A critical test of empirical mass loss formulae applied to individual giants and supergiants

To test our new, improved Reimers-type mass-loss relation, given by Schroder & Cuntz in 2005 (ApJL 630, L73), we take a look at the best studied galactic giants and supergiants - particularly those with spatially resolved circumstellar shells and winds, obtained directly or by means of a companion acting as a probing light source. Together with well-known physical parameters, the selected stars provide the most powerful and critical observational venues for assessing the validity of parameterized mass-loss relations for cool winds not driven by molecules or dust. In this study, star by star, we compare our previously published relation with the original Reimers relation (1975), the Lamers relation (1981), and the two relations by de Jager and his group (1988, 1990). The input data, especially the stellar masses, have been constrained using detailed stellar evolution models. We find that only the relationship by Schroder & Cuntz agrees, within the error bars, with the observed mass-loss rates for all giants and supergiants.