TITLE:   GCN CIRCULAR
NUMBER:  7921
SUBJECT: Swift Trigger 315630: Probably not a GRB
DATE:    08/07/02 13:25:50 GMT
FROM:    Scott Barthelmy at NASA/GSFC  <scott@lheamail.gsfc.nasa.gov>

S. D. Barthelmy (GSFC), J. Cummings (GSFC/UMBC), C. Markwardt (GSFC/UMD),
P. A. Evans (U Leicester), C. P. O'Brian (Leicester U), Pagani (PSU)
report on behalf of the Swift-BAT/XRT teams.

Using the data set from T-239 to T+963 sec from recent telemetry downlinks,
we report further analysis of Swift BAT trigger #315630
(Pagani, et al., GCN Circ. 7912).  The BAT ground-calculated position is
RA, Dec = 35.555, 39.818 deg which is 
   RA(J2000)  =  02h 22m 13.1s 
   Dec(J2000) = +39d 49' 03.4" 
with an uncertainty of 3.0 arcmin, (radius, sys+stat, 90% containment).
The partial coding was 67%.

This was a weak detection on-board, and manual ground analysis
improved the significance only slightly.

The mask-weighted lightcurve shows weak peaks in the middle two
energy bands (25-100 keV).  The duration is ~0.3 sec.

The best fit to the BAT data is a powerlaw with an exponential cutoff,
yielding an Epeak of 41 +- 5 keV and 15-150 keV band fluence of
7.5 +- 2.1 x 10^-9 erg/cm2.

The XRT began observing the field of the BAT trigger 315630
in Photon counting mode at 12:51:49 UT, 104 seconds after  
the BAT trigger.  The observations during the first orbit were  
immediately interrupted due to the Earth Limb constraint.  The total time
on target for that initial orbit was 17 sec.  No point source was detected.

During the second orbit beginning at T+3600 sec XRT collected 1380 sec of PC data,
no source is detected inside the BAT error circles (flight and ground).
The  3-sigma upper limit is 5.4e-3 counts/sec, which assuming an absorbed  
power law spectrum with photon index=2 and Galactic absorption  
would correspond to an observed 0.3-10 keV flux of 2e-13 erg cm^-2 s^-1.

In following orbits, no source is detected inside the BAT error circles
in 3965 sec of PC data taken from T+26.5 ksec to T+34.5 ksec.

Using past XRT afterglow data for all bursts, there is only a 20-30% chance
that no afterglow would have been detected from a real burst
at this time delay of 1.5 hours and an observation duration of 1380 sec.

Combining all these results, we conclude that this trigger is probably
not a real GRB -- however, we can not rule out its reality.

[GCN OPS NOTE(02jul08): Per author's request, P. O'Brian was added
to the author list.]