How to use the SEDs produced by synthesis models (inside and outside the VO)?

In this contribution we investigate how to describe the results and usage of evolutionary synthesis models. In particular, we look for an explicit and quantitative description of the parameter space of synthesis models and the evaluation of their associated uncertainties and dispersion. First, we need to understand what synthesis models actually compute: we show that a synthetic stellar population with fixed physical parameters (age, metallicity, star formation history, initial mass function and size of the system) can only be described in terms of probability distributions (i.e. there is an intrinsic dispersion in any model). Second, we need to identify and characterize the coverage in the parameter space of the models (i.e. the combinations of input parameters that yield meaningful models) and the different sources of systematic errors. Third, we need a way to describe quantitatively the intrinsic dispersion, the systematic error and the parameter space coverage of the models. Up to now, the parameter space coverage and uncertainties have been described qualitatively in the models' reference papers, with potential misinterpretations by models' users. We show how Virtual Observatory developments enable a correct use of synthesis models to obtain accurate (and not simply precise) results.
Comment: Talk contribution to "Workshop on Astronomical Spectroscopy and the Virtual Observatory" ESA pub in press. 4 pages, 4 figures. (Part I on metric of fitting for population synthesis)