Improving stellar parameter and abundance determinations of early B-type stars
In the past years we have made great efforts to reduce the statistical and systematic uncertainties in stellar parameter and chemical abundance determinations of early B-type stars. Both the construction of robust model atoms for non-LTE line-formation calculations and a novel self-consistent spectral analysis methodology were decisive to achieve results of unprecedented precision. They were extensively tested and applied to high-quality spectra of stars from OB associations and the field in the solar neighbourhood, covering a broad parameter range. Initially, most lines of hydrogen, helium and carbon in the optical/near-IR spectral range were reproduced simultaneously in a consistent way for the first time, improving drastically on the accuracy of results in published work.By taking additional ionization equilibria of oxygen, neon, silicon and iron into account, uncertainties as low as ~1% in effective temperature, ~10% in surface gravity and ~20% in elemental abundances are achieved - compared to ~5-10%, ~25% and a factor ~2-3 using standard methods. Several sources of systematic errors have been identified when comparing our methods for early B-type stars with standard techniques used in the nineties and also recently (e.g. VLT-FLAMES survey of massive stars). Improvements in automatic analyses are strongly recommended for meaningful comparisons of spectroscopic stellar parameters and chemical abundances ('observational constrains') with predictions of stellar and galactochemical evolution models.