CoChat
Sign Up
Cited by 1,044Meteorological Phenomena and SimulationsClimate variability and modelsClimate Change and Geoengineering

Deep learning to represent subgrid processes in climate models

Open access
Rasp, Stephan, Pritchard, Michael S., Gentine, Pierre·2018·Proceedings of the National Academy of Sciences
DOI 10.1073/pnas.1810286115PMC PMC6166853
Open this paper in CoChat — free account

Abstract

The representation of nonlinear subgrid processes, especially clouds, has been a major source of uncertainty in climate models for decades. Cloud-resolving models better represent many of these processes and can now be run globally but only for short-term simulations of at most a few years because of computational limitations. Here we demonstrate that deep learning can be used to capture many advantages of cloud-resolving modeling at a fraction of the computational cost. We train a deep neural network to represent all atmospheric subgrid processes in a climate model by learning from a multiscale model in which convection is treated explicitly. The trained neural network then replaces the traditional subgrid parameterizations in a global general circulation model in which it freely interacts with the resolved dynamics and the surface-flux scheme. The prognostic multiyear simulations are stable and closely reproduce not only the mean climate of the cloud-resolving simulation but also key aspects of variability, including precipitation extremes and the equatorial wave spectrum. Furthermore, the neural network approximately conserves energy despite not being explicitly instructed to. Finally, we show that the neural network parameterization generalizes to new surface forcing patterns but struggles to cope with temperatures far outside its training manifold. Our results show the feasibility of using deep learning for climate model parameterization. In a broader context, we anticipate that data-driven Earth system model development could play a key role in reducing climate prediction uncertainty in the coming decade.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1073/pnas.1810286115

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Methods based on various forms of artificial intelligence are already today being applied in a number of research fields related to climate change and environmental monitoring.

    “Methods based on various forms of artificial intelligence are already today being applied in a number of research fields related to climate change and environmental monitoring.”

    From Artificial intelligence, systemic risks, and sustainability (2021) · cites this paper as [25]

    The Discussion demonstrates an existing deep-learning application in climate modeling and notes that random forests have also been applied to model subgrid convection. It does not document applications in environmental monitoring or establish that AI is already being applied across a number of climate-related research fields; it describes uses beyond the atmosphere as possibilities. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • So far, hybrid models have mostly been limited to idealized scenarios such as aquaplanets.

    “So far, hybrid models have mostly been limited to idealized scenarios such as aquaplanets22,23.”

    From Neural general circulation models for weather and climate (2024) · cites this paper as [22]

  • Gentine et al. trained a deep neural network to emulate the behavior of a high-resolution cloud simulation, and found that the network gave similar results for a fraction of the cost and was stable in a simplified global model.

    “Gentine et al. trained a deep neural network to emulate the behavior of a high-resolution cloud simulation, and found that the network gave similar results for a fraction of the cost [275] and was stable in a simplified global model [670].”

    From Tackling climate change with machine learning (2023) · cites this paper as [670]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Rasp, S., Pritchard, M. S., & Gentine, P.
Affiliations
  • University of California, Irvine (US)
  • Columbia University (US)
Corresponding author
Rasp, Stephan
ORCID
Venue
Proceedings of the National Academy of Sciences
Publisher
National Academy of Sciences
Published
2018-09-06
Volume
115
Issue
39
Pages
9684-9689
ISSN
  • 0027-8424 (ISSN-L)
  • 1091-6490
DOI
10.1073/pnas.1810286115
PMC
PMC6166853
OpenAlex
W2808559252
Cited by
1,044
References
55
Citations by year
2026: 127 · 2025: 104 · 2024: 167 · 2023: 211 · 2022: 130 · 2021: 174 · 2020: 86 · 2019: 38 · 2018: 7
FWCI
37.15
Type
journal article
Language
en
Open access
Licence
Locations
  • Proceedings of the National Academy of Sciences (publishedVersion, cc-by-nc-nd)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
  • OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
  • arXiv (Cornell University) (cc-by)
  • Columbia Academic Commons (Columbia University)
  • AGU Fall Meeting Abstracts
Funders
  • Deutsche Forschungsgemeinschaft (SFB/TRR 165)
  • DOE | Office of Science (DE-SC0012152, DE-SC00-12548, DE-SC0014203)
  • NSF | GEO | Division of Atmospheric and Geospace Sciences (AGS-1419518, AGS-1734164, AGS-1734156, AGS-1649770)
  • National Aeronautics and Space Administration (NNX14AI36G)
Topics
  • Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Meteorological Phenomena and Simulations
  • Physical Sciences › Environmental Science › Global and Planetary Change › Climate variability and models
  • Physical Sciences › Environmental Science › Global and Planetary Change › Climate Change and Geoengineering
Keywords
deep learning, subgrid-scale modeling, cloud parameterization, cloud-resolving model, global climate models, multiscale modeling, convection, heavy rainfall, equatorial waves, energy conservation (physics), climate prediction uncertainty

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review