GRBM Observation

The strong outburst from SGR 1900+14triggered the GRBM on 1998 August 27 10:22:15.7 UT.

The event from SGR 1900+14was detected with a high signal-to-noise ratio by the 4 detection units, both in the main energy range (40-700 keV) and in the harder energy range (above 100 keV). The event occurred at an elevation angle of $48^{\circ}$ with respect to the GRBM equatorial plane, at an azimuthal angle of $29^{\circ}$ with respect to the unit #1 axis, and $61^{\circ}$ with respect to the unit #4. Since the shadow of the mirrors of LECS and MECS The average effective area of unit #1 at the event direction varies from $\sim$56 cm² at 60 keV to $\sim$365 cm² at 280 keV, while the count rates in units #2 and #3 are strongly affected by scattering effects. For the analysis presented here we used only the signal from unit #1 for which we have the highest signal-to-noise ratio and for which we have an analytical description of the efficiency as a function of energy and source direction. The estimated event duration is about 300 s. The total net counts were 1,007,699 in 40-700 keV and 301,235 above 100 keV. The net peak count rate in the (40-700 keV) energy range was 147,812 counts/s (the average background count rate was 880 counts/s) and 67,025 counts/s in the >100 keV band (the average background count rate was $\sim$1,000 counts/s). In particular, we corrected our 1-s light curve in the 40-700 keV range for one counter recycle (one recycle implies an additive contribution of 65535 counts/s), by comparing the 1-s resolution data with the high time resolution data (see below) and also comparing the GRBM light curve with that of Ulysses. The correction with only one recycle is what we consider the most likely correction to be applied. However some uncertainty remains, and we present lower limits to the peak flux and fluence.

The 40-700 keV peak flux is detected 1 s after the pulse rise and is $\gt 2.10 \times 10^{-4}$ erg cm^-2 s^-1. The measured fluence in the same band is $\gt 1.5 \times 10^{-3}$ erg cm^-2. Assuming a source distance of 5 kPc and isotropic emission, we obtain a luminosity of $6 \times 10^{41}$ erg s^-1 at the peak and a luminosity of $\sim3 \times 10^{40}$ erg s^-1 in the subsequent $\sim$70 s. The total energy detected by the GRBM in the 40-700 keV band corresponds to $\sim5 \times 10^{42}$ erg. These flux values are obtained assuming a power law spectrum whose index is computed on the 1-s time-scale from the two GRBM energy bands.

High time resolution data are only available for the first 98 s. After that time, only a 1-s resolved light curve is available. Given the intensity of the event, the counters of the high-resolution data were saturated (and therefore recycled several times) at the event peak. For this reason we cannot determine the event rise time for time-scales shorter than 1 s.

In Fig. 5.20 the 1-s background-subtracted light curve is shown (top panel), together with the 7.8 ms light curve (rebinned at 31.25 s, bottom panel, a) for the time period when it is not affected by saturation. The 5.16-s periodicity, is clearly detectable in the 1-s resolved light curve over-imposed to the general decay for the entire duration of the event. After a rapid decay during the first 2 seconds (faster in the 100-700 keV band, with respect to the 40-100 keV range), the decay can be approximated with two exponential laws, with time constant $\tau \sim$ 5 s for the first $\sim 15$ s and $\tau \sim$ 80 s for the subsequent decay. We also note a modulation of the oscillation with a $\sim$32 s period due to the sampling effect of a 5.16-s period at a 1 Hz frequency. The 31.25 ms light-curve shows that in addition to the 5.16-s periodicity after $35\sim40$ s from the beginning of the event the 5.16-s pulse is composed of 4 pulses equally separated in time by approximately 1 s and a dip (see next paragraph).