In order to estimate the mean rate of GRBs that have been on-line detected with the on-ground automatic S/W, based on the late SWTCs, the number of days, the quest indeed worked in, has to account for these periods of inactivity.
During the above time interval, 317 automatic on-line triggers have been detected; out of this, 229 have been classified as true GRBs, while the remaining 88 are due to solar hard x-ray flares, spurious events, and, above all, to short duration burst candidates: these, since they did not trigger the on-board logic, left their counts only in the 1 s ratemeters: unfortunately, from these time histories only the brightest bursts can be surely recognized as true, while the others remain uncertain (some of them could be due to high-energy particles crossing more than one GRBM unit, or simple statistical fluctuations). On the other hand, to date nobody observed any X-ray counterparts of a short burst. This is partially due to the selected GRBM on-board trigger parameters. We would change them to increase the sensitivity in the case of short burst candidates.
If one defines the reliability of the automatic on-line quest as the ratio of the total number of true GRBs automatically identified (229) over the number of on-line triggers (317), it comes out , in agreement with a previous estimate ([Guidorzi et al., 2001b]). In other words, at least three quarters of the on-line triggers are indeed cosmic GRBs. Out of the elapsed time (605 days), the on-line quest was not operating for 182 days for the reasons explained above. Therefore, the GRB mean rate turns out to be: 229 GRBs detected in (605-182)=423 days, i.e. GRB/day (no correction applied for sky exposure). This is lower than in the off-line case (sec. ) for the reasons explained therein (essentially, the on-line quest is more conservative and, therefore, less sensitive as for faint bursts).