GCN Circular 5628

D. N. Burrows (PSU), J. E. Hill (GSFC/USRA), L. Angelini (NASA-GSFC), A. 
Moretti (INAF-OAB), A. P. Beardmore (U. Leicester), and J. A. Nousek (PSU) 
report on behalf of the Swift/XRT team:

We wish to alert the community to an apparent change in the XRT boresight 
that affects XRT source positions.  Comparison between XRT and optical 
positions for GRBs with optical counterparts since January 2006 shows that 
the XRT position errors have increased since the beginning of May 
2006.  Analysis of bursts since 1 May 2006 shows that nearly 1/2 fall 
outside the quoted XRT 90% confidence radius (in most cases, only slightly 
outside the error circle), compared with ~6% for bursts before 1 May 
2006.  A plot of the ratio of the position error to the error circle radius 
shows a jump in early May.  This jump is presumed to be caused by a small 
shift in the XRT boresight.  The cause of the shift is unknown and is under 
investigation by the XRT team.  The XRT team is also in the process of 
recalibrating the boresight to correct for this shift.  In the meantime, we 
are increasing the radii of our error circles by 50% to account for the 
systematic error introduced by this boresight change.  Once we have 
recalibrated the XRT boresight, we will reduce our error circle radii back 
to an appropriate level.