Cosmic Ray Nuclei Research Articles

Boron isotopes in the PAMELA experiment

In this work, a new analysis of the isotopic composition of boron nuclei (B) in galactic cosmic rays (GCR) in the range of rigidities of 1–5 GV (nuclear energies 0.1–1.5 GeV/nucleon) was carried out using data from the PAMELA space experiment 2006–2014 on the rigidity of detected nuclei and their velocity (time-of-flight analysis and ionization losses in the instrument’s multilayer calorimeter). The new results of the PAMELA experiment expand the energy range of previous measurements, are consistent with the few existing data, and indicate deviations of the B isotope ratios from the GALPROP simulation data for the GCR, similar to the deviations for the Li and Be isotopes in the PAMELA data, which can be interpreted as evidence of observation against the background of the GCR the contribution of several local sources from explosions of nearby (hundreds of parsec) supernovae.

Cosmic-ray propagation models elucidate the prospects for antinuclei detection

Tentative observations of cosmic-ray antihelium by the AMS-02 collaboration have re-energized the quest to use antinuclei to search for physics beyond the standard model. However, our transition to a data-driven era requires more accurate models of the expected astrophysical antinuclei fluxes. We use a state-of-the-art cosmic-ray propagation model, fit to high-precision antiproton and cosmic-ray nuclei (B, Be, Li) data, to constrain the antinuclei flux from both astrophysical and dark matter annihilation models. We show that astrophysical sources are capable of producing 𝒪(1) antideuteron events and 𝒪(0.1) antihelium-3 events over 15 years of AMS-02 observations. Standard dark matter models could potentially produce higher levels of these antinuclei, but showing a different energy-dependence. Given the uncertainties in these models, dark matter annihilation is still the most promising candidate to explain preliminary AMS-02 results. Meanwhile, any robust detection of antihelium-4 events would require more novel dark matter model building or a new astrophysical production mechanism.

Secondary cosmic-ray nuclei in the Galactic halo model with nonlinear Landau damping

We employed our recent model of the cosmic-ray (CR) halo to compute the Galactic spectra of stable and unstable secondary nuclei. In this model, confinement of the Galactic CRs is entirely determined by the self-generated Alfvénic turbulence whose spectrum is controlled by nonlinear Landau damping. We analyzed the physical parameters affecting propagation characteristics of CRs and estimated the best set of free parameters providing accurate description of available observational data. We also show that agreement with observations at lower energies may be further improved by taking into account the effect of ion-neutral damping that operates near the Galactic disk.

Transport parameters from AMS-02 F/Si data and fluorine source abundance

The AMS-02 collaboration recently released cosmic ray F/Si data with an unprecedented accuracy. Cosmic ray fluorine is predominantly produced by fragmentation of heavier progenitors, while silicon is mostly accelerated at source. This ratio is thus maximally sensitive to cosmic ray propagation. We study the compatibility of the transport parameters derived from the F/Si ratio with those obtained from the lighter Li/C, Be/C, and B/C ratios. We also inspect the cosmic ray source abundance of F, which is one of the few elements that has a high first ionisation potential but is only moderately volatile and is a potentially key element to study the acceleration mechanism of cosmic rays. We used the 1D diffusion model implemented in the code and performed $ analyses accounting for several systematic effects (energy correlations in data, nuclear cross sections, and solar modulation uncertainties). We also took advantage of the nuclear database to update the F production cross sections for its most important progenitors (identified to be 56Fe, 32S, 28Si, 27Al, 24Mg, 22Ne, and 20Ne). The transport parameters obtained from AMS-02 F/Si data are compatible with those obtained from AMS-02 (Li,Be,B)/C data. The combined fit of all of these ratios leads to a $ 1.1$, with $ 10<!PCT!>$ adjustments of the B and F production cross sections (which are based on very few nuclear data points and would strongly benefit from new measurements). The F/Si ratio is compatible with a pure secondary origin of F, with a best-fit relative source abundance F/^ Si) CRS $ and an upper limit of $ $. Unfortunately, this limit is not sufficient to test global acceleration models of cosmic ray nuclei, for which values at the level of $ $ are required. Such levels could be attained with F/Si data with a few percent of accuracy at a few tens of TV, which is possibly within reach for the next generation of cosmic ray experiments.

