Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 332, 875-876 (1998)

Previous Section Next Section Title Page Table of Contents

3. Comments and conclusion

When performed for the same samples and in the same conditions, the calibrations (2.1.1, 2.3.1 & 2.3.3) give results remarkably close to those by Fernley et al. (1997) and definitely within the error bars.

This contradicts Hawley et al. (1986)'s assertion that the RJH algorithm would be `afflicted with considerable statistical bias'. They actually put this on Clube & Dawe (1980) whose comment was not quite as strong as that and whose opinion was rectified after a comparative study of performance (Jones et al. 1980) as recalled in Sect. 1.

It is not clear either whether the imprecisions given by Hawley et al. (1986)'s method are imprecisions on the numerical values or dispersions ([FORMULA]) of the parent population, two distinct results with the RJH algorithm which provides also the mean solar motion and the corresponding velocity ellipsoid for each calibration (not discussed here since not provided by Fernley et al. 1997).

The tests automatically carried out by the RJH algorithm on the individual space velocities also easily detect foreign members to a parent statistical population (well examplified in Calibration 2.3.2).

The validity of the earlier RJH approach based on the principle of maximum likelihood is confirmed by the convergence of results with a more recent algorithm.

The critical importance of good data at hand is stressed and, in this respect, the study by Fernley et al. (1997) based on a renewed sample of RR Lyrae stars involving Hipparcos proper motions and individual reddening corrections is definitely an important contribution.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: March 30, 1998