The Equilibrium Response of Atmospheric Machine‐Learning Models to Uniform Sea Surface Temperature Warming
- 1Now at Rosenstiel School of Marine Atmospheric and Earth Science, University of Miami Miami FL USA
- 2Program in Atmospheric and Oceanic Sciences Princeton University Princeton NJ USA
- 3Princeton University
Abstract
Machine learning (ML) models for the global atmosphere that are capable of producing stable, multi‐year simulations of Earth's climate have recently been developed. However, the ability of these ML models to generalize beyond the training distribution remains an open question. In this study, we evaluate the climate response of several state‐of‐the‐art ML models (ACE2‐ERA5, NeuralGCM, and cBottle) to a uniform sea surface temperature warming, a widely used benchmark for evaluating climate change. We assess each ML model's performance relative to a physics‐based general circulation model (NOAA's Geophysical Fluid Dynamics Laboratory AM4) across key diagnostics, including surface air temperature, precipitation, temperature, and wind profiles, and top‐of‐atmosphere radiation. While the ML models reproduce key aspects of the physical model response, particularly the response of precipitation, some exhibit notable departures from robust physical responses, including radiative responses and land region warming. Our results highlight the promise and current limitations of ML models for climate change applications and suggest that further improvements are needed for robust out‐of‐sample generalization.
Related
- Climate models lack decadal variability in precipitation because of a weak atmospheric response to sea surface temperature anomalies
- Atmospheric Predictability beyond 30 Days with Machine Learning
- SATFRONT: Detecting Atmospheric Fronts with Satellite Data using Machine Learning
- Machine-learning prediction of freezing temperature in Pb-contaminated soils remediated with biochar
- AQUAVERSE-AC: a machine learning-based atmospheric correction framework for inland and coastal waters
BibTeX
@article{zhang2026equilibrium,
title = {{The Equilibrium Response of Atmospheric Machine‐Learning Models to Uniform Sea Surface Temperature Warming}},
author = {Bosong Zhang and Timothy M. Merlis},
journal = {Journal of Advances in Modeling Earth Systems},
year = {2026},
month = sep,
doi = {10.1029/2026ms005815},
url = {https://doi.org/10.1029/2026ms005815},
note = {CC BY},
}
RIS
TY - JOUR
TI - The Equilibrium Response of Atmospheric Machine‐Learning Models to Uniform Sea Surface Temperature Warming
AU - Bosong Zhang
AU - Timothy M. Merlis
PY - 2026
DA - 2026/09/01
JO - Journal of Advances in Modeling Earth Systems
DO - 10.1029/2026ms005815
UR - https://doi.org/10.1029/2026ms005815
AB - Machine learning (ML) models for the global atmosphere that are capable of producing stable, multi‐year simulations of Earth's climate have recently been developed. However, the ability of these ML models to generalize beyond the training distribution remains an open question. In this study, we evaluate the climate response of several state‐of‐the‐art ML models (ACE2‐ERA5, NeuralGCM, and cBottle) to a uniform sea surface temperature warming, a widely used benchmark for evaluating climate change. We assess each ML model's performance relative to a physics‐based general circulation model (NOAA's Geophysical Fluid Dynamics Laboratory AM4) across key diagnostics, including surface air temperature, precipitation, temperature, and wind profiles, and top‐of‐atmosphere radiation. While the ML models reproduce key aspects of the physical model response, particularly the response of precipitation, some exhibit notable departures from robust physical responses, including radiative responses and land region warming. Our results highlight the promise and current limitations of ML models for climate change applications and suggest that further improvements are needed for robust out‐of‐sample generalization.
SN - 1942-2466
ER -