Inelastic Scattering of Dark Matter with Heavy Cosmic Rays

We investigate the impact of inelastic collisions between dark matter (DM) and heavy cosmic ray (CR) nuclei on CR propagation. We approximate the fragmentation cross-sections for DM-CR collisions using collider-measured proton-nuclei scattering cross-sections, allowing us to assess how these collisions affect the spectra of CR boron and carbon. We derive new CR spectra from DM-CR collisions by incorporating their cross-sections into the source terms and solving the diffusion equation for the complete network of reactions involved in generating secondary species. In a specific example with a coupling strength of b χ = 0.1 and a DM mass of m χ = 0.1 GeV, considering a simplified scenario where DM interacts exclusively with oxygen, a notable modification in the boron-to-carbon spectrum due to the DM-CR interaction is observed. Particularly, the peak within the spectrum, spanning from 0.1 to 10 GeV, experiences an enhancement of approximately 1.5 times. However, in a more realistic scenario where DM particles interact with all CRs, this peak can be amplified to twice its original value. Utilizing the latest data from AMS-02 and DAMPE on the boron-to-carbon ratio, we estimate a 95% upper limit for the effective inelastic cross-section of DM-proton as a function of DM mass. Our findings reveal that at m χ ≃ 2 MeV, the effective inelastic cross-section between DM and protons must be less than O(10−32)cm2 .

Loaded layer-cake model for cosmic ray interaction around exploding super-giant stars making black holes

The experiment AMS on the International Space-Station has produced accurate cosmic ray spectra for many chemical elements, both primaries like He, C, O, Fe, other cosmic ray (CR) primaries like Ne, Mg, and Si, secondaries like Li, Be, B, and of mixed provenance, like N, Na, and Al. The AMS spectra demonstrate that interaction is seriously diminishing fluxes up to a rigidity of about 100 GV, and so the existing models for CR interaction have to be re-examined. Based on earlier well-established ideas a model is proposed here that focusses on the cosmic ray interaction first in the wind shock shell of super giant stars, when the supernova driven shock races through, and second in the OB-Superbubble surrounding the SN: These stars include both red super-giant stars and blue super-giant stars; both produce black holes in their explosion, and drive winds and jets with electric currents. Variability of these winds or jets gives rise to temporary electric fields, as has recently been demonstrated, and discharge (so lightning) acceleration gives steep spectra, with synchrotron losses to p−5 in momentum p; these spectra are typically observed in both Galactic and some extragalactic radio filaments. Analogous hadron spectra p−4 excite a flat spectrum of magnetic irregularities in the bubble zone, which in turn yields a steep dependence of residence time versus energy, with power −5/3. This spectrum is indicated by the AMS data and appears to be required to explain the CR spectra below 100 GV. The emphasis in this paper is to work out the interaction of the freshly accelerated cosmic ray particles. In the model presented here the interaction is derived as a function of time, and then integrated, or developed to long times. The model gives a rigidity dependence of the secondary/primary ratio of slope −1/3 as well as the strong reduction of the primary fluxes below a rigidity of about 100 GV, relative to a power-law injection spectrum, with slope +2. The two key aspects based on blue super-giant stars and a magnetic irregularity spectrum in the bubble zone given by lightning are (i) a much larger column of interaction, allowed because of heavy element enrichment of the interaction zone, and (ii) even He, C, and O may have a small secondary contribution, as the difference to the Fe spectrum suggests; this small secondary component is visible in the 3He/4He ratio. The model may also explain the spectrum of CR anti-protons, the gamma-ray spectra of the Galaxy and the high energy neutrino spectrum of our Galaxy, including also red super-giant stars as sources. ISM-SNe, i.e. SN Ia and neutron-star SNe, contribute to CR protons and CR He nuclei.

