2 Observations

SIGMA is the coded-mask telescope operating in the 35-1300 keV energy band (Paul et al. 1991). The main field of view is $11.4^\circ\times10.5^\circ$, but some fraction of gamma-rays from sources within $\sim35^\circ$ to the pointing direction reaches the detector through the gaps in the passive shield and produces arc-shaped images. This "secondary optics'' was described in detail by Claret et al. (1994a).

  

Figure 1: GRB 920723 light curve with time resolution $\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}0.1$ in 35-300 keV band

Gamma-ray burst GRB 920723 was observed by SIGMA through the secondary optics. The burst lasted for $\approx 6$ s. The peak count rate in the 35-300 keV band was 7900 cnt s^-1, much higher than the average background rate 310 cnt s^-1. The WATCH all-sky monitor provided a $0.2^\circ$localization (Sazonov et al. 1998) and observed the fading X-ray emission in the 8-20 keV band during more than 40 s after the main burst (Terekhov et al. 1993). PHEBUS measured the peak flux $5\ 10^{-5}$ erg s^-1 cm^-2 and fluence $1.4\ 10^{-4}$ erg cm^-2 in the 100-500 keV energy band (Terekhov et al. 1995).

We present here the light curve of this burst observed by SIGMA. Since GRANAT operates on the high apogee orbit, the SIGMA background is very stable and usually does not show any significant variations on the time scales shorter than $\sim10^3$ s. Therefore, it can be accurately modeled by a low degree polynomial throughout the whole $\approx 20$ hour observation. The background can be reliably subtracted in the vicinity of the burst and the light curve can be accurately measured. A complete analysis of SIGMA background subtraction uncertainties is presented elsewhere (Burenin et al. 1999).

  

Figure 2: The background-subtracted burst light curve. Zero time is at the burst trigger