EUVIMEDObjevte EUVIMED
Open access

Climate-invariant machine learning

Tom Beucler iD, Pierre Gentine iD, Janni Yuval iD, Ankitesh Gupta iD, Liran Peng, Jerry Lin iD, Sungduk Yu iD, Stephan Rasp, Fiaz Ahmed, Paul A O'Gorman iD, J David Neelin iD, Nicholas J Lutsko iD and 1 more

Science Advances · 2024-02-09 · 10 (6)

DOI: 10.1126/sciadv.adj7250 ↗

At a glance

Journal article in Science Advances, published on 2024-02-09. Written by Tom Beucler and 12 co-authors. Institution: Faculty of Geosciences and Environment, University of Lausanne, Lausanne, VD 1015, Switzerland. Cited 110 times so far (Crossref); 113 references. The full text is freely available in PubMed Central.

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

Projecting climate change is a generalization problem: We extrapolate the recent past using physical models across past, present, and future climates. Current climate models require representations of processes that occur at scales smaller than model grid size, which have been the main source of model projection uncertainty. Recent machine learning (ML) algorithms hold promise to improve such process representations but tend to extrapolate poorly to climate regimes that they were not trained on. To get the best of the physical and statistical worlds, we propose a framework, termed “climate-invariant” ML, incorporating knowledge of climate processes into ML algorithms, and show that it can maintain high offline accuracy across a wide range of climate conditions and configurations in three distinct atmospheric models. Our results suggest that explicitly incorporating physical knowledge into data-driven models of Earth system processes can improve their consistency, data efficiency, and generalizability across climate regimes.

Authors and institutions

  • Faculty of Geosciences and Environment, University of Lausanne, Lausanne, VD 1015, Switzerland.
  • Department of Earth System Science, University of California, Irvine, CA 92697, USA.
  • Department of Earth and Environmental Engineering, Columbia University, New York, NY 10027, USA.
  • Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA.
  • Google Research, Mountain View, CA 94043, USA.
  • Department of Atmospheric and Oceanic Sciences, University of California, Los Angeles, Los Angeles, CA 90095, USA.

ORCID profiles: Tom Beucler, Pierre Gentine, Janni Yuval, Ankitesh Gupta, Jerry Lin, Sungduk Yu, Paul A O'Gorman, J David Neelin, Nicholas J Lutsko, Michael Pritchard

Cite

APA

Beucler, T., Gentine, P., Yuval, J., Gupta, A., Peng, L., Lin, J., Yu, S., Rasp, S., Ahmed, F., O'Gorman, P. A., Neelin, J. D., Lutsko, N. J., & Pritchard, M. (2024). Climate-invariant machine learning. Science Advances, 10(6). https://doi.org/10.1126/sciadv.adj7250

Vancouver

Beucler T, Gentine P, Yuval J, Gupta A, Peng L, Lin J, et al. Climate-invariant machine learning. Science Advances. 2024;10(6). doi:10.1126/sciadv.adj7250

BibTeX
@article{Beucler2024,
  author = {Beucler, Tom and Gentine, Pierre and Yuval, Janni and Gupta, Ankitesh and Peng, Liran and Lin, Jerry and Yu, Sungduk and Rasp, Stephan and Ahmed, Fiaz and O'Gorman, Paul A and Neelin, J David and Lutsko, Nicholas J and Pritchard, Michael},
  title = {Climate-invariant machine learning},
  journal = {Science Advances},
  year = {2024},
  volume = {10},
  number = {6},
  doi = {10.1126/sciadv.adj7250}
}
Start for free →