Direct measurements of cosmic – Ray iron and nickel with CALET on the International Space Station

Iron and nickel cosmic ray nuclei play a key role in the understanding of the acceleration and propagation mechanisms of charged particles in our Galaxy. In fact, iron and nickel are the most abundant nuclei among the heavy elements and provide favorable conditions for a low background measurement thanks to the negligible contamination from spallation of higher mass elements. CALET, operating on the ISS since 2015, has excellent capabilities of charge discrimination up to nickel and can measure the energy of cosmic ray nuclei thanks to a lead tungstate calorimeter providing a direct and precise measurement of heavy charged nuclei spectra. In this contribution, a direct measurement of iron and nickel nuclei spectra in the energy range from 10 GeV/n to 2 TeV/n and from 8.8 GeV/n to 240 GeV/n, respectively is presented. More than five years of data collected by CALET were used. A detailed study of systematic uncertainties is also illustrated. The measured spectra are compared with the ones measured by other experiments and are compatible with a single power law fit in the energy region from 50 GeV/n to 2 TeV/n and from 20 GeV/n to 240 GeV/n for iron and nickel respectively. Also, the ratio between nickel and iron spectra is reported.

The Silicon Charge Detector of the High Energy Cosmic Radiation Detection experiment

The High Energy Cosmic Radiation Detector (HERD) is an experiment designed for direct measurement of cosmic rays on the Chinese Space Station in 2027. Its goals include precise measurements of cosmic ray energy spectra, mass composition, electron/positron spectra, cosmic rays anisotropy, gamma ray astronomy and indirect searches for Dark Matter. HERD features a 55X0, homogeneous, 3D segmented imaging calorimeter and can detect particles from the top and four lateral sides, providing precise energy measurements and electron/proton separation for a wide field of view. A key detector in HERD is the Silicon Charge Detector (SCD). SCD measures the charge of particles before interaction with other materials, minimizing cosmic rays nuclei fragmentation and reducing systematics on nuclei flux measurement. Thorough studies, TCAD and/or SPICE simulations, and accelerator tests on prototypes have been conducted to evaluate the tracking and charge resolution capabilities of the SCD. Further testing with 300 μm detectors is planned in the coming months to fully characterize the SCD's performance. This paper presents the results of the simulation studies and of the measured performance with particles.

Probing the angular distribution of terrestrial cosmic-ray nuclei

Understanding how atmospheric cosmic-rays are distributed on Earth is crucial for simulating cosmic-ray setups and detectors for rare-event detection. The accuracy and reliability of such simulations and experiments depend on obtaining this information. This study focuses on analyzing the impact of key parameters in the PARMA model on the angular distributions of terrestrial cosmic-ray nuclei. These parameters include geomagnetic cut-off rigidity, atmospheric depth and solar activity. Moreover, angular distributions of various particles, including protons, electrons, neutrons, and muons, were systematically calculated over relevant rigidities, solar modulations, and altitudes. For this purpose, the latest version of the Excel-based Program for calculating Atmospheric Cosmic-ray Spectrum EXPACS (v.4.13), which relies on updated experimental data, was used. The results of the calculations showed clear trends in the angular distributions of all cosmic-ray species as a function of the model parameters. These findings provide valuable insights into the dynamics of cosmic rays in our solar system. Moreover, they can contribute to developing more accurate models for predicting the distribution of cosmic rays on the Earth's surface.

Prospects for combined analyses of hadronic emission from γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray sources in the Milky Way with CTA and KM3NeT

