Forecasting Local Geomagnetic Disturbances Using Machine Learning: A Multi‐Station Approach
Space Weather · 2026-07 · 24 (7)
DOI: 10.1029/2025sw004654 ↗Kurz gefasst
Fachartikel in der Zeitschrift Space Weather, online erschienen am 22.07.2026. Verfasst von Armando Collado‐Villaverde und 3 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Ministerio de Ciencia, Innovación y Universidades.
Abstract (Originalsprache)
Geomagnetic indices such as SYM‐H capture global disturbances but fail to resolve regional impacts critical for infrastructure resilience. We present a machine learning model to forecast the Local Disturbance Index (LDi), a validated regional geomagnetic index. Using solar wind data from the ACE satellite and station‐specific metadata (Magnetic Local Time and Magnetic coordinates) our model predicts the LDi 1 and 2 hr ahead for six mid‐latitude INTERMAGNET observatories, while also providing confidence intervals to quantify uncertainty. We tested the model on 21 recent storms (2017–2024) using ACE's preliminary data. Results show that local indices are essential for accurate regional impact assessment, improving risk mitigation for ground infrastructures such as pipelines or power grids. The model also is also capable to generalize its predictions to observatories not included in the training process.
Förderung
- Ministerio de Ciencia, Innovación y Universidades
Zitieren
APA
Collado‐Villaverde, A., Muñoz, P., Cid, C., & Maseda‐Zurdo, I. (2026). Forecasting Local Geomagnetic Disturbances Using Machine Learning: A Multi‐Station Approach. Space Weather, 24(7). https://doi.org/10.1029/2025sw004654
Vancouver
Collado‐Villaverde A, Muñoz P, Cid C, Maseda‐Zurdo I. Forecasting Local Geomagnetic Disturbances Using Machine Learning: A Multi‐Station Approach. Space Weather. 2026;24(7). doi:10.1029/2025sw004654
BibTeX
@article{ColladoVillaverde2026,
author = {Collado‐Villaverde, Armando and Muñoz, Pablo and Cid, Consuelo and Maseda‐Zurdo, Iván},
title = {Forecasting Local Geomagnetic Disturbances Using Machine Learning: A Multi‐Station Approach},
journal = {Space Weather},
year = {2026},
volume = {24},
number = {7},
doi = {10.1029/2025sw004654}
}
EUVIMED