Gamma/hadron discriminant variables in application to high-energy cosmic-ray air showers

TL;DR

This work addresses the challenge of gamma/hadron discrimination at ultra-high energies by introducing a novel trace-based discriminant, $C_{tail}$, that leverages time distributions of detector signals in extensive air showers. Built on time-resolved traces $S_i(k)$ from a 750 m-spaced water-Cherenkov array and using concentric ring analysis around the shower core, $C_{tail}$ aggregates per-station temporal information into an event-level discriminator, complementing traditional variables like $S_b$ and $X_{ ext{max}}$. Optimization across ring ranges, time bins, and zenith angles shows that the strongest discrimination occurs in rings 1–10 and time bins 1–2 for zenith angles up to $0^ ext{0}$–$40^ ext{0}$, confirming that temporal structure offers additional separation power. The approach demonstrates how shower-level temporal analysis can enhance primary identification in high-energy cosmic-ray observatories and can be integrated with existing discriminants to improve sensitivity to gamma-ray signals at EeV scales.

Abstract

Identification of primary cosmic rays on an event-by-event basis is a much-desired capability of cosmic-ray observatories. Several cosmic-ray air-shower experiments use so-called photon tags for gamma hadron primary particle discrimination. These photon tag variables are derived from the total signals measured by an array of detectors and are correlated with the total number of muons in the air shower. In this work, variables based on time distribution of signals in detectors (trace-based discriminant variables) are studied and compared to total-signal-based variables. This study relies on simulated high-energy cosmic-ray air showers with energies around 10^17.5eV. Since the variables discussed are derived from total signals and their time traces, which can be directly measured in real data, they are suitable for use as discriminant variables in the real ground-based cosmic ray experiments.

Figures (7)

• Figure 1: An exemplary array for this analysis.
• Figure 2: Signal trace $S_i(k)$ - time distribution of instantaneous signal in VEM units (Vertical Equivalent Muon) and its cumulative distribution $C_i(k)$, $k$ represents time bin numbers, $i$ is a station number.
• Figure 3: (a) Histogram of $C_{\text{tail}}$ distribution calculated for events with zenith angles 0-20 [deg], (b) final $C_{\text{tail}}$ cumulative distribution.
• Figure 4: Cumulative histograms of instantaneous signal $S_i(k)$ plotted for the same $ring \,3$, but for different time bins, and zenith angles of 0°, 30° and 60°. For showers with a 60° zenith angle, a pronounced difference between gamma-ray and proton showers is observed in the first time bin; however, this distinction diminishes in the subsequent time bins. In contrast, for vertical showers (zenith angle 0°), the difference between these two distributions, indicating discrimination capability, persists up to time bin 11.
• Figure 5: Dependence of the L1 difference between gamma and proton $C_{tail}$ distributions on ring number, zenith angle, and time bin number.