The Cherenkov Telescope Array and the KM3NeT neutrino telescopes are major upcoming facilities in the fields of γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray and neutrino astronomy, respectively. Possible simultaneous production of γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document} rays and neutrinos in astrophysical accelerators of cosmic-ray nuclei motivates a combination of their data. We assess the potential of a combined analysis of CTA and KM3NeT data to determine the contribution of hadronic emission processes in known Galactic γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray emitters, comparing this result to the cases of two separate analyses. In doing so, we demonstrate the capability of Gammapy, an open-source software package for the analysis of γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray data, to also process data from neutrino telescopes. For a selection of prototypical γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray sources within our Galaxy, we obtain models for primary proton and electron spectra in the hadronic and leptonic emission scenario, respectively, by fitting published γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray spectra. Using these models and instrument response functions for both detectors, we employ the Gammapy package to generate pseudo data sets, where we assume 200 h of CTA observations and 10 years of KM3NeT detector operation. We then apply a three-dimensional binned likelihood analysis to these data sets, separately for each instrument and jointly for both. We find that the largest benefit of the combined analysis lies in the possibility of a consistent modelling of the γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray and neutrino emission. Assuming a purely leptonic scenario as input, we obtain, for the most favourable source, an average expected 68% credible interval that constrains the contribution of hadronic processes to the observed γ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\gamma $$\\end{document}-ray emission to below 15%.

Origin of 60Fe nuclei in cosmic rays: the contribution of local OB associations

ABSTRACT The presence of live 60Fe nuclei (lifetime of 3.8 Myr) in cosmic rays detected by the Advanced Composition Explorer/Cosmic Ray Isotope Spectrometer instrument suggests a nearby nucleosynthesis source. 60Fe is primarily produced in core-collapse supernovae, and we aim to clarify whether the detected 60Fe nuclei can be associated with a particular local supernova. We consider 25 OB associations and subgroups located within 1 kpc of the Solar system based on recent Gaia census. A model is developed that combines stellar population synthesis within these OB associations, cosmic ray acceleration within associated superbubbles, and cosmic ray transport to the Solar system. The most critical model parameter impacting 60Fe cosmic ray production is the explodability criterion, which determines if a massive star ends its life as a supernova. Our study points to the Scorpius–Centaurus (Sco–Cen) OB association as the most probable origin of the observed 60Fe nuclei, particularly suggesting they were accelerated in the Sco–Cen superbubble by a young supernova aged ≤500 kyr with a progenitor mass of approximately 13–20 M⊙. A less likely source is the supernova at the origin of the Geminga pulsar 342 kyr ago, if the progenitor originated in the Orion OB1 association. The contribution of local OB associations to the cosmic ray density of stable 56Fe is estimated to be around 20 per cent, with some sensitivity to cosmic ray acceleration efficiency and diffusion coefficient. These findings shed light on the origins of cosmic ray nuclei, connecting them to nucleosynthesis events within our local cosmic neighbourhood.

Between the Cosmic-Ray “Knee” and the “Ankle”: Contribution from Star Clusters

We show that massive, young star clusters may be possible candidates that can accelerate Galactic cosmic rays (CRs) in the range of 107–109 GeV (between the “knee” and “ankle”). Various plausible scenarios, such as acceleration at the wind termination shock and supernova shocks inside these young star clusters, have been proposed, since it is difficult to accelerate particles up to the 107–109 GeV range in the standard paradigm of CR acceleration in supernova remnants. We consider a model for the production of different nuclei in CRs from massive stellar winds using the observed distribution of young star clusters in the Galactic plane. We present a detailed calculation of CR transport in the Galaxy, taking into account the effect of diffusion, interaction losses during propagation, and particle reacceleration by old supernova remnants to determine the all-particle CR spectrum. Using the maximum energy estimate from the Hillas criterion, we argue that a young, massive star cluster can accelerate protons up to a few tens of PeV. Upon comparison with the observed data, our model requires a CR source spectrum with an exponential cutoff of 5 × 107 Z GeV (50 Z PeV) from these clusters, together with a CR injection fraction of ∼5% of the wind kinetic energy. We discuss the possibility of achieving these requirements in star clusters, as well as the associated uncertainties, in the context of considering star clusters as the natural accelerator of the “second component” of Galactic CRs.

