Causal Climate Emulation with Bayesian Filtering

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Hickman, Sebastian, Trajkovic, Ilija, Kaltenborn, Julia, Pelletier, Francis, Archibald, Alex, Gurwicz, Yaniv, Nowack, Peer, Rolnick, David, Boussard, Julien
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915574508945408
author Hickman, Sebastian
Trajkovic, Ilija
Kaltenborn, Julia
Pelletier, Francis
Archibald, Alex
Gurwicz, Yaniv
Nowack, Peer
Rolnick, David
Boussard, Julien
author_facet Hickman, Sebastian
Trajkovic, Ilija
Kaltenborn, Julia
Pelletier, Francis
Archibald, Alex
Gurwicz, Yaniv
Nowack, Peer
Rolnick, David
Boussard, Julien
contents Traditional models of climate change use complex systems of coupled equations to simulate physical processes across the Earth system. These simulations are highly computationally expensive, limiting our predictions of climate change and analyses of its causes and effects. Machine learning has the potential to quickly emulate data from climate models, but current approaches are not able to incorporate physically-based causal relationships. Here, we develop an interpretable climate model emulator based on causal representation learning. We derive a novel approach including a Bayesian filter for stable long-term autoregressive emulation. We demonstrate that our emulator learns accurate climate dynamics, and we show the importance of each one of its components on a realistic synthetic dataset and data from two widely deployed climate models.
format Preprint
id arxiv_https___arxiv_org_abs_2506_09891
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Causal Climate Emulation with Bayesian Filtering
Hickman, Sebastian
Trajkovic, Ilija
Kaltenborn, Julia
Pelletier, Francis
Archibald, Alex
Gurwicz, Yaniv
Nowack, Peer
Rolnick, David
Boussard, Julien
Machine Learning
Artificial Intelligence
Computational Engineering, Finance, and Science
Atmospheric and Oceanic Physics
Traditional models of climate change use complex systems of coupled equations to simulate physical processes across the Earth system. These simulations are highly computationally expensive, limiting our predictions of climate change and analyses of its causes and effects. Machine learning has the potential to quickly emulate data from climate models, but current approaches are not able to incorporate physically-based causal relationships. Here, we develop an interpretable climate model emulator based on causal representation learning. We derive a novel approach including a Bayesian filter for stable long-term autoregressive emulation. We demonstrate that our emulator learns accurate climate dynamics, and we show the importance of each one of its components on a realistic synthetic dataset and data from two widely deployed climate models.
title Causal Climate Emulation with Bayesian Filtering
topic Machine Learning
Artificial Intelligence
Computational Engineering, Finance, and Science
Atmospheric and Oceanic Physics
url https://arxiv.org/abs/2506.09891