Swift Subthreshold Triggers


The GCN system has been modified to incorporate the distribution
of the Swift-BAT Subthreshold Notices.  The testing period is over
and now the Swift-BAT team is ready to release this new data product
to the public.

About once per day, the BAT flight software will trigger on
a rate increase that produces a peak in the image domain
that has a significance below that of normal image-domain threshold (=6.5 sigma),
hence the name Subthreshold Notices (image sigma is in the 5.7-6.5 range).

The goal is to go after fainter GRBs, some of which might be more distant
than the typical normal trigger GRB.  The trade-off is that the percentage
of false positives will greatly increase.  For those Subthreshold trigger locations
that the spacecraft can slew to (~50% promptly), XRT will be used to confirm
GRB-ness if it detects and point source.  (These confirmed bursts
will be "promoted" to the normal BAT_Position Notices/etc and distributed
to the normal set of Swift notice recipients.)

If, after reviewing the information below, you want to receive
these Subthreshold Notices, please send me a request to have them enabled.
Forwarding this email to  scott@milkyway.gsfc.nasa.gov is useful
because then I have your exact "site name".
Visit the Configuration Change webpage:   http://gcn.gsfc.nasa.gov/config_builder.html

Since the vast majority of these Subthreshold triggers will not be real
(i.e. just noise in the image domain), it is envisioned that these will
only be useful to robotic telescopes (i.e. no immediate consumption of human
resources).  Of course anyone is free to receive this new notice type.

The sequence for the on-board software and the ground software
for these Subthreshold Notices is as follows:
1) BAT produces a Subthreshold trigger.  (Please note that in this process
   of lowering the trigger threshold, proceedures have been implemented
   to reduce triggers around the SAA.)
2) The BAT Figure-of-Merit task and the spacecraft (SC) make their usual decisions
   about slewing to the new location or not.
3) If not, then the sequence terminates after BAT produces the usual set 
   of BAT_Position, FOM_Observe, SC_Slew, and Lightcurve TDRSS messages
   (and captures some event-by-event data).
4) If the S/C does slew, then the XRT makes its usual set of observations
   and produces the usual set of real-time Notices.  These observations terminate
   at the point in the orbit when the location becomes unobservable
   for the first time (1-40 min).  (Longer follow-up observations
   are scheduled manually if a GRB is confirmed.)
5) If you have BAT_Subthreshold enabled, then you will receive these
   BAT and XRT follow-on Notices  if you normally have them enabled.
   (UVOT does take data for these Subthreshold triggers, but it is not
   the standard Automated-Target response sequence.  This is because
   of filter wheel rotation lifetime concerns.  The UVOT exposures
   are longer and produce no real-time notices.)
   If you do not have BAT_Subthreshold enabled, then you will receive none
   of these Notices that are derived from the original BAT Subthreshold trigger.
6) These notices also go to an automated processing operation (Phil Evans, U.Leicester)
   which looks for an XRT point source within the BAT position error circle.
   (The notices used are the XRT_Position, _Image, _Threshpix, and _SPER.
   The later comes down for a period of time (1-~40 min, see item 4).)
7) If it finds a point source (presumably the afterglow of the GRB),
   then it sends that result back to GCN.  This can take up to 2-45 min
   as more and more of the SPER data are integrated.
8) GCN then "promotes" all the stored copies of the various notices to "real GRB" status
   and distributes them to all the sites that have those Notices Types enabled
   (and meets all their other filter criteria).  (If the trigger is not promoted
   there is no corresponding Nack distributed -- just silence.)

The time from the trigger to the Subthreshold notice being distributed
on the ground is in the 10-20 sec range (the same for normal triggers).

The location uncertainties are in the 2-4 arcmin range (90% CL);
slightly worse than for normal triggers.

It should be noted that these Subthreshold Notices will contain
a very large fraction of false positives.  During the testing phase,
the rate has been 3 real astrophysical sources (2 GRBs and 1 hard x-ray transient)
in about 75 subthreshold triggers (i.e. 96% false positives).

