Academic Journal
ArchesWeatherGen: Skillful and compute-efficient probabilistic weather forecasting with machine learning.
| Τίτλος: | ArchesWeatherGen: Skillful and compute-efficient probabilistic weather forecasting with machine learning. |
|---|---|
| Συγγραφείς: | Couairon G; INRIA, Paris, France., Singh R; INRIA, Paris, France., Charantonis A; INRIA, Paris, France., Lessig C; ECMWF, Bonn, Germany., Monteleoni C; INRIA, Paris, France.; University of Colorado Boulder, Boulder, CO, USA. |
| Πηγή: | Science advances [Sci Adv] 2026 Apr 24; Vol. 12 (17), pp. eadx2372. Date of Electronic Publication: 2026 Apr 22. |
| Τύπος έκδοσης: | Journal Article |
| Γλώσσα: | English |
| Στοιχεία περιοδικού: | Publisher: American Association for the Advancement of Science Country of Publication: United States NLM ID: 101653440 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 2375-2548 (Electronic) Linking ISSN: 23752548 NLM ISO Abbreviation: Sci Adv Subsets: MEDLINE; PubMed not MEDLINE |
| Imprint Name(s): | Original Publication: Washington, DC : American Association for the Advancement of Science, [2015]- |
| Περίληψη: | Weather forecasting plays a vital role in today's society, from agriculture and logistics to predicting the output of renewable energies and preparing for extreme weather events. Deep learning weather forecasting models trained with the next state prediction objective on ERA5 have shown great success compared to numerical global circulation models. Here, we propose a methodology to leverage deterministic weather models in the design of probabilistic weather models, leading to improved performance and reduced computing costs. We design a probabilistic weather model based on flow matching, a modern variant of diffusion models, that is trained to project deterministic weather predictions to the distribution of ERA5 weather states. Our model ArchesWeatherGen surpasses IFS ENS and NeuralGCM on all WeatherBench headline variables (except for NeuralGCM's geopotential). Our work also aims to democratize the use of generative machine learning models in weather forecasting research. |
| References: | Nature. 2024 Aug;632(8027):1060-1066. (PMID: 39039241) Sci Adv. 2024 Mar 29;10(13):eadk4489. (PMID: 38552014) Proc Natl Acad Sci U S A. 2025 Apr 29;122(17):e2420288122. (PMID: 40279391) Science. 2023 Dec 22;382(6677):1416-1421. (PMID: 37962497) Nature. 2015 Sep 3;525(7567):47-55. (PMID: 26333465) Nature. 2023 Jul;619(7970):533-538. (PMID: 37407823) Nature. 2025 Jan;637(8044):84-90. (PMID: 39633054) |
| Entry Date(s): | Date Created: 20260422 Latest Revision: 20260813 |
| Update Code: | 20260912 |
| PubMed Central ID: | PMC13101865 |
| DOI: | 10.1126/sciadv.adx2372 |
| PMID: | 42018632 |
| Βάση Δεδομένων: | MEDLINE |
| ISSN: | 2375-2548 |
|---|---|
| DOI: | 10.1126/sciadv.adx2372 |