Off-line Quest Limits

The ratio between the number of on-board triggered that could not have been scanned (693) and the total number of on-board triggers (15,857) also reflects the fraction of 1 s bins that have not been scanned throughout the overall archive: actually, the GRB sample of unscanned triggers is by no means different from any other random subset, apart from occurring in the nearby of data gaps (but this property looks independent from the on-board trigger mechanism); therefore, a parameter called ``off-line quest coverage factor'', $f_{\mbox{\small {cov}}}$, can be defined, expressing such fraction:

$$f_{\mbox{\small {cov}}} \ = \ \frac{\mbox{\char93 scanned 1 s bi... ...ar93 scanned on-board triggers}}{\mbox{total number of on-board triggers}} \ =$$

$$= \ \frac{15,857 - 693}{15,857} \ = \ (95.6 \pm 0.2)\%$$ (37)

To estimate the number $N_{\mbox{\small {missed}}}$ of the on-board triggers, corresponding to true GRBs, that could not have been scanned by the off-line quest, the total number of GRBs automatically detected on the archive data ($N=894$, see table ) is required, and the result is expressed by the following expression :

$$N_{\mbox{\small {missed}}} \ = \ \frac{N}{f_{\mbox{\small {cov}}}... ...{f_{\mbox{\small {cov}}}} - 1.0\Big ) \ \simeq \ 894 \times 0.046 \ \simeq \ 41$$ (38)

From eq.  it comes out that the measured number of bursts automatically detected by the off-line quest has to be increased by its ($4.6 \pm 0.2$%) in order to account for the missed bursts (in this case: $\sim$ 40) occurred in the nearby of the ratemeters gaps, where the off-line quest, based on the late SWTCs, cannot work.

On the other side, since one the advantages of the early SWTCs is that they can be applied also in these ``forbidden'' regions, the 10 GRBs recovered thanks to them (table ) come from this $\sim$ 40 missed GRBs. Furthermore, the low efficiency of the early SWTCs with respect to the late ones is apparent, as well: in fact, from the rate of GRBs detected with the late SWTCs, $\sim$ 40 GRBs are expected to be scanned by the only early SWTCs, while the formers are able to catch only a quarter. In conclusion, this clearly shows that the early SWTCs sensitivity is significantly worse than for the late SWTCs, in agreement with the belief that the formers have a better efficiency.