WYFFOS/AUTOFIB2 on the WHT

Astrometry

What sort of accuracy do I need?

Currently the postioning/guiding error of Autofib2 amounts to approximately 0.4 arcsec; see the Autofib2 general description page. Thus, errors in the positions supplied should be no greater than 0.4 arcsec. Great care should also be taken to ensure that the fiducial (giude) stars are on the same astrometric reference frame as the program stars. For the present, we recommend that observers use positions based on APM scans of photographic sky survey plates (POSS or UKST) which can be accessed via the CATALOGUES program at Cambridge. If working with objects from CCD frames, you must ensure that you set up the secondary astrometric reference frame based on the APM positions of the brighter (unsaturated) stars in the CCD field. Recent work with COSMOS suggests that positions obtained with this measuring machine may not be sufficiently accurate over the large field available with WYFFOS/AUTOFIB2.

What sort of fiducials do I need?

There are quite severe limitations imposed in the placing of fiducial guide fibres within the WYFFOS/AF2 field. A large number (>>20) of candidate fiducial stars must be input into the CONFIGURE program for each WYFFOS/AF2 field, in order to ensure a suffient yield (>3) of fiducial stars with fibres. In principle aquisition and guiding can be achieved off fiducial stars as faint as B=16-17. However, under these circumstances, the integration times on the TV are long, and accurate guiding is difficult. We recommend fiducial stars in the magnitude range B=14-15; this will maintain a narrow magnitude range (and thus keeping within the dynamic range of the TV), while avoiding the bright stars (B<13) which are likely to have measured positions systematically shifted with respect to the fainter stars on the photographic plate due to strong diffraction spikes and haloes.

How often do I need to re-configure the field?

Due to atmospheric refraction the positions of objects in the field will slowly shift as a function of zenith distance (z). For the 1 degree field of WYFFOS/AF2 this is not too serious a problem for fields reasonably close to the zenith. Using the approximate formula (suitable for z < 60 degrees):

delta(z) ~ 60.4arcsec tan (z)

we find that the shift over the field amounts to +/-0.35 arcsec for observations with zd < 40 degrees. Thus a field at dec=+30 degrees could, in principle, be tracked for 6 hours (-3h < HA < +3h) without the need for reconfiguration. However, this will NOT be true of fields at other declinations, and observers should estimate the likely shifts for fields observed at zd > 40 degrees using the formula above (re-configuring when necessary). This formula breaks down at z > 60 degrees (why are you observing that far over anyway??) where

delta(z) ~ 60.29arcsec tan (z) - 0.06688arcsec tan^3(z)

is more appropriate.

Brian Boyle 29 February 1996