Date: Mon, 03 Sep 2007 11:33:21 -0500 (CDT)
From: Stephen Kent
Subject: Re: 2.5m obslog 20070902 (MJD 54346)
> Skippy was never happy (see problems, HJD54345)
I checked some fields from run 6920.  We are scanning 82 N as requested,
but the right ascension in the header does not match the actual pointing,
being too high by about 0.18 degrees.
The u' CCD straddles the camera boresight in RA. The RA in the fits header
corresponds to the location of the telescope boresight at the time when
row 0 is read out.  In a u' frame, this RA should match the actual sky
position at row 1024.
I checked this by finding images of the globular cluster M2, which crosses
camcol 1 at ra=323.36.
In run 1620, I find the cluster in frame u1-0236 around row 205.  The FITS
header RA is 323.626.  Based on the header, I would predict an RA of
323.626 - (1024-205)*.4/3600 = 323.54.  This is 0.18 degrees (roughly 1600
pixels) away from the actual position or 43 seconds of RA.
Does anyone know what would cause this offset?
Steve