Preprint · 2024 · Open access

Data driven weather forecasts trained and initialised directly from observations

A. P. McNally, Christian Lessig, Peter Lean, Eulalie Boucher, Mihai Alexe, Ewan Pinnington, Matthew Chantry, Lang, Simon, Burrows, Chris, Marcin Chrust, Florian Pinault, Ethel Villeneuve and 2 more

arXiv (Cornell University) · Published 22 July 2024 · DOI 10.48550/arxiv.2407.15586

Free PDFView at the publisherCite this paper

In brief

“Data driven weather forecasts trained and initialised directly from observations” is a 2024 preprint by A. P. McNally and 13 co-authors, published in arXiv (Cornell University). OpenAlex counts 51 citations as of 15 November 2026. It is open access under a CC BY licence: the full text is free to read at arXiv (Cornell University).

Citations
51
Most cited in
2026 (23)
References
7
Authors
14

Abstract

Skilful Machine Learned weather forecasts have challenged our approach to numerical weather prediction, demonstrating competitive performance compared to traditional physics-based approaches. Data-driven systems have been trained to forecast future weather by learning from long historical records of past weather such as the ECMWF ERA5. These datasets have been made freely available to the wider research community, including the commercial sector, which has been a major factor in the rapid rise of ML forecast systems and the levels of accuracy they have achieved. However, historical reanalyses used for training and real-time analyses used for initial conditions are produced by data assimilation, an optimal blending of observations with a physics-based forecast model. As such, many ML forecast systems have an implicit and unquantified dependence on the physics-based models they seek to challenge. Here we propose a new approach, training a neural network to predict future weather purely from historical observations with no dependence on reanalyses. We use raw observations to initialise a model of the atmosphere (in observation space) learned directly from the observations themselves. Forecasts of crucial weather parameters (such as surface temperature and wind) are obtained by predicting weather parameter observations (e.g. SYNOP surface data) at future times and arbitrary locations. We present preliminary results on forecasting observations 12-hours into the future. These already demonstrate successful learning of time evolutions of the physical processes captured in real observations. We argue that this new approach, by staying purely in observation space, avoids many of the challenges of traditional data assimilation, can exploit a wider range of observations and is readily expanded to simultaneous forecasting of the full Earth system (atmosphere, land, ocean and composition).

Cite this paper

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

APA 7

McNally, A. P., Lessig, C., Lean, P., Boucher, E., Alexe, M., Pinnington, E., Chantry, M., Simon, L., Chris, B., Chrust, M., Pinault, F., Villeneuve, E., Bormann, N., & Healy, S. B. (2024). Data driven weather forecasts trained and initialised directly from observations. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2407.15586

MLA 9

McNally, A. P., et al. “Data driven weather forecasts trained and initialised directly from observations.” arXiv (Cornell University), 2024. https://doi.org/10.48550/arxiv.2407.15586.

Chicago

McNally, A. P., Christian Lessig, Peter Lean, Eulalie Boucher, Mihai Alexe, Ewan Pinnington, Matthew Chantry, et al. 2024. “Data driven weather forecasts trained and initialised directly from observations.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2407.15586.

Harvard

McNally, A. P. et al. (2024) ‘Data driven weather forecasts trained and initialised directly from observations’, arXiv (Cornell University). doi:10.48550/arxiv.2407.15586.

IEEE

A. P. McNally et al., “Data driven weather forecasts trained and initialised directly from observations,” arXiv (Cornell University), 2024, doi: 10.48550/arxiv.2407.15586.

BibTeX

@misc{mcnally2024data,
  title     = {{Data driven weather forecasts trained and initialised directly from observations}},
  author    = {McNally, A. P. and Lessig, Christian and Lean, Peter and Boucher, Eulalie and Alexe, Mihai and Pinnington, Ewan and Chantry, Matthew and Simon, Lang, and Chris, Burrows, and Chrust, Marcin and Pinault, Florian and Villeneuve, Ethel and Bormann, Niels and Healy, Sean B.},
  howpublished = {arXiv (Cornell University)},
  year      = {2024},
  doi       = {10.48550/arxiv.2407.15586},
  url       = {https://doi.org/10.48550/arxiv.2407.15586}
}

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

Citations by year

YearCitationsShare of the best year
201811
202311
202477
20251919
20262323

Topic and keywords

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

  • data-driven prediction
  • observation space
  • data assimilation
  • numerical weather prediction
  • wind forecasting

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?

McNally, A. P., Lessig, C., Lean, P., Boucher, E., Alexe, M., Pinnington, E., Chantry, M., Simon, L., Chris, B., Chrust, M., Pinault, F., Villeneuve, E., Bormann, N., & Healy, S. B. (2024). Data driven weather forecasts trained and initialised directly from observations. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2407.15586

How many times has this paper been cited?

51 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 arXiv (Cornell University).

Who wrote this paper?

A. P. McNally, Christian Lessig, Peter Lean, Eulalie Boucher, Mihai Alexe, Ewan Pinnington, Matthew Chantry, Lang, Simon, Burrows, Chris, Marcin Chrust, Florian Pinault, Ethel Villeneuve and 2 more.

What is the DOI of this paper?

10.48550/arxiv.2407.15586. The permanent link is https://doi.org/10.48550/arxiv.2407.15586.

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.