Emulating the Forced Response of Climate Models with Flow Matching

Graham Clyne, Julia Kaltenborn, Peer Nowack, Claire Monteleoni, Anasatase Charantonis · 2026

Read Emulating the Forced Response of Climate Models with Flow Matching by Graham Clyne, Julia Kaltenborn, Peer Nowack, Claire Monteleoni, Anasatase Charantonis (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-16929.txt.

Introduction

Clyne's team builds ArchesClimate-SSP, an emulator that mimics how a climate model responds to external forcings across scenarios. Shared Socioeconomic Pathways sketch economic and demographic futures, and their greenhouse-gas and aerosol emissions shift Earth's energy balance, but running an Earth system model is far too costly for large ensembles. Trained on several pathways, the emulator reproduces unseen IPSL-CM6A-LR scenarios and stays physically consistent even under strong extrapolation. Checked against the statistical emulator MESMER-M, it models climate-state responses rather than copying seen scenarios.

Catalog record

CREATORGraham Clyne, Julia Kaltenborn, Peer Nowack, Claire Monteleoni, Anasatase Charantonis
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE73194
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.16929

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Emulating the Forced Response of Climate Models with Flow Matching Graham Clyne, Julia Kaltenborn, Peer Nowack, Claire Monteleoni, Anasatase Charantonis 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 in turn lead to the human impact on Earth’s energy balance over time. These forcings are fundamental boundary conditions in climate models in order to gain insight into the potential climatic impacts of these changes described by each SSP. Running a climate model, however, is extremely computationally expensive, conflicting with the need for large ensembles of simulations for each model to give, e.g., more robust estimates in the presence of internal variability (the inherent, chaotic fluctuations within the climate system) and scenario uncertainty. Recent research has demonstrated the ability to capture climate model dynamics using machine learning when conditioned on forcings from different climatic scenarios. We here train a Deep Learning (DL) model on multiple SSPs and successfully generate scenarios unseen during training. Our emulator is validated against MESMER-M, a statistical emulator of land surface temperature. Our research demonstrates the capacity to generate such changing climate states in response to a variety of simultaneous climate forcings (e.g., carbon dioxide, methane, nitrous oxide, sulphate aerosols, and ozone). In particular, our ablation studies underline a need to include a range of different forcings to represent long-term atmospheric trends with a DL emulator. 1 Introduction Humanity is facing the consequences of a changing climate, ranging from vanishing ecosystem services (benefits that humans derive from the natural environment) to increased exposure to extreme events, putting the lives of humans, flora, and fauna at risk [ 61 ] . To move forward, we need to mitigate, adapt, and develop our future sustainably [ 15 ] . A cornerstone of climate change adaptation is humanity’s capacity to project the future of our climate under different scenarios. Policymakers need to understand how different emission plans might affect Earth’s climate to make informed decisions on behalf of the groups they represent. For decades, scientists have contributed to understanding Earth’s future climate by developing climate models [ 12 ] . Climate models started by modeling only the atmosphere and ocean, but recently they have developed into Earth System Models (ESMs), i.e. they simulate different parts of the Earth in tandem, using physical equations that formalize what we know about how the atmosphere, ocean, cryosphere and biosphere operate and interact [ 49 , 48 ] . ESMs are run for many different “what-if” experiments, helping scientists better understand the Earth’s climate [ 12 , 27 ] . One set of such experiments is called ScenarioMIP [ 60 ] , which compares the climate outputs of different ESMs under different future scenarios. These experiments essentially ask “how will our Earth’s climate look like if humanity behaves in a certain way?”. Those future scenarios of human behaviour are called Shared Socio-Economic Pathways (SSPs) [ 66 ] . They describe the future in terms of economic, societal, and demographic changes over the coming decades and as such also formulate assumptions about future emissions of greenhouse gases and aerosols. SSPs have further implications for the secondary formation of other forcing agents through atmospheric chemical reactions, in particular tropospheric and stratospheric concentrations of ozone and concerning the atmospheric oxidizing capacity (relevant for the degradation of greenhouse gases such as methane). Changes in anthropogenic greenhouse gases and aerosols expected from each SSP are provided through Input4MIPs , a coordinated activity that provides standardized, version-controlled datasets, providing time-dependent emission maps or atmospheric concentration fields that are then used as inputs for the ESMs participating in ScenarioMIP [ 53 ] . In summary, we can, e.g., input a future emission scenario into an ESM to simulate its long-term climate response for a given SSP. The resulting projections have several types of uncertainty. In the near-term, i.e. 30 years, model uncertainty (how unsure we are about the output of the model) is a major component of the the total uncertainty, followed by the climate system’s internal variability (the natural, stochastic fluctuations within the climate system), and a small amount of scenario uncertainty (the range of potential future climate outcomes resulting from different plausible pathways of human activity). As the projection moves farther into the future, scenario uncertainty can start to become the dominant uncertainty given a wide range of SSPs [ 47 , 43 ] . Unlike model uncertainty, which could be narrowed through, e.g., improved representations of model physical processes and computing power, scenario uncertainty is irreducible because it pertains to future human agency. Understanding and quantifying this uncertainty is critical for policymakers; it underpins the exploration of strategies that remain effective across various trajectories. By generating different scenarios, researchers can better demonstrate how immediate policy interventions could directly alter the long-term climate. ESMs are computationally expensive, restricting the generation of a wide variety of future scenarios. This work aims to help climate models in exploring additional scenarios by providing a robust method to quickly generate spatially and temporally explicit future climate simulations that respond appropriately to a multitude of external forcings. In this study, we investigate whether generative Deep Learning (DL) can generate unseen future climate scenarios. We aim to investigate the need to accurately represent a combination of different forcings on the emulated climate scenarios. We use ArchesClimate as the DL framework, an autoregressive probabilistic climate model emulator based on ArchesWeather [ 16 , 17 ] and here train ArchesClimate on a set of climate projections produced with the IPSL-CM6A-LR ESM [ 7 ] . We extend the original ArchesClimate framework by conditioning on sets of forcing agents that represent trajectories under different SSP scenarios. We will refer to the extended model generally as ArchesClimate-SSP, with each specific forcing setup being indicated as AC_{forcing\_setup} , where forcing\_setup will be defined in subsection 4.2 . We train our model using monthly averages from 2015-2100 for 8 different future scenarios, in addition to historical runs and control runs. We held out three SSP scenarios (SSP4-3.4, SSP5-3.4 and SSP5-8.5) for testing. We structure the paper as follows: we first provide a review of the current state of climate model emulation. We then explain the methods used in our research. Following this, we show that we can accurately emulate SSP scenarios unseen at training time. To do so, we first evaluate if the generated scenarios represent temporal and spatial variability of the IPSL climate model, and we compare our results to MESMER-Ms, a state-of-the-art spatial statistical emulator, [ 57 , 4 ] . Finally, we perform ablations of the model by removing different forcings. This shows how different forcings inform the accuracy of ArchesClimate-SSP when generating future scenarios. The contributions of this research are as follows: • We introduce two mechanisms that help long-term stability and flexibility: training with random future timesteps, and using dropout of the forcing conditions. • We show that we are able to train a DL model that correctly reproduces the temporal and spatial dynamics of SSPs

Read the complete text