In order to add new clues that may clarify whether the excess recorded with AIROBICC is related to GRB 920925c or not, the BATSE team (Kippen 1996) investigated the reason why BATSE was not triggered by this GRB. There are two reasons why BATSE may have missed the burst: either it was not strong enough to overwrite the previous trigger (happened about one hour before) or the Earth (or the Moon) occulted the source. The BATSE team kindly provided us with a map of the sky, as seen by BATSE (Kippen 1996), at the time that GRB 920925c occurred. It shows that the AIROBICC location was completely occulted by the Earth throughout the whole burst. Therefore the AIROBICC observation is compatible with BATSE not seeing the GRB. The WATCH position was on the Earth's limb at the beginning of the burst and was occulted about one minute later. The first peak of GRB 920925c (which is within the first minute after the initial trigger) had lower peak flux than the previous BATSE trigger and thus it would have not produced a trigger overwrite even neglecting the additional attenuation by the Earth's atmosphere. So the position given by WATCH is also consistent with the non-detection of the GRB with BATSE.
A cross-check of burst positions given by BATSE, WATCH and IPN since 1991 to 1994 reveals disagreements for some events at a level of 5 or even more, so an error in the WATCH location cannot be discarded (if the error is due to timing it would also affect the IPN location estimate). Furthermore it is interesting to note that assuming that the peak registered with AIROBICC corresponds to the peak which triggered ULYSSES, we can recompute the IPN ring using the time of detection of AIROBICC (neglecting the WATCH-AIROBICC distance compared to the ULYSSES-Earth distance) yielding a new IPN ring which is compatible with AIROBICC but not with WATCH (see Fig. 7).
At this moment it is not possible to decide the nature of the excess recorded with AIROBICC. Note that this GRB had a large fluence at keV energies, but was not very intense. The GRB occurred under a small zenith angle (12°) and could thus be studied with a low energy threshold of the AIROBICC array. The events of the excess differ significantly from background events as shown in Fig. 8. The figure plots the mean of the ratio of scintillator fired huts to AIROBICC fired huts for groups of 11 events. The histogram fitted by a Gaussian function represents the groups of background events. The vertical dashed line, which is at 2.0 as given by the Gaussian fit, represents the mean for the 11 events of the excess. It has been shown (Arqueros et al. 1996) that the ratio particles to light for EAS at observation level is sensitive to the chemical composition and the difference observed in Fig. 8 points to a gamma origin of the excess. This is confirmed by a MC study in which simulated EAS are analyzed in the same way as real data (Martínez et al. 1995; Cortina 1997). The results for real data and for MC are summarized in Table 4. If the excess has indeed gamma-ray nature its significance would be higher due to the much lower diffuse gamma-ray background compared to the charged particles background (Karle et al. 1995c). On the other hand the angular separation between AIROBICC and WATCH locations is significant.
Table 4. Comparison between the ratio of scintillator fired huts to AIROBICC fired huts for groups of 11 events as predicted by the MC and as seen in the excess and in the background of the real data. The MC seems to confirm a gamma origin of the excess.
© European Southern Observatory (ESO) 1998
Online publication: August 6, 1998