The CaloCube calorimeter for high-energy cosmic-ray measurements in space: Response of a large-scale prototype to protons

Direct observation of cosmic rays nuclei is currently limited to energies of the order of hundreds of TeV. In order to extend these observations to higher energies, detectors capable of operating in space with high geometric factor and energy resolution are needed. In particular, highly performing calorimeters based on the CaloCube design can allow to carry out cosmic ray measurements in the PeV energy region. The CaloCube R&D project foresees the installation in space of a homogeneous and isotropic calorimeter composed of cubic scintillator crystals arranged to form a cube of about tons weight, with a high acceptance and capable of detecting particles coming from any direction. A prototype, composed of 5 × 5 × 18 CsI(Tl) crystals, has been tested on high-energy particle beams at CERN SPS accelerator and the results relative to the calorimeter response to protons are reported in this document.

The High Energy Light Isotope eXperiment program of direct cosmic-ray studies

HELIX is a new NASA-sponsored instrument aimed at measuring the spectra and composition of light cosmic-ray isotopes from hydrogen to neon nuclei, in particular the clock isotopes 10Be (radioactive, with 1.4 Myr lifetime) and 9Be (stable). The latter are unique markers of the production and Galactic propagation of secondary cosmic-ray nuclei, and are needed to resolve such important mysteries as the proportion of secondary positrons in the excess of antimatter observed by the AMS-02 experiment. By using a combination of a 1 T superconducting magnet spectrometer (with drift-chamber tracker) with a high-resolution time-of-flight detector system and ring-imaging Cherenkov detector, mass-resolved isotope measurements of light cosmic-ray nuclei will be possible up to 3 GeV/n in a first stratospheric balloon flight from Kiruna, Sweden to northern Canada, anticipated to take place in early summer 2024. An eventual longer Antarctic balloon flight of HELIX will yield measurements up to 10 GeV/n, sampling production from a larger volume of the Galaxy extending into the halo. We review the instrument design, testing, status and scientific prospects.

Leptonic Model for Neutrino Emission from Active Galactic Nuclei.

It is often stated that the observation of high-energy neutrinos from an astrophysical source would constitute indisputable proof for the acceleration of hadronic cosmic rays. Here, we point out that there exists a purely leptonic mechanism to produce TeV-scale neutrinos in astrophysical environments. In particular, very high-energy synchrotron photons can scatter with x rays, exceeding the threshold for muon-antimuon pair production. When these muons decay, they produce neutrinos without any cosmic-ray protons or nuclei being involved. In order for this mechanism to be efficient, the source in question must produce very high-energy photons which interact in an environment that is dominated by keV-scale radiation. We find that such a source could potentially generate an observable neutrino flux through muon pair production for reasonable choices of physical parameters.

The Milky Way revealed to be a neutrino desert by the IceCube Galactic plane observation

The Galactic diffuse emission (GDE) is formed when cosmic rays leave the sources where they were accelerated, diffusively propagate in the Galactic magnetic field and interact with the interstellar medium and interstellar radiation field. GDE in γ-rays (GDE-γ) has been observed up to subpetaelectronvolt energies, although its origin may be explained by either cosmic-ray nuclei or electrons. Here we show that the γ-rays accompanying the high-energy neutrinos recently observed by the IceCube Observatory from the Galactic plane have a flux that is consistent with the GDE-γ observed by the Fermi-LAT and Tibet ASγ experiments around 1 TeV and 0.5 PeV, respectively. The consistency suggests that the diffuse γ-ray emission above ~1 TeV could be dominated by hadronuclear interactions, although a partial leptonic contribution cannot be excluded. Moreover, by comparing the fluxes of the Galactic and extragalactic diffuse emission backgrounds, we find that the neutrino luminosity of the Milky Way is one-to-two orders of magnitude lower than the average of distant galaxies. This finding implies that our Galaxy has not hosted the type of neutrino emitters that dominates the isotropic neutrino background at least in the past few tens of kiloyears.

