The availability of parallaxes from the Hipparcos mission for a large number of MK standards permits to examine two fundamental questions, namely the accuracy of the MK luminosity calibrations and the validity of the spectroscopic approach.
It is well known that up to now, 90% of all the stellar distances known were based upon the spectroscopic method, which consists in using the MK luminosity class for obtaining the absolute magnitude. The calibration of the absolute magnitudes upon luminosity classes has been made through a combination of methods; for late type stars, mostly trigonometric parallaxes were used, whereas for the earlier types, use was made of both cluster sequences and statistical methods. The result of such methods are compiled in tabulations like the classical ones by Blaauw (1963) and by Schmidt-Kaler (1982). The sore point about these calibrations is that the uncertainties involved are seldomly known. Schmidt-Kaler says "The intrinsic (cosmic) dispersion (of the absolute magnitude) varies generally around = 0.7 mag; on the main sequence it is about 0.3 mag". The availability of the Hipparcos parallaxes makes it possible to extend the trigonometric calibrations well into early type stars, leaving aside use of the less certain statistical methods and the cluster sequences.
The second point which can also be examined is the relation between luminosity classes and absolute magnitudes. The authors of the MK system have always emphasized that the luminosity class is a description of the spectrum (for instance Morgan, 1984). Calibration is thus an independent step. At some intermediate step (never stated clearly), the assumption is then made that a strict relation exists between luminosity class and absolute magnitude. In practice this translates into the fact that the luminosity class is treated as if it were a coded absolute magnitude. To our knowledge, an examination of the relation between luminosity class and absolute magnitude has never been made for early type stars, probably because of the lack of a sufficient number of stars with both known luminosity classes and absolute magnitudes.
In what follows we shall consider the two questions successively, first the calibration and second the relation between luminosity class and absolute magnitude. In the last years another problem has appeared, which tends to complicate the problem. Several authors -for instance Slettebak (1975) and Gray & Garrison (1987)- have emphasized that when stars with higher dispersion than the original one of the MK system (125 Å/mm) are used, care must be taken to set up separate standards for high and low velocity rotating stars. We shall also consider what can be said concerning this point.
© European Southern Observatory (ESO) 1998
Online publication: January 16, 1998