Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Searching for EeV photons with Telescope Array Surface Detector and neural networks

Telescope Array Collaboration, R. U. Abbasi, T. Abu-Zayyad, M. Allen, J. W. Belz, D. R. Bergman, F. Bradfield, I. Buckland, W. Campbell, B. G. Cheon, K. Endo, A. Fedynitch, T. Fujii, K. Fujisue, K. Fujita, M. Fukushima, G. Furlich, A. Gálvez Ureña, Z. Gerber, N. Globus, T. Hanaoka, W. Hanlon, N. Hayashida, H. He, K. Hibino, R. Higuchi, D. Ikeda, D. Ivanov, S. Jeong, C. C. H. Jui, K. Kadota, F. Kakimoto, O. Kalashev, K. Kasahara, Y. Kawachi, K. Kawata, I. Kharuk, E. Kido, H. B. Kim, J. H. Kim, J. H. Kim, S. W. Kim, R. Kobo, I. Komae, K. Komatsu, K. Komori, A. Korochkin, C. Koyama, M. Kudenko, M. Kuroiwa, Y. Kusumori, M. Kuznetsov, Y. J. Kwon, K. H. Lee, M. J. Lee, B. Lubsandorzhiev, J. P. Lundquist, H. Matsushita, A. Matsuzawa, J. A. Matthews, J. N. Matthews, K. Mizuno, M. Mori, S. Nagataki, K. Nakagawa, M. Nakahara, H. Nakamura, T. Nakamura, T. Nakayama, Y. Nakayama, K. Nakazawa, T. Nonaka, S. Ogio, H. Ohoka, N. Okazaki, M. Onishi, A. Oshima, H. Oshima, S. Ozawa, I. H. Park, K. Y. Park, M. Potts, M. Przybylak, M. S. Pshirkov, J. Remington, C. Rott, G. I. Rubtsov, D. Ryu, H. Sagawa, N. Sakaki, R. Sakamoto, T. Sako, N. Sakurai, S. Sakurai, D. Sato, K. Sekino, T. Shibata, J. Shikita, H. Shimodaira, H. S. Shin, K. Shinozaki, J. D. Smith, P. Sokolsky, B. T. Stokes, T. A. Stroman, H. Tachibana, K. Takahashi, M. Takeda, R. Takeishi, A. Taketa, M. Takita, Y. Tameda, K. Tanaka, M. Tanaka, M. Teramoto, S. B. Thomas, G. B. Thomson, P. Tinyakov, I. Tkachev, T. Tomida, S. Troitsky, Y. Tsunesada, S. Udo, F. R. Urban, M. Vrábel, D. Warren, K. Yamazaki, Y. Zhezher, Z. Zundel, J. Zvirzdin

TL;DR

The paper tackles the problem of constraining ultra-high-energy photon flux to probe cosmogenic processes and beyond-Standard-Model physics. It introduces a neural network classifier that jointly leverages composition-sensitive observables and time-resolved surface-detector waveforms, with burn-sample fine-tuning to reduce MC-data biases and a blind threshold optimization to produce the tightest possible photon-flux limits. The analysis of 14 years of Telescope Array Surface Detector data yields no photon excess above background and sets stringent 95% CL upper limits on the diffuse photon flux at high energies, improving over prior results and providing competitive constraints in the Northern Hemisphere relative to the Southern Hemisphere experiments. The approach demonstrates the viability and impact of machine-learning aided analyses on ground-based cosmic-ray detectors and informs future searches for cosmogenic photons and exotic dark-matter scenarios.

Abstract

Ultra-high-energy photons play an important role in probing astrophysical models and beyond-Standard-Model scenarios. We report updated limits on the diffuse photon flux using Telescope Array's Surface Detector data collected over 14 years of operation. Our method employs a neural network classifier to effectively distinguish between proton-induced and photon-induced events. The input data include both reconstructed composition-sensitive parameters and raw time-resolved signals registered by the Surface Detector stations. To mitigate biases from Monte Carlo simulations, we fine-tune the network with a subset of experimental data. The number of observed photon candidates is found to be consistent with the expected hadronic background, yielding upper limits on photon flux $Φ_γ(E_γ> 10^{19} \text{eV}) < 2.3 \cdot 10^{-3} $, and $Φ_γ(E_γ> 10^{20} \text{eV}) < 3.0 \cdot 10^{-4} $ $ (\text{km}^2 \cdot \text{sr} \cdot \text{yr})^{-1} $.

Searching for EeV photons with Telescope Array Surface Detector and neural networks

TL;DR

The paper tackles the problem of constraining ultra-high-energy photon flux to probe cosmogenic processes and beyond-Standard-Model physics. It introduces a neural network classifier that jointly leverages composition-sensitive observables and time-resolved surface-detector waveforms, with burn-sample fine-tuning to reduce MC-data biases and a blind threshold optimization to produce the tightest possible photon-flux limits. The analysis of 14 years of Telescope Array Surface Detector data yields no photon excess above background and sets stringent 95% CL upper limits on the diffuse photon flux at high energies, improving over prior results and providing competitive constraints in the Northern Hemisphere relative to the Southern Hemisphere experiments. The approach demonstrates the viability and impact of machine-learning aided analyses on ground-based cosmic-ray detectors and informs future searches for cosmogenic photons and exotic dark-matter scenarios.

Abstract

Ultra-high-energy photons play an important role in probing astrophysical models and beyond-Standard-Model scenarios. We report updated limits on the diffuse photon flux using Telescope Array's Surface Detector data collected over 14 years of operation. Our method employs a neural network classifier to effectively distinguish between proton-induced and photon-induced events. The input data include both reconstructed composition-sensitive parameters and raw time-resolved signals registered by the Surface Detector stations. To mitigate biases from Monte Carlo simulations, we fine-tune the network with a subset of experimental data. The number of observed photon candidates is found to be consistent with the expected hadronic background, yielding upper limits on photon flux , and .

Paper Structure

This paper contains 11 sections, 6 equations, 10 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Data sets and Monte Carlo simulations
  4. Neural network
  5. Architecture
  6. Training Process
  7. Model dependence
  8. Results
  9. Conclusions
  10. Reconstruction procedure
  11. Results without fine-tuning

Figures (10)

  • Figure 1: Schematic representation of the arrangement of the TA SD and FD stations, abu2012surface.
  • Figure 2: An example of processed and calibrated waveforms registered by a TA SD station. Blue and orange lines depict signals from the upper and lower scintillator layers, respectively.
  • Figure 3: Energy distribution of the MC sample used for neural network training with mixed photon-induced and proton-induced events after applying event selection criteria.
  • Figure 4: Architecture of the neural network used to distinguish proton-induced and photon-induced events: (1) -- spatial station bundle, (2) -- temporal station bundle, and (3) -- combined analysis. The waveform encoder architecture, which is a sub-network in the temporal station bundle, is shown inside the dotted box. Solid arrows depict data flow, dotted lines connect elements with their detailed description. Layer names and parameters follow TensorFlow API.
  • Figure 5: Architecture of the temporal station bundle. See the temporal station bundle in Figure \ref{['nn_arch']} for layers characteristics.
  • ...and 5 more figures