ACE2-NEMO: Coupling an ML atmospheric emulator to a full-depth dynamical ocean model

Bobby Antonio, Kristian Strommen, Pablo Ortega, Hannah M. Christensen

arXiv preprint · 30 March 2026 · ECMWF, University of Oxford

Why it matters. Early multi-decadal coupling of ACE2 atmosphere emulator to NEMO ocean; reveals weak ENSO-like variability, informing ML-based coupled climate modelling.

In one sentence. ACE2 atmospheric emulator is coupled to the NEMO ocean for 70-year runs, giving realistic fast air-sea coupling but unrealistically weak tropical Pacific ENSO-like variability.

Abstract

Understanding how fast atmospheric variability shapes slow climate variability and sensitivity remains a central challenge in Earth-system science. Recent advances in machine-learned (ML) atmospheric models have demonstrated remarkable skill on weather timescales, but their emergent behaviour in a fully coupled climate system remains largely unexplored. We present early results from a new hybrid modelling framework, in which the ACE2 ML atmospheric emulator is interactively coupled to the NEMO ocean model. We report on a set of 70-year coupled simulations (1950-2020 historical forcing and fixed-1950s control). These experiments represent, to our knowledge, one of the first multi-decadal integrations of a machine-learned atmosphere interacting with a full-depth dynamical ocean. Several historical and fixed-1950s control simulations from the fully dynamic global coupled climate model EC-Earth, which has the same ocean component used in ACE2-NEMO, are also considered for comparison. We assess the behaviour of the coupled system, with particular focus on low-frequency tropical variability and the climate response to greenhouse-gas forcing. Analysis of potentially emergent El Nino-like variability reveals realistic fast timescale air-sea coupling in the tropical Pacific, but the temporal variability is unrealistic, with very low amplitude oscillations; this appears to be due to weak atmospheric feedback in the tropical Pacific. The response to CO2 forcing shows initial agreement with EC-Earth3P but deviates due to reduced downward short-wave radiation in ACE2. These results provide a unique test of physical realism and generalisability for atmospheric emulators and evaluate the possible role of entirely machine-learned components in next-generation Earth system models.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.