[vsnet-grb-info 3952] GRB 060801: Swift/XRT Astrometry Correction

[GCN Circulars]

TITLE:   GCN CIRCULAR
NUMBER:  5773
SUBJECT: GRB 060801: Swift/XRT Astrometry Correction
DATE:    06/10/30 16:30:13 GMT
FROM:    Judith Racusin at PSU  <racusin at astro.psu.edu>

J. L. Racusin, D. N. Burrows (PSU), and N. Gehrels (NASA/GSFC) report on 
behalf of the Swift XRT team:

We have re-analyzed the full XRT data set of the short hard GRB 060801. 
XRT observed the field of GRB 060801 for a total of 75 ks between August 1 
and August 5, 2006 in Photon Counting mode in order to obtain a more 
accurate position of the short-lived X-ray afterglow.

In the full data set we find 23 serendipitous X-ray sources detected with 
the XIMAGE detect algorithm with S/N > 3, 12 of which have near-by optical 
counterparts in the SDSS.  We match these sources to obtain a best fit 
mean frame shift, carefully accounting for several instrumental factors 
described below.

At the time of these observations there was a problem with the bias maps 
that caused spurious hot pixels to remain in the cleaned event lists after 
standard processing.  To remove this contamination, we filtered out events 
with energies < 0.4 keV.  For more information on the bias map issue, see 
http://www.swift.ac.uk/xrtdigest.shtml.

We also apply an exposure map correction to the images before input into 
XRTCENTROID to account for any sources landing on or near the bad columns 
in the XRT CCD. In the first orbit, the GRB did in fact lie on one of 
these columns.  We use XRTCENTROID to calculate the positions of the GRB 
and the serendipitous sources.

Additional position error arises from the Swift star tracker solution 
uncertainty.  This error is small once the spacecraft has completely 
settled after a slew, but is not achieved until ~100s after the XRT 
observations begin.  Therefore, we exclude the first 100s of each orbit, 
reducing the total used exposure used time to 69 ks.

We calculate the statistical position errors using the empirical fits as 
described in Moretti et al. (2006, A&A, 448, L9), assuming that the 
astrometric correction removes the 3.5" systematic error normally applied 
to XRT positions to account for errors in the star tracker attitude 
solution.

The result of this analysis leads to a mean frame shift from the 
previously reported position (Racusin et al., GCN 5382), of:

dRA: -3.60"   dDec: -0.90"  +/- 0.64"

and a new XRT refined position of:

RA(J2000):   14h 12m 01.31s
Dec(J2000): +16d 58' 54.0"

with an estimated uncertainty of 2.1 arcseconds (90% containment).

This position is 4.2 arcseconds from the refined XRT position given in 
Racusin et al. (GCN 5382), 1.9 arcseconds from the boresight corrected XRT 
position using the new TELDEF described by Burrows et al. (GCN 5750), and 
0.5 arcseconds from the frame shifted position given in Butler et al. (GCN 
5389) based upon only the first 25 ks of data.  Potential host galaxies B 
& D from Castro-Tirado et al. and Piranomonte et al. (GCNs 5384 & 5386) 
both lie inside the new XRT refined error circle.  A figure comparing all 
of these positions is available at: 
http://www.swift.psu.edu/images/grb060801_astrometry.gif

Galaxy B has a redshift of 1.131 as given by Cucchiara et al. (GCN 5460). 
We encourage observations of galaxy D to obtain its redshift.

This Circular is an official product of the Swift XRT team.