Background level and variations in the GRBM and AC ratemeters

The measured in-orbit background level in the GRBM and AC ratemeters ranges between $\sim$900 counts/s to $\sim$1100 counts/s. In Fig. 4.1 we show a typical time profile along two satellite orbits of the four GRBM and AC ratemeters with a default set of thresholds. The very good stability and low modulation of the background along the orbit in the GRBM band is apparent, as well as the number of spurious signals ("spikes") that make the time variation of the ratemeters not consistent with simple counting Poisson statistics. These spikes are mainly due to cosmic rays of high atomic number Z exciting meta-stable states in the CsI crystal, giving rise to a phosphorescence on short time scales, but long enough when compared to the response of the detector electronics, therefore allowing the electronics to detect a high number of counts. From the behavior of the simultaneous AC ratemeters, however, we see that the typical amplitude of the signals composing the spikes corresponds to a small equivalent energy. This fact was confirmed by the in-flight thresholds calibration, that showed that they progressively diminish as the energy threshold rises.

Since the satellite is in a low-earth, almost equatorial orbit it crosses the South Atlantic Geomagnetic Anomaly (SAGA) once per orbit, at a latitude that varies during the day from about +4 to -4 because of the Earth rotation. During the SAGA crossing, the experiment is switched off, to avoid fatiguing or damages to the PMT due to the very strong particles flux. The CsI activation during "deep" passages over the SAGA causes an increase of the measured background level when the instrument is turned on again.

  

Figure 4.1: Typical on-orbit background level of the four GRBM and LS ratemeters of the GRBM experiment with GRBM energy thresholds LLT=1 and ULT=6, and ACT=0. The data gaps are due to the satellite passage close to the South Atlantic Anomaly and therefore identify the single orbits. The low energy channel (GRBM) shows the presence of several spikes, whose associated energy is within the energy difference between the GRBM LLT and the ACT.

Moreover, an unpredicted situation whose origin is not clear and is still under study was discovered. In an orbit portion preceding the SAGA of about 10 minutes, sometime we detect an increase in the count rate of all detectors, mainly concentrated in the lower energies (i.e. smaller than 100 keV). The increase is typically negligible, but at times it becomes important, and rarely dramatic. We call this phenomenon "pre-SAGA", and we tried to investigate its origin. We first have searched for a correlation with solar flares, but we did not find anything significant. Considering that the phenomenon shows up in a fixed portion of the orbit, we expect it could be due to some Geomagnetic effect.

The excess intensity is at its maximum during those orbits with the most distant passage from the SAGA, that are clearly distinguishable thanks to the PDS Particle Monitor. In Fig. 4.2 the most dramatic example of the pre-SAGA we have detected is shown. As it is visible the pre-SAGA was exceptionally intense at time around -5000, and it was still intense in the next orbit. We note that during this second orbit, over-imposed to the pre-SAGA, we detected the famous GRB970228 (more visible from the expanded views in the two bottom panels). This GRB was triggered by the GRBM thanks to the fact that because of the pre-SAGA effect the LIT (see section 1.2.3) was set at 32 s(the initial value was 128 s), in order to properly follow these possible variation of the background.

  

Figure 4.2: GRB970228 as detected in the LS1 ratemeter (T= 0 s, corresponding to the trigger time: 1997, Feb. 28, 02:58:00 UT). The pre-SAGA anomaly is huge in the previous orbit and evident in the orbit of GRB970228 itself. In the middle panel the GRB region is enlarged in order to make GRB970228 visible. In the bottom panel the burst profile (31.25 ms bins) as seen with the high time resolution data, in which the minor structures of the first peak can be clearly identified.

From Fig. 4.1 it appears that the GRBM ratemeters of LS2 has a higher background with respect to the other three. This is due to the fact that, as discussed in section 2.5.3, this detector is illuminated by the four ¹⁰⁹Cd radioactive sources (main emission 88 keV, 462 days half life) that are part of the in-flight calibration system of the HPGSPC. The 88 keV line is efficiently detected by the shield LS2 in the energy range of the GRBM ratemeters, but due to its higher low energy threshold the AC ratemeter is not affected by the same problem. On the other hand, LS4 is influenced by another "local" source of background. This is the linear Co57 movable calibration source of the PDS experiment itself, whose parking position is located just above the top of shield LS4. The standard emission (122-136 keV) of the source is efficiently shielded by a properly heavy "parking strip", but a secondary line at 692 keV is emitted by the source with a branch of 0.16% (compared to the 68% and 3.8% of the main lines) and due to its high energy it is not efficiently shielded, therefore causing a higher background in the shield LS4.