Journal article · 2022 · Open access

Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System

Patrick Laloyaux, Thorsten Kurth, Peter Dominik Dueben, David Matthew Hall

Journal of Advances in Modeling Earth Systems, 14(6) · Published 16 May 2022 · DOI 10.1029/2022ms003016

Open PDFView at the publisherCite this paper

In brief

“Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System” is a 2022 journal article by Patrick Laloyaux and 3 co-authors, published in Journal of Advances in Modeling Earth Systems. OpenAlex counts 48 citations as of 20 November 2026, which puts it in the top 4% of works published in 2022 by citations. It is open access under a CC BY-NC licence: the full text is free to read at Journal of Advances in Modeling Earth Systems.

Citations
48
Most cited in
2024 (17)
Field-weighted impact
3.57
References
58
Authors
4

Abstract

Model bias is one of the main obstacles to improved accuracy and reliability in numerical weather prediction conducted with state‐of‐the‐art atmospheric models. To deal with model bias, a modification of the standard four‐dimensional variational (4D‐Var) algorithm, called weak‐constraint 4D‐Var, has been developed where a forcing term is introduced into the model to correct for the bias that accumulates along the model trajectory. This approach reduced the temperature bias in the stratosphere by up to 50% and is implemented in the European Centre for Medium‐Range Weather Forecasts operational forecasting system. Despite different origins and applications, data assimilation (DA) and Deep Learning are both able to learn about the Earth system from observations. In this paper, a deep learning approach for model bias correction is developed using temperature retrievals from radio occultation (RO) measurements. Neural networks (NNs) require a large number of samples to properly capture the relationship between the temperature first‐guess trajectory and the model bias. As running the Integrated Forecasting System (IFS) DA system for extended periods of time with a fixed model version and at realistic resolutions is computationally very expensive, we have chosen to train, the initial NNs are trained using the ERA5 reanalysis before using transfer learning on 1 year of the current IFS model. Preliminary results show that convolutional NNs are adequate to estimate model bias from RO temperature retrievals. The different strengths and weaknesses of both deep learning and weak constraint 4D‐Var are discussed, highlighting the potential for each method to learn model biases effectively and adaptively.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Laloyaux, P., Kurth, T., Dueben, P. D., & Hall, D. M. (2022). Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System. Journal of Advances in Modeling Earth Systems, 14(6). https://doi.org/10.1029/2022ms003016

MLA 9

Laloyaux, Patrick, et al. “Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System.” Journal of Advances in Modeling Earth Systems, vol. 14, no. 6, 2022. https://doi.org/10.1029/2022ms003016.

Chicago

Laloyaux, Patrick, Thorsten Kurth, Peter Dominik Dueben, and David Matthew Hall. 2022. “Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System.” Journal of Advances in Modeling Earth Systems 14 (6). https://doi.org/10.1029/2022ms003016.

Harvard

Laloyaux, P. et al. (2022) ‘Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System’, Journal of Advances in Modeling Earth Systems, 14(6). doi:10.1029/2022ms003016.

IEEE

P. Laloyaux, T. Kurth, P. D. Dueben, and D. M. Hall, “Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System,” Journal of Advances in Modeling Earth Systems, vol. 14, no. 6, 2022, doi: 10.1029/2022ms003016.

BibTeX

@article{laloyaux2022deep,
  title     = {{Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System}},
  author    = {Laloyaux, Patrick and Kurth, Thorsten and Dueben, Peter Dominik and Hall, David Matthew},
  journal   = {Journal of Advances in Modeling Earth Systems},
  year      = {2022},
  volume    = {14},
  number    = {6},
  doi       = {10.1029/2022ms003016},
  url       = {https://doi.org/10.1029/2022ms003016}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Citations by year

YearCitationsShare of the best year
202244
202366
20241717
20251515
202666

Its field-weighted citation impact is 3.57: it has been cited 3.57 times as often as the average work of the same field, type and year (1.00 is the world average).

Topic and keywords

Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Meteorological Phenomena and Simulations

  • model bias correction
  • data assimilation
  • deep learning
  • radio occultation
  • temperature retrieval
  • convolutional neural networks
  • transfer learning
  • ERA5 reanalysis

Work with this paper

Most-cited papers in Earth and Planetary Sciences

All highly cited Earth and Planetary Sciences papers

Questions about this paper

How do I cite this paper in APA style?

Laloyaux, P., Kurth, T., Dueben, P. D., & Hall, D. M. (2022). Deep Learning to Estimate Model Biases in an Operational NWP Assimilation System. Journal of Advances in Modeling Earth Systems, 14(6). https://doi.org/10.1029/2022ms003016

How many times has this paper been cited?

48 times, according to OpenAlex on 20 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Journal of Advances in Modeling Earth Systems: sign in and open the PDF from this page.

Who wrote this paper?

Patrick Laloyaux, Thorsten Kurth, Peter Dominik Dueben, David Matthew Hall.

What is the DOI of this paper?

10.1029/2022ms003016, permanently at https://doi.org/10.1029/2022ms003016. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 20 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.