Error analysis for stellar population synthesis as an inverse problem

Abstract

Stellar population synthesis can be approached as an inverse problem. The physical information is extracted from the observations through an inverse model. The process requires the transformation of the observational errors into model errors. A description is given for the error analysis to obtain objectively the errors in the model. Finding a solution for overdetermined and underdetermined cases was the purpose of two preceding papers. This new one completes the problem of stellar population synthesis by means of a data base, by providing practical formulae defining the set of acceptable solutions. All solutions within this set are compatible, at a given confidence level, with the observations.