# Study of cosmic ray events with high muon multiplicity using the ALICE detector at the CERN Large Hadron Collider

ALICE is one of four large experiments at the CERN Large Hadron Collider near Geneva, specially designed to study particle production in ultra-relativistic heavy-ion collisions. Located 52 meters underground with 28 meters of overburden rock, it has also been used to detect muons produced by cosmic ray interactions in the upper atmosphere. In this paper, we present the multiplicity distribution of these atmospheric muons and its comparison with Monte Carlo simulations. This analysis exploits the large size and excellent tracking capability of the ALICE Time Projection Chamber. A special emphasis is given to the study of high multiplicity events containing more than 100 reconstructed muons and corresponding to a muon areal density $\rho_{\mu} > 5.9~$m$^{-2}$. Similar events have been studied in previous underground experiments such as ALEPH and DELPHI at LEP. While these experiments were able to reproduce the measured muon multiplicity distribution with Monte Carlo simulations at low and intermediate multiplicities, their simulations failed to describe the frequency of the highest multiplicity events. In this work we show that the high multiplicity events observed in ALICE stem from primary cosmic rays with energies above $10^{16}$ eV and that the frequency of these events can be successfully described by assuming a heavy mass composition of primary cosmic rays in this energy range. The development of the resulting air showers was simulated using the latest version of QGSJET to model hadronic interactions. This observation places significant constraints on alternative, more exotic, production mechanisms for these events.

Figures

## Figure 1

 A single atmospheric muon event. The thin outer cylinder is the Time Of Flight detector (1) The large inner cylinder is the Time Projection Chamber (2) and the smaller cylinder at the centre isthe silicon Inner Tracking System (3). Muons are reconstructed as two TPC tracks, one in the upper half oft he detector ($up$ track) and the other in the lower half ($down$ track), which are then joined to create a single muon track.

## Figure 2

 Root-mean-square and mean values of the relative difference between the number of generated and reconstructed muons for events simulated with different muon multiplicities.

## Figure 3

 The ratio of muons reconstructed as single tracks to the total number of reconstructed muons (both single and matched tracks) in the data and simulations with proton and iron primaries.

## Figure 4

 TOF trigger efficiency as a function of muon multiplicity.

## Figure 5

 Muon multiplicity distribution of the whole sample of data (2010-2013) corresponding to 30.8 days of data taking.

## Figure 6

 The measured muon multiplicity distribution compared with the values and fits obtained from CORSIKA simulations with proton andiron primary cosmic rays for 30.8 days of data taking. The errors are shown separately (statistical and systematic) for data, while for Monte Carlo they are the quadrature sum of the statistical and systematic uncertainties.

## Figure 7

 Event display of a multi-muon event with 276 reconstructed muons crossing the TPC.

## Figure 8

 Zenithal and azimuthal distribution of the multi-muon eventwith 276 reconstructed muons.

## Figure 9

 Spatial distribution of the 276 recostructed muons indicatingmatched and single-track muons.

## Figure 10

 The surface level spatial distribution of the cores of simulated EAS giving rise to more than 100 muons in the ALICE Time Projection Chamber. The simulation was for iron primaries in the energy range $10^{16}-10^{18}$ eV and corresponds to the equivalent of 5 years of data taking