Transient Event Class | Number (%) | Number (%) | Number (%) | Number (%) | Total (%) |
(1 s) | (2 s) | (4 s) | (8 s) | (merged) | |
GRBs | 730 (39.9) | 87 ( 8.1) | 54 ( 5.5) | 23 ( 5.9) | 894 (21.0) |
Likely GRBs | 217 (11.9) | 164 (15.3) | 83 ( 8.5) | 10 ( 2.6) | 474 (11.1) |
Suspected GRBs | 671 (36.6) | 512 (47.8) | 346 (35.6) | 11 ( 2.8) | 1540 (36.1) |
Solar X-ray Flares | 72 ( 3.9) | 27 ( 2.5) | 38 ( 3.9) | 37 ( 9.5) | 174 ( 4.1) |
SAGA | 16 ( 0.9) | 48 ( 4.5) | 30 ( 3.1) | 153 (39.4) | 247 ( 5.8) |
Occultation Events | 3 ( 0.2) | 11 ( 1.0) | 100 (10.3) | 101 (26.0) | 215 ( 5.0) |
Suspected TGFs | 33 ( 1.8) | 25 ( 2.3) | 16 ( 1.6) | 0 ( 0.0) | 74 ( 1.7) |
SGRs | 1 ( 0.1) | 2 ( 0.2) | 1 ( 0.1) | 0 ( 0.0) | 4 ( 0.1) |
Unkown | 29 ( 1.6) | 8 ( 0.7) | 13 ( 1.3) | 1 ( 0.3) | 51 ( 1.2) |
False | 59 ( 3.2) | 187 (17.5) | 292 (30.0) | 52 (13.4) | 590 (13.8) |
Total | 1831 (100) | 1071 (100) | 973 (100) | 388 (100) | 4263 (100) |
From a quick look of the results reported in table ,
some properties of each quest stand out: first of all, the 1 s quest
has the highest fraction of GRBs (
40%),
and this is quite easy to understand:
actually, the strong GRBs, that triggered all quests, have been accounted
for only in the 1 s results, thus leaving the weaker GRBs to the other
quests; as these are more difficult to recognize, the samples of bursts
collected with the other quests are proportionally smaller.
Another interesting observation concerns the number of off-line triggered
occultation events, as a function of the rebin time quest: while at low
bin times (1 and 2 s) there is a few of them (
1%),
at higher rebin times (4 and 8 s) this fraction rises to 10% and 26%,
respectively. To some extent, the same observation is still true for
the off-line triggered SAGA events; a possible explanation of this fact
is that these background variations are strong enough to trigger the
SWTCs only when high integration times are used.
Concerning the triggered solar hard X-ray flares, one remark is needed:
the number of flares reported in table is little
higher than the real number of distinct flares automatically caught
(see chapter 6):
actually, it sometimes happened that the longest ones, with several
hundreds seconds durations, have triggered the SWTCs several times,
not only once: only after a visual inspection of the overall sample,
it has been possible to limit this redundancy and to count the number
of separate solar flares exactly.
In summary, out of 4263 total off-line triggers automatically collected
by the samples (see the columns of table at
the right side), 894 (21%) have been classified as true GRBs.
Among the remaining subsets, the Suspected GRBs class is the largest
(36%); then, it comes out that 14% out of the total correspond to
False Events, while a similar fraction (11%) describes the population
of Likely GRBs.
If the ratio between the number of off-line triggers that have been
classified as GRBs over the total number of off-line triggers is used
to define a sort of ``off-line quest reliability''
, then it
turns out that from
, for the only 1 s quest, the
overall
falls to
.
This shows that the merging of multiple rebin quests
has a worse reliability than the 1 s quest only (that is why in the
on-line case this is the only quest working); on the other hand,
the multiple rebin search allows to catch a greater number of faint,
slow-varying, shallow GRBs,
thus improving the search algorithms' sensitivity.