Trajectory determination of muons using scintillators and a novel self-organizative map

https://doi.org/10.1016/j.nima.2020.164166Get rights and content

Abstract

In this work we propose a method for the determination of the impact point of muons in scintillators using a novel type of self-organizative maps called Self-Equalizing Map (SEM) and comparing the relative pulse height obtained by four photomultipliers (PMTs) at each scintillator. Using two 1 m2 scintillators and calculating the impact point in both of them, we can also estimate the angle of incidence of these particles. This method has been specifically designed for a muon telescope called MITO (Muon Impact Tracer and Observer) which is part of the ORCA (Antarctic Cosmic Ray Observatory). Data from tests using MITO in Livingston Island, Antarctica have been used to evaluate the feasibility of this method. The obtained directions have been found to be consistent with the expected incident directions of atmospheric muons produced by the interaction between CRs and atmospheric atoms.

Introduction

Primary Cosmic Ray (CRs) and Solar Energetic Particles (SEPs) interact with air nuclei when they arrive at the top of Earth’s atmosphere, producing secondary CRs. These secondary CRs, in turn, can interact with other nuclei and produce additional secondary particles. CRs and SEPs with energies above 500 MeV can produce secondary particles that can be measured by instruments operating at ground level. The CR secondaries most commonly measured at ground level are pions, muons, neutrons, protons, electrons and gammas being the muons the most abundant ones.

The main responsible of muon flux at ground level are primary CRs with energies from tens to hundreds of GeV. Otherwise, neutrons observed by neutron monitors at ground level are produced by primary CRs with energies from 500 MeV up to 50 GeV, that is around the detection limit of neutron monitors. Therefore, muon and neutron observations at ground level are complementary in this primary CR range of energies [1]. On the other hand, while the arrival direction of CRs at the magnetosphere limit is almost isotropic, there are studies that indicate that sometimes that isotropy breaks slightly in favor of certain directions as a result of the arrival of huge magnetic structures at Earth’s orbit, such as magnetic clouds embedded in interplanetary coronal mass ejections [2], [3]. Both muon flux measurement and determination of arrival direction at Earth’s surface is typically performed by telescope arrays, for instance, the Nagoya Multi-directional Muon Telescope [4], or the GRAPES-3 Experiment [5]. However, the Muon Impact Tracer and Observer (MITO) is a single telescope designed to measure both muon flux and incident directions. In this telescope, the incident trajectory is derived from the muon impact point observed at two piled 1 m2 scintillators [6], allowing the study of predominant directions. MITO is part of ORCA (Antarctic Cosmic Ray Observatory), which has been recently deployed by the University of Alcalá at the Juan Carlos I Scientific Spanish Base, in Antarctica [7].

ORCA is a combination of a neutron monitor, NEMO, which is a direct heritage of CaLMa (Castilla-La Mancha Neutron Monitor) [8], and the aforementioned MITO muon telescope. Its main objectives are to measure the flux of CRs in a region not covered by the Neutron Monitor Data Base (NMDB) and to study solar activity, which can be inferred from CR flux temporal variations. In this article, though, we will focus on MITO.

Muon tracking has traditionally been performed using multiple scintillators laid out in a two layer matrix (for instance, two layers of 6 × 6 scintillators), separated by a lead layer to filter out lower energy particles, with a photomultiplier (PMT) gathering the light generated at each scintillator. Another muon trackers are Nagoya (6 × 6 array of 1m2 detectors) [4], São Martinho da Serra (two layers of 4 × 8 m with scintillators of 1 m2) and Kuwait telescope (3 × 5 × 1 m with an intermediate lead layer) among others. When a coincidence is registered between two detectors, one at each layer, a trajectory can be determined limited by the resolution provided by each scintillator matrix and the distance between them. Apart from the resolution limitation, these instruments are usually very large and their construction cost is also very high, so another approach is to use just two large scintillators instead of two scintillator matrices, and determine the impact point at each of them in order to calculate a trajectory [9]. This is also the approach used in MITO, which obtains the point of impact by comparing the level of the pulses detected in several PMTs.

