Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 359, 682-694 (2000)

Previous Section Next Section Title Page Table of Contents

3. Monte Carlo simulations

EAS events were simulated using the CORSIKA code in its versions 5.20 with the QGSJET/GHEISHA options (Knapp & Heck 1993; Heck et al. 1998)). This generator is based on the quark-gluon string model (QGS) with an allowance for semihard processes (JET) (Kalmykov et al. 1997). Complex nuclei were treated with the "complete fragmentation" ansatz. The energy cutoff for particles of the electromagnetic cascade was set to 3 MeV. Proton, helium, oxygen and iron induced showers were produced at zenith angles of 0o,6o, 12o and 18o at discrete energies between 300 TeV and 10 PeV (4400 independent showers of which 1000 are above 1 PeV) as well as an independent sample with a continuous energy distribution at zenith angles of 6o and 12o between 50 TeV and 1 PeV (4330 independent showers) and following a power law of E-1 between 2.5 PeV and 6.5 PeV (240 showers). The events continuously distributed in energy were spectrally weighted and used in the fits to infer the chemical composition (Sect. 5.2), while the samples with discrete energies were employed to develop the reconstruction methods and to correct the results obtained with biased estimators of the primary energy (Sect. 5.1). Note that the simulation of an EAS induced by a 1 PeV primary proton including the Cherenkov light production requires about 3 h CPU time on a 300 MHz Pentium-II PC (during the same time the HEGRA experiment registers more than 350 showers with energies larger than 1 PeV).

The detector performance was modelled with two independent detector simulations: a full detector simulation (Martinez et al. 1995), and an empirical simulation using measured response functions (Haustein 1996; Horns 1997). Each independent generated EAS was used 20 times with core positions inside and outside the HEGRA area to take into account the detector related fluctuations of observables and to check the event selection criteria. With the standard cuts described below, each shower was used on the average two times in the discrete and once in the continuous MC sample.

Special care has been taken to simulate the density profile and absorption features of the atmosphere above La Palma correctly. Weather balloon measurements as well as comparisons between TeV photon data registered by the HEGRA system of imaging air Cherenkov telescopes and simulations were employed for this purpose (Bernlöhr 1999; Konopelko et al. 1999). The shower development and light emission were modelled with the U.S. standard atmosphere, and the light propagation was then simulated with a special program able (Haustein 1996), assuming a tropical maritime atmosphere for the summer. This atmosphere is a good approximation for the conditions at Tenerife, an island neighbouring the experimental site (Bernlöhr 1999).

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: July 7, 2000