2004A&A...413..693M

We study mass-loss and wind momentum rates of 29 Galactic O-type stars with luminosity classes I, III and V by means of a pure Hα profile analysis and investigate to what extent the results compare to those originating from a state-of-the-art, complete spectral analysis. Our investigation relies on the approximate method developed by Puls et al. (1996A&A...305..171P) which we have modified to account for the effects of line-blanketing. Effective temperatures and gravities needed to obtain quantitative results from such a simplified approach have been derived by means of calibrations based on most recent spectroscopic NLTE analyses and models of Galactic stars by Repolust, Puls and Herrero (RPH) (2003A&A accepted) and Martins et al. (2002A&A...382..999M). Comparing (I) the derived wind-densities to those determined by RPH for eleven stars in common and (II) the Wind-momentum Luminosity Relationship (WLR) for our sample stars to those derived by other investigations, we conclude that our approximate approach is actually able to provide consistent results. Additionally, we studied the consequences of ``fine tuning'' some of the direct and indirect parameters entering the WLR, especially by accounting for different possible values of stellar reddening and distances. Combining our data set with the corresponding data provided by Herrero et al. (2002A&A...396..949H) and RPH we finally study the WLR for the largest sample of Galactic O-type stars gathered so far, including an elaborate error treatment. The established disagreement between the theoretical predictions and the ``observed'' WLRs being a function of luminosity class is suggested to be a result of wind clumping. Different strategies to check this hypothesis are discussed, particularly by comparing the Hα mass-loss rates with the ones derived from radio observations.