Unstable cosmic ray nuclei constrain low-diffusion zones in the Galactic disc

ABSTRACT Observations of the vicinity of a variety of galactic gamma-ray sources have indicated a local suppression of diffusivity of cosmic rays (CRs) by up to three orders of magnitude. However, the impact of these low-diffusion zones on global properties of CR transport is, however, only poorly understood. Here, we argue that CR nuclear ratios, like the boron-to-carbon ratio and relative abundances of Beryllium isotopes are sensitive to the filling fraction of such low-diffusion zones and hence their measurements can be used to constrain the typical sizes and ages of such regions. We have performed a careful parameter study of a CR transport model that allows for different diffusion coefficients κdisc and κhalo in the galactic disc and halo, respectively. Making use of preliminary data from the AMS-02 experiment on the ratio of Beryllium isotopes, we find a 3.5σ preference for a suppression of the diffusion coefficient in the disc with a best-fitting value of $\kappa _{\mathrm{disc}}/\kappa _{\mathrm{halo}} = 0.20^{+0.10}_{-0.06}$. We forecast that with upcoming data from the HELIX balloon experiment, the significance could increase to 6.8σ. Adopting a coarse-graining approach, we find that such a strong suppression could be realized if the filling fraction of low-diffusion zones in the disc was $\sim 66~{{\ \rm per\ cent}}$. We conclude that the impact of regions of suppressed diffusion might be larger than usually assumed and ought to be taken into account in models of Galactic CR transport.

Antistars as possible sources of antihelium cosmic rays

A minor population of antistars in galaxies has been predicted by some of non-standard models of baryogenesis and nucleosynthesis in the early Universe, and their presence is not yet excluded by the currently available observations. Detection of an unusually high abundance of antinuclei in cosmic rays can probe the baryogenesis scenarios in the early Universe. Recent report of the AMS-02 collaboration on the tentative detection of a few antihelium nuclei in GeV cosmic rays provided a great hope on the progress in this issue. We discuss possible sources of antinuclei in cosmic rays from antistars which are predicted in a modified Affleck-Dine baryogenesis scenario by Dolgov and Silk (1993). The model allows us to estimate the expected fluxes and isotopic content of antinuclei in the GeV cosmic rays produced in scenarios involving antistars. We show that the flux of antihelium CRs reported by the AMS-02 experiment can be explained by Galactic anti-nova outbursts, thermonuclear anti-SN Ia explosions, a collection of flaring antistars, or an extragalactic source with abundances not violating existing gamma-ray and microlensing constraints on the antistar population.

Fragmentation of Cosmic Ray Nuclei and Its Experimental Study

Fragmentation of Cosmic Ray Nuclei and Its Experimental Study

Application of Machine Learning Ensemble Methods to ASTRI Mini-Array Cherenkov Event Reconstruction

The Imaging Atmospheric Cherenkov technique has opened up previously unexplored windows for the study of astrophysical radiation sources in the very high-energy (VHE) regime and is playing an important role in the discovery and characterization of VHE gamma-ray emitters. However, even for the most powerful sources, the data collected by Imaging Atmospheric Cherenkov Telescopes (IACTs) are heavily dominated by the overwhelming background due to cosmic-ray nuclei and cosmic-ray electrons. As a result, the analysis of IACT data necessitates the use of a highly efficient background rejection technique capable of distinguishing a gamma-ray induced signal through identification of shape features in its image. We present a detailed case study of gamma/hadron separation and energy reconstruction. Using a set of simulated data based on the ASTRI Mini-Array Cherenkov telescopes, we have assessed and compared a number of supervised Machine Learning methods, including the Random Forest method, Extra Trees method, and Extreme Gradient Boosting (XGB). To determine the optimal weighting for each method in the ensemble, we conducted extensive experiments involving multiple trials and cross-validation tests. As a result of this thorough investigation, we found that the most sensitive Machine Learning technique applied to our data sample for gamma/hadron segregation is a Stacking Ensemble Method composed of 42% Extra Trees, 28% Random Forest, and 30% XGB. In addition, the best-performing technique for energy estimation is a different Stacking Ensemble Method composed of 45% XGB, 27.5% Extra Trees, and 27.5% Random Forest. These optimal weightings were derived from extensive testing and fine-tuning, ensuring maximum performance for both gamma/hadron separation and energy estimation.