Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 337, 25-30 (1998)

Previous Section Next Section Title Page Table of Contents

4. Image analysis and Monte Carlo simulations

Prior to image analysis a data integrity check is performed to test for the presence of cloud or electronics problems. For a subset of observations tracking calibration is tested by monitoring changes in single-fold rates as local stars rotate through the field of view. Using this method the pointing direction can be inferred to an accuracy of around [FORMULA].

For each event trigger the tubes associated with the image are selected using the following criteria: (i) The TDC of the tube must indicate that the tube has exceeded its discrimination threshold within 50 ns around the trigger time of the event, and (ii) The ADC signal in the tube must be at least 1 standard deviation above the RMS of background noise for that tube.

An image is considered suitable for parameterization if it contains more than 4 selected tubes, and if the total signal for all tubes in the image exceeds 200 ADC counts (around 20 p.e.). About 25% of raw images are rejected by these two selection conditions. Surviving images are parameterized after Hillas 1985, with the gamma-ray domains for our observations being :

[FORMULA] Concentration [FORMULA]
alpha [FORMULA]

The cumulative percentages of events passing each selection condition are given in Table 1. The numbers shown are based on image cuts applied to Monte Carlo gamma-rays and protons. The efficiency of the cuts for the Monte Carlo protons is consistent with that seen for real data.


Table 1. Cumulative percentages of events passing gamma-ray selection cuts.

To estimate fluxes and upper limits for our data we use an exposure calculation based on a detailed Monte Carlo simulation of the response of our telescope to gamma-ray initiated EAS. Simulations have been based on the MOCCA simulation package (Hillas 1982) which models all relevant particle production processes and includes atmospheric absorption effects for the erenkov photons that are produced. The energies of simulated gamma-ray primaries were selected from a power law with integral exponent -1.4, with a minimum primary energy of 500 GeV. Core distances were selected from an appropriate distribution out to a limiting core distance of 250 m. Simulated gamma-ray images were then subjected to the same selection criteria as the real data. Fluxes and upper limits are calculated by comparing measured gamma-ray excess rates to those predicted by the simulations. If we assume that our telescope model is correct, this method of flux estimation has only two free parameters - the source spectral index and source cutoff energy. For total flux calculations the estimate of cutoff energy is not critical - the generally steep nature of gamma-ray source spectral indices ensures that the bulk of the flux is around the threshold energy.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: August 6, 1998