Detection of the Crab Nebula using a randomforest analysis of the first TAIGA IACT data

Abstract

ABSTRACT The Tunka Advanced Instrument for Gamma- and cosmic-ray Astronomy (TAIGA) is a multicomponent experiment for the measurement of TeV to PeV gamma- and cosmic rays. Our goal is to establish a novel hybrid direct air shower technique, sufficient to access the energy domain of the long-sought Pevatrons. The hybrid air Cherenkov light detection technique combines the strengths of the HiSCORE shower front sampling array, and two ∼4 m class, ∼9.6° field of view Imaging Air Cherenkov Telescopes (IACTs). The HiSCORE array provides good angular and shower core position resolution, while the IACTs provide the image shape and orientation for gamma-hadron separation. In future, an additional muon detector will be used for hadron tagging at ≥100 TeV energies. Here, only data from the first IACT of the TAIGA experiment are used. A randomforest algorithm was trained using Monte Carlo (MC) simulations and real data, and applied to $85\, \mathrm{h}$ of selected observational data tracking the Crab Nebula at a mean zenith angle of 33.5°, resulting in a threshold energy of 6 TeV for this data set. The analysis was performed using the gammapy package. A total of 163.5 excess events were detected, with a statistical significance of 8.5 σ. The observed spectrum of the Crab Nebula is best fit with a power law above 6 TeV with a flux normalization of (3.20 ± 0.42) · 10−10 TeV−1 cm−2 s−1at a reference energy of $13\, \mathrm{TeV}$ and a spectral index of −2.74 ± 0.16.