3 Expected performance

The expected performance in terms of sensitivity is summarised in Fig. 1. This is the performance we expect for the second phase of the project in which high quantum efficiency hybrid photo sensors will replace the photo-multipliers of the telescope's camera. In the first phase our threshold will be $\approx 30$ GeV, in the second phase 10 GeV. It is not yet clear how soon the second phase will follow the first. The sensitivity at energies > 30 GeV is to a good approximation independent of these changes.

  

Figure 1: The fluence sensitivity of the MAGIC Telescope (MAGICT) compared to other experiments (MAGICT refers to observations at zenith angles up to $\approx$ 30$^\circ$, "MAGICT large ZA'' to those at $\approx$ 70$^\circ$). The fluence sensitivity at threshold E0 is defined here as the total fluence above E0 which is necessary for a $5 \sigma$ detection of the burst above that threshold. A total burst duration of 60 s and a differential spectral index of 2.6 is assumed. For MAGICT the actual observation time is assumed to be only 30 s taking into account the reaction time. For all other detectors the observation time is assumed to be equal to the burst duration

Given the diverse shape of GRB light-curves it is difficult to predict an average counting rate for gamma-rays in successful burst observations. We note instead that the MAGIC Telescope will have an effective collection area for primary gamma photons of 10⁸ cm² at the threshold rising to 10⁹ cm² at 100 GeV. For a hypothetical counting rate of 0.1Hz for EGRET above 100 MeV, we expect (assuming a spectral index of 2.0) a counting rate of $\approx 100$ Hz above 10 GeV. A rate of $\approx 6$ Hz for 30 s will suffice for a 5 $\sigma$ detection with a moderate background rejection applied. For strong bursts such as GRB 930131, gamma counting rates of the order of 1 kHz may occur. The data acquisition will therefore be prepared to sustain such rates without additional dead-time.