Emulating the forced response of climate models with generative machine learning
- 1INRIA
- 2MILA
- 3Karlsruhe Institute of Technology (KIT)
Abstract
Global climate models are essential tools to simulate past and potential future pathways of climate change, as well as associated climate impacts. Shared Socioeconomic Pathways (SSPs) describe a range of future scenarios of global economic and demographic development. These SSPs are intrinsically linked to changes in climate forcings—the external drivers, such as greenhouse gas and aerosol emissions, which change Earth’s energy balance over time. These forcings act as boundary conditions in Earth System Models (ESM), providing insight into the potential climatic impacts of each SSP. Running an ESM, however, is extremely computationally expensive, conflicting with the need for large ensemble runs to provide robust estimates in the presence of internal variability and scenario uncertainty. Machine Learning emulators provide a promising avenue toward fast and cheap scenario generation, but until recently lacked uncertainty quantification and the ability to condition on external forcings. Here, we build upon recent work \cite{clyneArchesClimateProbabilisticDecadal2025} and extend it by training on multiple SSPs. We successfully generate scenarios of IPSL-CM6A-LR unseen during training and that remain largely physically consistent with the underlying climate model, including under moderate extrapolation. Under an abrupt-4×CO2 step, however, the model does not reproduce the transient climate response. Our emulator is validated against MESMER-M, a statistical emulator of land surface temperature. Our research demonstrates that our model, ArchesClimate-SSP, does not simply imitate scenarios seen during training, but is actually capable of modeling the response of a climate state to diverse forcings. This is an important step toward reliable and rapid climate model scenario generation.
Related
- Quantifying nonlinear thresholds of grey-green space morphology on urban carbon metabolism across local climate zones: a machine learning study in Changchun city
- A machine learning model for early season wildfire risk in Ghana’s Guinea savannah assessing discrimination, operational prevalence, spatial transferability, and CMIP6 climate scenarios
- Regional climate risk assessment from climate models using probabilistic machine learning
- Advancing spatial hurricane storm surge risk assessment using machine learning and simulations under changing climate scenarios
- Response to comment on “Using genomic data and machine learning to predict antibiotic resistance: A tutorial paper”
BibTeX
@article{clyne2026emulating,
title = {{Emulating the forced response of climate models with generative machine learning}},
author = {Graham Clyne and Julia Kaltenborn and Peer Johannes Nowack and Claire Monteleoni and Anasatase Charantonis},
journal = {Machine Learning Earth},
year = {2026},
month = sep,
doi = {10.1088/3049-4753/aeac2f},
url = {https://doi.org/10.1088/3049-4753/aeac2f},
note = {CC BY},
}
RIS
TY - JOUR
TI - Emulating the forced response of climate models with generative machine learning
AU - Graham Clyne
AU - Julia Kaltenborn
AU - Peer Johannes Nowack
AU - Claire Monteleoni
AU - Anasatase Charantonis
PY - 2026
DA - 2026/09/24
JO - Machine Learning Earth
DO - 10.1088/3049-4753/aeac2f
UR - https://doi.org/10.1088/3049-4753/aeac2f
AB - Global climate models are essential tools to simulate past and potential future pathways of climate change, as well as associated climate impacts. Shared Socioeconomic Pathways (SSPs) describe a range of future scenarios of global economic and demographic development. These SSPs are intrinsically linked to changes in climate forcings—the external drivers, such as greenhouse gas and aerosol emissions, which change Earth’s energy balance over time. These forcings act as boundary conditions in Earth System Models (ESM), providing insight into the potential climatic impacts of each SSP. Running an ESM, however, is extremely computationally expensive, conflicting with the need for large ensemble runs to provide robust estimates in the presence of internal variability and scenario uncertainty. Machine Learning emulators provide a promising avenue toward fast and cheap scenario generation, but until recently lacked uncertainty quantification and the ability to condition on external forcings. Here, we build upon recent work \cite{clyneArchesClimateProbabilisticDecadal2025} and extend it by training on multiple SSPs. We successfully generate scenarios of IPSL-CM6A-LR unseen during training and that remain largely physically consistent with the underlying climate model, including under moderate extrapolation. Under an abrupt-4×CO_2 step, however, the model does not reproduce the transient climate response. Our emulator is validated against MESMER-M, a statistical emulator of land surface temperature. Our research demonstrates that our model, ArchesClimate-SSP, does not simply imitate scenarios seen during training, but is actually capable of modeling the response of a climate state to diverse forcings. This is an important step toward reliable and rapid climate model scenario generation.
SN - 3049-4753
ER -