Basically, MITO is composed by a stack of two devices 136.5 cm apart from each other, each of them consisting of a scintillator and four photomultiplier tubes gathering the light emanating from its lateral sides. This allows the determination of the particle impact point at each device by means of pulse height analysis; and when the point of impact on each device is found, the angle of incidence of the particle can be obtained [10]. MITO will be described in detail in Section 2.

However, the determination of the impact position at each plane as a function of the measured pulse heights is difficult, not only because of the difficulty of developing a reliable reconstruction algorithm, but also because the measurement depends on multiple factors such as the response linearity of each PMT and the associated electronics or the ambient temperature. Furthermore, the response depends on the specific plane and PMT. On the other hand, the instrument would require a precise calibration process that should be repeated over time to ensure correct results as operating conditions change. What is proposed in this manuscript is a method that facilitates the determination of the point of impact and avoids this need for calibration, taking into account that the expected distribution of impact points follows a certain criteria obtained by simulation. To do so, we use a novel self-organize map called Self-Equalizing Map (SEM) that modifies the distribution function obtained by a simple reconstruction algorithm to tailor it to the expected distribution. The use of neural networks on particle detectors is not new, as they have been used to discriminate neutrons and gamma rays in scintillators [11] and more recently to maximize the Signal-to-Noise Ratio (SNR) [12]. The method to find out the point of impact is described in Section 3.

Section 4 describes the results obtained in the determination of impact points and incidence angles obtained from data captured by MITO during several days at the scientific base Juan Carlos I, located in Livingston Island (Antarctica). Finally, Section 5 covers the conclusions.

Section snippets

The experiment

MITO is an instrument composed of two identical devices, each consisting of a organic scintillator and four PMTs tubes mounted in an aluminum chassis. Each Saint-Gobain BC-400 scintillator is made of polyvinyl toluene with 65% anthracene, and is shaped like a square prism of 100 × 100 × 5 cm. The light yielded by the scintillator when a particle goes through it is transported via light guides and collected by 4 Hamamatsu R2154-02 PMTs, located in front of each lateral side of the prism, at each

Procedure to measure the angle of incidence

The purpose of this section is to elaborate a method to calculate the angle of impact of muons based on their point of impact on each of the two MITO planes. To calculate this, we estimate the distance between the impact point and each of the four PMTs at each plane, which is a function of the pulse height of each captured event. MITO, in its configuration in Antarctica, has a 10 cm thick lead layer located above the bottom scintillator. Although certain muon dispersion is expected because of

Results

Finally, a test with real data to check the proposed method has been performed. The data were collected using the MITO muon telescope from January 17, 2019 to January 20, 2019 in the Juan Carlos I scientific base, located in Livingston Island, Antarctica (S62°39’46”, W60°23’20”, 12 m asl).

As explained in Section 2, the raw data obtained from the eight PMTs was gathered using the ARACNE module and pulse height for each event in coincidence was stored in a text file that can be used repeatedly,

Conclusions

We have developed a method based on a novel self-organizative map called Self-Equalizing Map (SEM) to calculate the point of impact of a particle in a scintillator. This map is trained using unsupervised learning to produce a specific discretized distribution function preserving the topological properties of the input space. The learning process has been attained through the training of SEMs according to a previously generated simulation, assuming that the angle of incidence of muons is

CRediT authorship contribution statement

Alberto Regadío: Conceptualization, Methodology, Software, Formal analysis, Visualization, Writing - original draft. J. Ignacio García Tejedor: Visualization, Formal analysis, Validation, Investigation, Data curation, Writing - review & editing. Sindulfo Ayuso: Investigation, Validation, Writing - review & editing. Óscar García Población: Investigation. Juan José Blanco: Funding acquisition, Validation, Supervision, Writing - review & editing. Sebastián Sánchez-Prieto: Supervision, Writing -

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Thanks to project CTM2016-77325-C2-1-P funded by Ministerio de Economía y Competitividad and by the European Regional Development Fund, FEDER .

References (17)

There are more references available in the full text version of this article.

Cited by (0)

View full text