[daip] Re: CVEL accuracy ?

[Eric Greisen]

I really know very little about this - the crucial subroutine is
$APLNOT/DOPLR and its precursor comments say


C  This version of dop takes into account components of the observer's
C  motion due to the rotation of the earth, the revolution of the
C  earth-moon barycenter about the sun, and the motion of the earth's
C  centre about the earth-moon barycenter.  The perturbations of the
C  earth's orbit due to planets are neglected.  The absolute precision
C  of this version of dop is about 0.004 km/sec, but since the dominant
C  error term is slowly varying the relative error will be considerably
C  less for times up to a week or so.
C
C  The routine omits the planetary perturbations on the earth's orbit.
C  They amount to about 0.003 km/sec and are thought to be the largest
C  contribution to the error in the velocity.

The AIPS routines do not use the elaborate code of calc that takes
into account earth tides and GR and lots of other things.

As to documentation - what you see is what you get with aips.  It is
amazing to me that the astrometry in aips is as good as it is given
that no astrometrist has been willing to work with the group very
much.

Eric Greisen