- Physical Review Journals
All JournalsPhysics Magazine

Physical Review D

Reuse & Permissions

It is not necessary to obtain permission to reuse this article or its components as it is available under the terms of the Creative Commons Attribution 4.0 International license. This license permits unrestricted use, distribution, and reproduction in any medium, provided attribution to the author(s) and the published article's title, journal citation, and DOI are maintained. Please note that some figures may have been included with permission from other third parties. It is your responsibility to obtain the proper permission from the rights holder directly for these figures.

Export citation

Export citation

Choose format for download:

Download Citation
  • Open Access

Hybrid-graph neural network method for muon fast reconstruction in neutrino telescopes

Cen Mo and Liang Li*

  • State Key Laboratory of Dark Matter Physics, Key Laboratory for Particle Astrophysics and Cosmology (MOE), Shanghai Key Laboratory for Particle Physics and Cosmology (SKLPPC), School of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China

  • *Contact author: liangliphy@sjtu.edu.cn
PDF

Phys. Rev. D 112, 072012 – Published 24 October, 2025

DOI: https://doi.org/10.1103/z6dg-58fd

Abstract

Fast and accurate muon reconstruction is crucial for neutrino telescopes to improve experimental sensitivity and enable online triggering. This paper introduces a hybrid-graph neural network (GNN) method tailored for efficient muon track reconstruction, leveraging the robustness of GNNs, alongside traditional physics-based approaches. The “light GNN model” achieves a run-time of 0.19–0.29 ms per event on GPUs, offering a 3 orders of magnitude speedup compared to traditional likelihood-based methods, while maintaining a high reconstruction accuracy. For high-energy muons (10–100 TeV), the median angular error is approximately 0.1°, with errors in reconstructed Cherenkov photon emission positions being below 3–5 m, depending on the GNN model used. Furthermore, the semi-GNN method offers a mechanism to assess the quality of event reconstruction, enabling the identification and exclusion of poorly reconstructed events. These results establish the GNN-based approach as a promising solution for next-generation neutrino telescope data reconstruction.

View figure in article

Physics Subject Headings (PhySH)

Article Text

References (30)

