Astrophysics Research Group, University of Antwerp (RUCA),
Groenenborgerlaan 171, B-2020 Antwerpen, Belgium
Measuring accurate radial velocities of rotating early-type (O-B-A) stars
using cross-correlation
techniques is hampered by the fact that object-template spectrum mismatch
causes systematic errors that do not cancel out sufficiently in these spectra.
This series of papers aims at quantifying those mismatch errors, understanding
their astrophysical origin, and developing strategies to avoid them maximally.
In this first paper, we employ synthetic spectra to
study mismatch between A-type main-sequence stars caused solely by
differences in
and log
g.
We show that this spectral-type mismatch varies greatly throughout the
spectrum, though with some degree of systematic dependence on
rotational velocity and, to a lesser degree, on temperature.
We propose a scheme for selecting spectral regions that should provide,
for main-sequence A-type stars, accuracies better than 1 km s
-1 for
vsin
i
150 kms
-1 and between 1 - 2 kms
-1for
vsin
i up to 300 kms
-1.
The scheme includes sufficient spectral information to keep
random errors conveniently small, but excludes all wavelength sub-intervals
which produce systematic errors much larger than the above mentioned
accuracy.
Our predictions confirm the success of the methodology of Fekel
(1985, 1999).
We conclude that the proposed scheme needs
further testing on a broad sample of real
A-type spectra to see under what conditions of stellar individuality
it may break down.
Key words: methods: numerical --
techniques: radialvelocities -- techniques: spectroscopic --
stars: early-type -- stars: kinematics