Spatiotemporal weather forecasting via multi-scale graph neural networks and latent diffusion models
PLOS One · 04.06.2026 · 21 (6) · e0348354
DOI: 10.1371/journal.pone.0348354 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 04.06.2026. Verfasst von ZhiPeng Wu. Beteiligte Einrichtung: School of Finance, Hefei University of Economics, Hefei, Anhui, China. Hauptthemen laut MEDLINE: Weather. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Anhui Provincial Quality Engineering Projects, Action Project for Training Young and Middle-aged Teachers of Education Department of Anhui Province, Youth and Middle-aged Scientific and Technological Innovation Leading Talents Program of the Corps.
Abstract (Originalsprache)
Accurate weather prediction is crucial in agriculture, disaster prevention, and public safety. Challenge: Traditional numerical models have high computational costs and struggle with atmospheric nonlinearity and chaos, while existing deep learning methods face limitations in handling spatial heterogeneity and non-Euclidean data. Solution: This paper introduces the STGLDWeather method. It combines multi-scale spatiotemporal graph neural networks (MS-ST-GNN) and latent diffusion models (LDM) to capture multi-scale spatiotemporal dependencies in weather data and model the temporal evolution of weather conditions in latent space. Conclusion: Experiments on real weather datasets show that STGLDWeather significantly outperforms existing state-of-the-art baselines in prediction accuracy and computational efficiency, particularly excelling in temperature, geopotential height, and wind speed forecasts.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- School of Finance, Hefei University of Economics, Hefei, Anhui, China.
ORCID-Profile: ZhiPeng Wu
Förderung
- Anhui Provincial Quality Engineering Projects
- Action Project for Training Young and Middle-aged Teachers of Education Department of Anhui Province
- Youth and Middle-aged Scientific and Technological Innovation Leading Talents Program of the Corps
Zitieren
APA
Wu, Z. (2026). Spatiotemporal weather forecasting via multi-scale graph neural networks and latent diffusion models. PLOS One, 21(6), e0348354. https://doi.org/10.1371/journal.pone.0348354
Vancouver
Wu Z. Spatiotemporal weather forecasting via multi-scale graph neural networks and latent diffusion models. PLOS One. 2026;21(6):e0348354. doi:10.1371/journal.pone.0348354
BibTeX
@article{Wu2026,
author = {Wu, ZhiPeng},
title = {Spatiotemporal weather forecasting via multi-scale graph neural networks and latent diffusion models},
journal = {PLOS One},
year = {2026},
volume = {21},
number = {6},
pages = {e0348354},
doi = {10.1371/journal.pone.0348354}
}
EUVIMED