EUVIMEDEUVIMED entdecken
Crossref

Deep Learning with Magnetic Parameter Constraints for Short-Term Prediction of Solar Active Region Vector Magnetic Fields

Yuqing Zhou, Hui Liu, Zhenyu Jin, Yuyang Li, Sizhong Zou, Jiaben Lin, Mingfu Shao, Zhuoheng Huang

Solar Physics · 2026-06 · 301 (6)

DOI: 10.1007/s11207-026-02687-1 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Solar Physics, online erschienen am 10.06.2026. Verfasst von Yuqing Zhou und 7 weiteren Autor:innen. Die Arbeit nennt 24 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Forecasting the dynamic evolution of solar magnetic fields is a critical technique for enabling space-weather warnings. Addressing the limitations of existing methods in predicting all vector magnetic field components and in maintaining consistency with solar surface magnetic-field-related quantities, this study proposes a deep learning prediction method that integrates dynamic masks of active regions with multiple magnetic parameter constraints. By constructing a three-channel representation of vector magnetic fields, applying dynamic masks to enhance attention to strong-field regions, and incorporating multi-parameter magnetic parameter constraints, we developed an end-to-end short-term (12-hour) predictive model of solar vector magnetic field evolution. Using SDO/SHARP vector magnetogram data, the model predicts and analyses field evolution across all components. Quantitative evaluations demonstrate that our approach achieves horizon-averaged structural similarity index measure (SSIM) of 0.912 (per-hour range: 0.909 – 0.916) and correlation coefficient (CC) of 0.998 for the radial component Br\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$B_{r}$\end{document} (root-mean-square error (RMSE) 13.0 – 21.0 G); the horizontal components achieve Bϕ\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$B_{\phi }$\end{document} SSIM 0.760 – 0.800 (CC 0.910 – 0.945, RMSE 38.5 – 50.0 G) and Bθ\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$B_{\theta }$\end{document} SSIM 0.728 – 0.750 (CC 0.895 – 0.920, RMSE 38.5 – 49.0 G). The model maintains unsigned magnetic-flux prediction errors at 7.82% (95% confidence interval (CI): ±0.11%). These results demonstrate strong image-domain performance together with consistency under the magnetic-parameter diagnostics used here, suggesting initial potential for supporting future space weather forecasting efforts.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Zhou, Y., Liu, H., Jin, Z., Li, Y., Zou, S., Lin, J., Shao, M., & Huang, Z. (2026). Deep Learning with Magnetic Parameter Constraints for Short-Term Prediction of Solar Active Region Vector Magnetic Fields. Solar Physics, 301(6). https://doi.org/10.1007/s11207-026-02687-1

Vancouver

Zhou Y, Liu H, Jin Z, Li Y, Zou S, Lin J, et al. Deep Learning with Magnetic Parameter Constraints for Short-Term Prediction of Solar Active Region Vector Magnetic Fields. Solar Physics. 2026;301(6). doi:10.1007/s11207-026-02687-1

BibTeX
@article{Zhou2026,
  author = {Zhou, Yuqing and Liu, Hui and Jin, Zhenyu and Li, Yuyang and Zou, Sizhong and Lin, Jiaben and Shao, Mingfu and Huang, Zhuoheng},
  title = {Deep Learning with Magnetic Parameter Constraints for Short-Term Prediction of Solar Active Region Vector Magnetic Fields},
  journal = {Solar Physics},
  year = {2026},
  volume = {301},
  number = {6},
  doi = {10.1007/s11207-026-02687-1}
}
Kostenlos starten →