, with a horizontal resolution of 3 kilometers around the world. This is 30 times finer than the typical resolution of Global Climate Models (GCMs). The new model significantly mitigates many long-standing systematic errors in standard-resolution GCMs. For example, the new model improves the modeling of precipitation in terms of the timing of its diurnal cycle and the distribution of light versus heavy rain. The new model also captures the pattern of significant weather events, such as tropical and extratropical cyclones, atmospheric rivers, and cold air outbreaks, which are poorly captured by typical GCMs. This paper is an important step towards a revolution in DOE climate modeling where simulations of unprecedented detail and realism pave the way for much more accurate climate predictions.


This research was supported under the E3SM project funded by the DOE Office of Science, Office of Biological and Environmental Research. It used resources from the National Energy Research Scientific Computing Center (NERSC), a DOE Office of Science user facility located at Lawrence Berkeley National Laboratory. This research also utilized the resources of the Argonne Leadership Computing Facility at Argonne National Laboratory.


Comments are closed.