On the use of neural networks for energy reconstruction in high-granularity calorimeters

We contrasted the performance of deep neural networks — Convolutional Neural Network (CNN) and Graph Neural Network (GNN) — to current state of the art energy regression methods in a finely 3D-segmented calorimeter simulated by GEANT4. This comparative benchmark gives us some insight to assess the particular latent signals neural network methods exploit to achieve superior resolution. A CNN trained solely on a pure sample of pions achieved substantial improvement in the energy resolution for both single pions and jets over the conventional approaches. It maintained good performance for electron and photon reconstruction. We also used the Graph Neural Network (GNN) with edge convolution to assess the importance of timing information in the shower development for improved energy reconstruction. We implement a simple simulation based correction to the energy sum derived from the fraction of energy deposited in the electromagnetic shower component. This serves as an approximate dual-readout analogue for our benchmark comparison. Although this study does not include the simulation of detector effects, such as electronic noise, the margin of improvement seems robust enough to suggest these benefits will endure in real-world application. We also find reason to infer that the CNN/GNN methods leverage latent features that concur with our current understanding of the physics of calorimeter measurement.

Segmentation of EM showers for neutrino experiments with deep graph neural networks

We introduce a first-ever algorithm for the reconstruction of multiple showers from the data collected with electromagnetic (EM) sampling calorimeters. Such detectors are widely used in High Energy Physics to measure the energy and kinematics of in-going particles. In this work, we consider the case when many electrons pass through an Emulsion Cloud Chamber (ECC) brick, initiating electron-induced electromagnetic showers, which can be the case with long exposure times or large input particle flux. For example, SHiP experiment is planning to use emulsion detectors for dark matter search and neutrino physics investigation. The expected full flux of SHiP experiment is about 1020 particles over five years. To reduce the cost of the experiment associated with the replacement of the ECC brick and off-line data taking (emulsion scanning), it is decided to increase exposure time. Thus, we expect to observe a lot of overlapping showers, which turn EM showers reconstruction into a challenging point cloud segmentation problem. Our reconstruction pipeline consists of a Graph Neural Network that predicts an adjacency matrix and a clustering algorithm. We propose a new layer type (EmulsionConv) that takes into account geometrical properties of shower development in ECC brick. For the clustering of overlapping showers, we use a modified hierarchical density-based clustering algorithm. Our method does not use any prior information about the incoming particles and identifies up to 87% of electromagnetic showers in emulsion detectors. The achieved energy resolution over 16,577 showers is σE/E = (0.095 ± 0.005) + (0.134 ± 0.011)/√(E). The main test bench for the algorithm for reconstructing electromagnetic showers is going to be SND@LHC.

Direct Measurements of Cosmic Rays (TeV and beyond) Using an Ultrathin Calorimeter: Lessening Fluctuation Method

In this paper, we propose a method that makes it possible to use an ultrathin calorimeter for direct measurements of cosmic rays with energies of TeV and higher. The problems of determining the primary energy with a thin calorimeter, due to large fluctuations in shower development, the low statistics of analyzed events and the large size required for the calorimeter, are considered in detail. A solution to these problems is proposed on the basis of a lessening fluctuation method. This method is based on the assumption of the universality of the development of cascades initiated by particles of the same energy and mass. For energy reconstruction, so-called correlation curves are used. The main analyzed quantities are the size of the cascade and the rate of its development. The method was tested using the calorimeter of the PAMELA collaboration. Based on simulations, it is shown that the primary energy can be determined on the ascending branch of the cascade curve. This fact solves the problems associated with the need to increase the calorimeter thickness with an increase in primary energy and with the limitation of the analyzed events. The proposed technique is universal for different energies and different nuclei.

Particle identification

The identity of a particle is fixed by its mass, lifetime and quantum numbers such as charge, spin, parity and flavour. A particle’s identity can be inferred by observing its interactions in matter, as for example the shower development of an electron or a photon, the specific energy loss of charged particles, the emission of radiation by a particle or the penetration capability of a muon. The mass of a particle can be determined by measurements of specific energy loss, time-of-flight or Cherenkov radiation when combined with a momentum measurement. High energy electrons can be separated from heavier particles through transition radiation. For particles which decay in the detector the mass can often be kinematically reconstructed from the decay products and the lifetime can be determined by the reconstruction of secondary vertices.

Atmospheric Muons Measured with IceCube

IceCube is a cubic-kilometer Cherenkov detector in the deep ice at the geographic South Pole. The dominant event yield is produced by penetrating atmospheric muons with energies above several 100 GeV. Due to its large detector volume, IceCube provides unique opportunities to study atmospheric muons with large statistics in detail. Measurements of the energy spectrum and the lateral separation distribution of muons offer insights into hadronic interactions during the air shower development and can be used to test hadronic models. We will present an overview of various measurements of atmospheric muons in IceCube, including the energy spectrum of muons between 10 TeV and 1 PeV. This is used to derive an estimate of the prompt contribution of muons, originating from the decay of heavy (mainly charmed) hadrons and unflavored mesons. We will also present measurements of the lateral separation distributions of TeV muons between 150m and 450m for several initial cosmic ray energies between 1 PeV and 16 PeV. Finally, the angular distribution of atmospheric muons in IceCube will be discussed.

Average shape of longitudinal shower profiles measured at the Pierre Auger Observatory

The average profiles of cosmic ray shower development as a function of atmospheric depth are measured for the first time with the Fluorescence Detectors at the Pierre Auger Observatory. The profile shapes are well reproduced by the Gaisser-Hillas parametrization at the 1% level in a 500 g/cm2 interval around the shower maximum, for cosmic rays with log(E/eV) > 17.8. The results are quantified with two shape parameters, measured as a function of energy. The average profiles carry information on the primary cosmic ray and its high energy hadronic interactions. The shape parameters predicted by the commonly used models are compatible with the measured ones within experimental uncertainties. Those uncertainties are dominated by systematics which, at present, prevent a detailed composition analysis.

Characteristics of air showers with energy more than 1017 eV reconstructed by the Yakutsk array radio emission measurements

The paper presents results on the longitudinal development of air showers of ultra-high energies obtained from radio emission measurements at the Yakutsk array. The energy, the depth of maximum development of individual showers are determined and a statistical analysis of Xmax in order to estimate the fluctuation of air shower development σ(Xmax) in the energy region 1017-1018 eV is performed. It is shown that σ(Xmax) in the energy region 1017-1018 eV is equal to 50-60 g·cm-2, which doesn’t contradict with a mixed composition of cosmic rays - protons and helium nuclei. This is also indicated by data of the Xmax value dependence on energy.

Determination of cosmic-ray primary mass on an event-by-event basis using radio detection

Traditionally, the depth of maximum shower development Xmax has been used as a surrogate observable for composition. Here we present the possibility of a new methodology to discriminate between light and heavy cosmic-ray primaries on an event-by-event basis. This method is based on comparisons between detected radio signals and Monte Carlo simulations, but instead of first reconstructing Xmax, we try to infer the cosmic-ray composition directly. We show that a large discrimination efficiency could in principle be reached for zenith angles above θ≃65°, even when some of the typical uncertainties in radio detection are taken into account.