When XRT makes follow-up observations of the BAT Subthreshold position,
ther is a chance that a random, unassociated field source will be
in the BAT error circle.
The probabilities (based on 2XMM) of finding a serendipitous X-ray source in
in the BAT field of view in N seconds of SPER data:

SPER exposure time:    P(seren)
  [sec]                 [%]

0--1                    <0.004
1--10                   0.004--0.09 
10--100                 0.09--0.5 
100--500                0.5--2 
500--1000               2--6 
1000--1500              6--9 
1500--2000              9--13 
>2000s                  >13

Can be found at:  http://gcn.gsfc.nasa.gov/gcn/swift.html
The Notices are archived at:

Here is an example of a full-format email BAT_Subthreshold Notice.
It has all the same fields as a normal (above-threshold) BAT_Position
Notice -- only the NOTICE_TYPE field has changed, and there is an extra
COMMENT line indicating the low image signifiance nature of these triggers.

      TITLE:           GCN/SWIFT NOTICE
      NOTICE_DATE:     Tue 25 Aug 09 16:59:46 UT
      NOTICE_TYPE:     Swift-BAT Sub-Threshold Position
      TRIGGER_NUM:     360857,   Seg_Num: 0
      GRB_RA:          257.802d {+17h 11m 12s} (J2000),
                       257.897d {+17h 11m 35s} (current),
                       257.306d {+17h 09m 14s} (1950)
      GRB_DEC:         +28.045d {+28d 02' 41"} (J2000),
                       +28.033d {+28d 02' 01"} (current),
                       +28.105d {+28d 06' 17"} (1950)
      GRB_ERROR:       4.00 [arcmin radius, statistical only]
      GRB_INTEN:       7654 [cnts]    Image_Peak=193 [image_cnts]
      TRIGGER_DUR:     2.048 [sec]
      TRIGGER_INDEX:   264     E_range: 25-100 keV
      BKG_INTEN:       83797 [cnts]
      BKG_TIME:        61122.97 SOD {16:58:42.97} UT
      BKG_DUR:         24 [sec]
      GRB_DATE:        15068 TJD;   237 DOY;   09/08/25
      GRB_TIME:        61155.99 SOD {16:59:15.99} UT
      GRB_PHI:         -156.31 [deg]
      GRB_THETA:        30.70 [deg]
      SOLN_STATUS:     0x883
      RATE_SIGNIF:     5.38 [sigma]
      IMAGE_SIGNIF:    6.07 [sigma]
      MERIT_PARAMS:     +1  +0  +0  +1  +2  -4  +0  +0 +10  +1
      SUN_POSTN:       154.58d {+10h 18m 20s}  +10.54d {+10d 32' 26"}
      SUN_DIST:         96.54 [deg]   Sun_angle= -6.9 [hr] (East of Sun)
      MOON_POSTN:      217.95d {+14h 31m 49s}  -20.31d {-20d 18' 51"}
      MOON_DIST:        61.87 [deg]
      MOON_ILLUM:      33 [%]
      GAL_COORDS:       50.16, 33.17 [deg] galactic lon,lat of the burst (or transient)
      ECL_COORDS:      252.85, 50.76 [deg] ecliptic lon,lat of the burst (or transient)
      COMMENTS:        SWIFT-BAT Sub-Threshold GRB Coordinates.
      COMMENTS:        This is a rate trigger.
      COMMENTS:        A point_source was found.
      COMMENTS:        This does not match any source in the on-board catalog.
      COMMENTS:        This does not match any source in the ground catalog.
      COMMENTS:        This is a GRB.
      COMMENTS:        The IMAGE_SIGNIF is VERY LOW, this is very likely not real.
      COMMENTS:        This trigger occurred at longitude,latitude = 74.27,-13.64 [deg].