High Accuracy at Low Cost: Using Machine Learning to Emulate the Department of Energy's Global Cloud-Resolving Climate Model

Peter Caldwell | 22-ERD-052

Executive Summary

We will use machine-learning-based nudging to emulate the Department of Energy's global cloud-resolving atmosphere model using coarse-resolution simulations. The resulting model will combine the accuracy of very fine resolution with the speed needed to separate climate-change trends from weather noise.

Publications, Presentations, and Patents

Peter Caldwell, “Climate Prediction at LLNL,” (Presentation, to Geraldine Richmond, Under Secretary for Science and Innovation at the Department of Energy, Livermore, CA, June 23, 2023).

Peter Caldwell, "The E3SM Atmosphere Model in C++ (EAMxx)" (Presentation, Energy Exascale Earth System model Workshop, Denver, CO, June 20, 2023).