This is to announce that we now have a complete solution for the distortions in the WF/PC-1, and also a (limited) mosaicking capability for both cameras. The distortion correction allows accuracies of the order of 10 mas everywhere on each CCD chip. The mosaicking still has some residual systematic effects near the borders (of order 15-20 mas), but for most applications this should not affect the astrometry too much. Moreover, the mosaicking algorithm has been determined from data obtained at only one epoch, and no guarantee exists yet that it is stable with time (each CCD belongs to a separate camera). New data will be obtained in the next few weeks to check on the stability of the solution. We are willing, however, to go ahead on a risk basis with the astrometry derived from mosaics.
There are still a few problems which do not make it straightforward, as yet, to determine the coordinates. We will therefore undertake this task on your behalf. One of the problems is related to the header, the other to the orientation of the WF/PC-1 in the focal plane. The header problem is particularly complex (it does not depend only on the time of observation), so our approach is to reprocess all Early Acquisition images with the new, corrected software.
The task of the GO will be to provide RSB (contact person: Danny Golombek) with the x,y coordinates of their objects (usually an offset star plus the target). Alternatively they can send us a hardcopy of the image with the objects marked and we will measure their positions (however, the responsibility for the identification rests with the GO). We will take care of the corrections and will determine the coordinates. It is then up to the GOs to send an update on their proposals to USB.
The derivation of the coordinates of the objects of interest follows these steps:
1. Reprocessing (to make sure that the pointing information
in the header is accurate)
2. Measure x,y of objects (done by GO or one of our SDAs on
the basis of a GO provided finding chart)
3. Correct for distortions (done by TIB. An IRAF task will
follow shortly)
4. Create a mosaic, if necessary. The way this is done is to
put all x,y's in the reference frame of WF2 (or P6) so
that we can use the same pointing information for all CCDs)
5. Convert x,y to RA,DEC. We have redetermined the WFC
orientation and found it to be off by about 0.3 degrees
from the value in the header. Therefore, for the time being,
this task will be performed by a program which accounts for
this rotation. Once this rotation has been updated in our
database, this task will be performed by IRAF routines
(e.g. xy2rd etc).
At the end of this procedure we will have rather accurate RELATIVE coordinates of the target and the selected offset star. The ABSOLUTE coordinates, however, will be affected by the errors both in the Guide Star catalog and in the exact position of WF/PC-1 in the focal plane. These errors can easily compound to 1-2 arcseconds. The ONBOARD Acquisition you perform on the offset star will remove the zero point error, and the maneuver to the target will be based on the (much more accurate) relative positions.
Great care should be exercised in computing the onboard acquisition exposure time for the offset star, and in selecting the aperture to avoid problems with the error in the absolute coordinates. Note that for fields containing star(s) in common with the digitized plates, the zero point error can be minimized by using GASP to determine an offset to be applied to the WF/PC-1 derived coordinates. This task will also be performed by the STScI and the coordinates we will return to you will be corrected for this if possible.
-Roberto Gilmozzi