S.D. Barthelmy (GSFC), D.M. Palmer (LANL) and Neil Gehrels (GSFC)
for the Swift Team:
Swift has started a new campaign structured to detect fainter transients
in nearby galaxies -- in particular subluminous long GRBs and SGR Superflares.
(This campaign started ~13:00 UT 14 may 2012.)
Swift/BAT lowered the triggering threshold for peaks in the image domain
that are near locations of known sources in the on-board catalog.
It is called the "interesting source" threshold. We are lowering it
from 6.40 sigma (where it has been for years) to 5.80 sigma.
To satisfy the interesting source threshold, a peak in the image domain
must be greater than 5.80 sigma AND it must be within a 12 arcmin radius
of one of these sources in the catalog AND the peak intensity minus
the current on-board threshold intensity for that source must be
greater than 2 sigma. We have about 600 nearby galaxies (distance < 20 Mpc)
in the on-board catalog, ~70 tiles covering M31, and ~600 other known
hard x-ray sources. For the 670 galaxy entries (nearby and M31),
their thresholds have been set to 0. For nearly all of the 600
known sources, their thresholds are twice their previous trigger level.
It is the "(peak_intensity - threshold_intensity) > 2 sigma"
that inhibits this new Interesting Source trigger from firing
on the many known sources. We are going after the underluminous
long GRBs and SGRs in the nearby galaxies.
The following bullets describe key aspects about the information
you will receive for these Interesting Source triggers:
* The full, normal set of Swift GCN Notices will be produced for these triggers.
And these Notices will have their full set of information content;
there is no format or content change for these triggers.
These triggers (in the 5.8 - 6.4 sigma range) will still cause BAT to trigger
and produce the normal set of messages down TDRSS, through GCN, and out to
the world, ie BAT_POS, BAT_LC, BAT_SCALEDMAP, FOM, SC_SLEW, XRT_POS_[NACK], etc.
* If the spacecraft decides it is safe to slew (just like normal),
then the spacecraft will slew and the XRT & UVOT will produce their normal set
of observations and GCN notices (including updates). (On rare occaisions,
the merit value of the new interesting source will be below the merit value
of the target currently being observed, and so no slew will occur until
a planned target of lower merit value comes along AND the interesting source
is still visible.)
* Since, by definition, these triggers will be spatially coincident with a source
in the on-board BAT catalog, there will be a COMMENT line in the full-format GCN email
notice that says "this matches source XYZ123 in the on-board catalog; delta=0.NNN [deg]".
And a COMMENT line "this matches source XYZ123 in the ground catalog; delta=0.NNN [deg]".
Because the matching tolerance (12 arcmin=0.20 deg) is much larger than the typical BAT
position error (3 arcmin=0.05 deg), most spurious events will have a delta > 0.07 deg.
However, a large delta can also indicate a true object in the halo of a galaxy
(35 kpc @ 10 Mpc = 0.2 deg).
* Because of these catalog matches, GCN will label the trigger as a TRANSIENT
(not a GRB). (So there will be the usual two GCN notices go out for these
catalog-matching triggers: (a) the BAT_GRB_POS notice, and (b) the BAT_TRANSIENT notice.
* Other noteworthy pieces of information for these Interesting Sources are:
(a) these Interesting Source triggers can result from both Rate Triggers (4msec - 32 sec)
and from Image Triggers (Trigger_Dur > 60 sec).
(b) the "IMAGE_SIGNIF: 5.83 [sigma]" line in the BAT_POS notice
will have a value between 5.8 & 6.4.
This will also cause a COMMENT line to appear to the effect of "low image significance...".
(c) If the "TRIGGER_DUR: 0.032 [sec]" has a small value (probably 4 to ~128 msec),
then it is likely an SGR Superflare or a Short GRB in nearby galaxy or noise.
If it is a big value, then it is either a long GRB or a hard x-ray transient
in a nearby galaxy, or noise.
We currently estimate that the false positive rate will be ~1 per week.
If this turns out to be significantly off, we will adjust the 5.80 value
higher or lower accordingly.
The rate of reals is somewhat harder to calculate, but is expected
to be in the 0.5-1 per month range.
The Swift Team will issue its rapid-response Circular only if the trigger is
a real astrophysical source, but not for noise-caused triggers (this is because
the Team is getting smaller and this would increase the work-load).
These rapid-response Circulars are typically issued within 30 min of the trigger time.