Conference paper · 2024 · Open access

Towards a machine learning model for data assimilation and forecasting directly trained from observations

Christian Lessig, Peter Lean, Tony McNally, Mihai Alexe, Simon T. K. Lang, Matthew Chantry, Peter Dominik Dueben

EGU General Assembly Conference Abstracts · Published 8 March 2024 · DOI 10.5194/egusphere-egu24-11724

Read the free full textCite this paper

In brief

“Towards a machine learning model for data assimilation and forecasting directly trained from observations” is a 2024 conference paper by Christian Lessig and 6 co-authors, published in EGU General Assembly Conference Abstracts. OpenAlex counts 0 citations as of 15 November 2026. It is open access under a CC BY licence: the full text is free to read at EGU General Assembly Conference Abstracts.

Citations
0
Authors
7

Abstract

State-of-the-art data assimilation systems, such as the 4DVar system of the European Centre for Medium-Range Weather Forecasts (ECMWF), are highly successful in producing state estimates of the atmosphere constrained by millions of observations. However, existing systems require substantial approximations, e.g. in forward operators, and employ conventional models in their optimization loop. This limits the amount of information that can be extracted from observations, e.g. the assimilation of visible satellite channels is still challenging. The current approach also separates the use of observations into a data assimilation step and a forecasting one, with observations only being used indirectly for forecasting, for example for tuning of parametrizations and for evaluation.Here, we explore the possibility to train large machine learning models for data assimilation and forecasting directly from observations. In particular, we build a generative transformer neural network that models the joint the probability distribution p(y,x) over output states y for an input x. The input are observations from a temporal window, e.g. 6h or 12h, and the output y can either be an estimate of the state within the window or a short-term forecast, e.g. for another 12h. Different observations are processed by different embedding networks but then fused in the backbone transformer network. To obtain an integrated and consistent representation of the atmospheric state that corresponds to the different input data streams, we train with a variation of the masked token model training objective from natural language processing that impels the network to learn the correlation between the different input streams and channels. To properly represent the statistical nature of the estimation of y given x, our network provides an ensemble prediction as a nonparametric model for the probability distribution over y.We present results for a network trained with a substantial amount of data, including different satellite observations (such as AMSU-A microwave sounders from NOAA 15-19 and the METOP satellites as well as IASI), radiosondes, and ground station-based measurements. The skill for both data assimilation and forecasting is analyzed and compared to ECMWF’s operational 4DVar system. We also ablate the effect different observations have on the skill of the network output.

Cite this paper

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

APA 7

Lessig, C., Lean, P., McNally, T., Alexe, M., Lang, S. T. K., Chantry, M., & Dueben, P. D. (2024). Towards a machine learning model for data assimilation and forecasting directly trained from observations. EGU General Assembly Conference Abstracts. https://doi.org/10.5194/egusphere-egu24-11724

MLA 9

Lessig, Christian, et al. “Towards a machine learning model for data assimilation and forecasting directly trained from observations.” EGU General Assembly Conference Abstracts, 2024. https://doi.org/10.5194/egusphere-egu24-11724.

Chicago

Lessig, Christian, Peter Lean, Tony McNally, Mihai Alexe, Simon T. K. Lang, Matthew Chantry, and Peter Dominik Dueben. 2024. “Towards a machine learning model for data assimilation and forecasting directly trained from observations.” EGU General Assembly Conference Abstracts. https://doi.org/10.5194/egusphere-egu24-11724.

Harvard

Lessig, C. et al. (2024) ‘Towards a machine learning model for data assimilation and forecasting directly trained from observations’, EGU General Assembly Conference Abstracts. doi:10.5194/egusphere-egu24-11724.

IEEE

C. Lessig et al., “Towards a machine learning model for data assimilation and forecasting directly trained from observations,” EGU General Assembly Conference Abstracts, 2024, doi: 10.5194/egusphere-egu24-11724.

BibTeX

@inproceedings{lessig2024towards,
  title     = {{Towards a machine learning model for data assimilation and forecasting directly trained from observations}},
  author    = {Lessig, Christian and Lean, Peter and McNally, Tony and Alexe, Mihai and Lang, Simon T. K. and Chantry, Matthew and Dueben, Peter Dominik},
  booktitle = {EGU General Assembly Conference Abstracts},
  year      = {2024},
  doi       = {10.5194/egusphere-egu24-11724},
  url       = {https://doi.org/10.5194/egusphere-egu24-11724}
}

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

Topic and keywords

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

  • data assimilation
  • weather forecasting
  • generative transformers
  • joint probability distribution
  • masked token prediction
  • ensemble forecasting
  • satellite observations
  • IASI

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?

Lessig, C., Lean, P., McNally, T., Alexe, M., Lang, S. T. K., Chantry, M., & Dueben, P. D. (2024). Towards a machine learning model for data assimilation and forecasting directly trained from observations. EGU General Assembly Conference Abstracts. https://doi.org/10.5194/egusphere-egu24-11724

How many times has this paper been cited?

0 times, according to OpenAlex on 15 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: read the free full text at EGU General Assembly Conference Abstracts.

Who wrote this paper?

Christian Lessig, Peter Lean, Tony McNally, Mihai Alexe, Simon T. K. Lang, Matthew Chantry, Peter Dominik Dueben.

What is the DOI of this paper?

10.5194/egusphere-egu24-11724. The permanent link is https://doi.org/10.5194/egusphere-egu24-11724.

Bibliographic data from OpenAlex (CC0), last checked 15 November 2026. ResearchGuru does not host this paper; the links go to the publisher or the repository that does.