Stellar synthetic spectroscopy in the Virtual Observatory era

Abstract

Arguments for including in the VO large grids of synthetic spectra fully covering the HR diagram are presented. One obvious need is that of population synthesis at all redshifts. Theoretical spectra have also the power to predict peculiar behaviors, that could remain unnoticed in observed spectra, or lead to erroneous conclusions. One interesting example is given of the Ca II H & K lines in extremely metal-poor stars. In carbon-rich atmospheres, the H & K lines become much weaker relative to the continuum, which will lead to an underestimate of the metallicity. What actually happens is a displacement of the continuum, obvious in absolute-flux synthetic spectra, but not visible in observed continuum-normalized spectra. Libraries of synthetic spectra are described, as well as the codes used to compute them. A few remarks are made on issues with making these libraries available through the VO, as well as on the necessary input data, i.e. line lists and model atmospheres.