  1. IceCube Collaboration, Evidence for high-energy extraterrestrial neutrinos at the IceCube detector, Science 342, 1242856 (2013).
  2. Z. P. Ye, F. Hu, W. Tian et al., A multi-cubic-kilometre neutrino telescope in the western pacific ocean, Nat. Astron. 7, 1497 (2023).
  3. KM3NeT Collaboration, Letter of intent for KM3NeT 2.0, J. Phys. G 43, 084001 (2016).
  4. A. Avrorin et al., The gigaton volume detector in lake baikal, Nucl. Instrum. Methods Phys. Res., Sect. A 639, 30 (2011).
  5. P. H. Barrett, L. M. Bollinger, G. Cocconi, Y. Eisenberg, and K. Greisen, Interpretation of cosmic-ray measurements far underground, Rev. Mod. Phys. 24, 133 (1952).
  6. J.-H. Koehne, K. Frantzen, M. Schmitz, T. Fuchs, W. Rhode, D. Chirkin, and J. B. Tjus, Proposal: A tool for propagation of charged leptons, Comput. Phys. Commun. 184, 2070 (2013).
  7. IceCube Collaboration, Neutrino emission from the direction of the blazar TXS 0506+056 prior to the IceCube-170922A alert, Science 361, 147 (2018).
  8. IceCube Collaboration, Evidence for neutrino emission from the nearby active galaxy NGC 1068, Science 378, 538 (2022).
  9. S. Aiello et al. (KM3NeT Collaboration), Observation of an ultra-high-energy cosmic neutrino with KM3NeT, Nature (London) 638, 376 (2025).
  10. F. J. Yu, J. Lazar, and C. A. Argüelles, Trigger-level event reconstruction for neutrino telescopes using sparse submanifold convolutional neural networks, Phys. Rev. D 108, 063017 (2023).
  11. M. Jin, Y. Hu, and C. A. Argüelles, Two watts is all you need: Enabling in-detector real-time machine learning for neutrino telescopes via edge computing, J. Cosmol. Astropart. Phys. 06 (2024) 026.
  12. S. A. et al. (GEANT4 Collaboration), geant4—A simulation toolkit, Nucl. Instrum. Methods Phys. Res., Sect. A 506, 250 (2003).
  13. A. et al. (GEANT4 Collaboration), geant4 developments and applications, IEEE Trans. Nucl. Sci. 53, 270 (2006).
  14. J. A. et al., Recent developments in geant4, Nucl. Instrum. Methods Phys. Res., Sect. A 835, 186 (2016).
  15. S. G. Parker et al., Optix: A general purpose ray tracing engine, Acm Trans. Graphic. 29, 1 (2010).
  16. S. Blyth, Opticks: GPU optical photon simulation for particle physics using NVIDIA® OptiXTM, EPJ Web Conf. 214, 02027 (2019).
  17. F. Hu, Z. Li, and D. Xu, Exploring a PMT+SiPM hybrid optical module for next generation neutrino telescopes, Proc. Sci., ICRC2021 (2021) 1043.
  18. R. Abbasi et al., A convolutional neural network based cascade reconstruction for the IceCube neutrino observatory, J. Instrum. 16, P07041 (2021).
  19. S. Aiello et al., Event reconstruction for KM3net/ORCA using convolutional neural networks, J. Instrum. 15, P10005 (2020).
  20. Y. Wang, Y. Sun, Z. Liu, S. E. Sarma, M. M. Bronstein, and J. M. Solomon, Dynamic graph CNN for learning on point clouds, arXiv:1801.07829.
  21. H. Qu and L. Gouskos, Jet tagging via particle clouds, Phys. Rev. D 101, 056019 (2020).
  22. S. Ioffe and C. Szegedy, Batch normalization: Accelerating deep network training by reducing internal covariate shift, arXiv:1502.03167.
  23. X. Glorot, A. Bordes, and Y. Bengio, Deep sparse rectifier neural networks, in Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, Proceedings of Machine Learning Research, Vol. 15, edited by G. Gordon, D. Dunson, and M. Dudík (PMLR, Fort Lauderdale, FL, 2011), pp. 315–323.
  24. K. He, X. Zhang, S. Ren, and J. Sun, Deep residual learning for image recognition, arXiv:1512.03385.
  25. A. Paszke et al., Pytorch: An imperative style, high-performance deep learning library, in Advances in Neural Information Processing Systems 32 (Curran Associates, Inc., Red Hook, NY, USA, 2019), pp. 8024–8035.
  26. M. Fey and J. E. Lenssen, Fast graph representation learning with PyTorch Geometric, in ICLR Workshop on Representation Learning on Graphs and Manifolds (2019).
  27. D. P. Kingma and J. Ba, Adam: A method for stochastic optimization, arXiv:1412.6980.
  28. M. J. D. Powell, An efficient method for finding the minimum of a function of several variables without calculating derivatives, Computer Journal (UK) 7, 155 (1964).
  29. P. Virtanen et al. (SciPy 1.0 Contributors), scipy 1.0: Fundamental algorithms for scientific computing in python, Nat. Methods 17, 261 (2020).
  30. C. Mo, Hybrid-gnn v1.0, https://github.com/cmo-ft/Hybrid-GNN/ (2025).

Outline

Information

Phys. Rev. D 112, 072012– Published 24 October, 2025
  • Received 6 June 2025
  • Accepted 10 September 2025
Crossmark - Check for updates button
Creative Commons Logo

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI. Funded by SCOAP3.

Published by the American Physical Society

Sign In to Your Journals Account

Filter

Filter

Article Lookup

Enter a citation