Quantifying correlations between galaxy emission lines and stellar continua using a PCA-based technique

Abstract

We analyse the correlations between continuum properties and emission line equivalent widths of star-forming and narrow-line active galaxies from SDSS. We show that emission line strengths can be predicted reasonably well from PCA coefficients of the stellar continuum using local multiple linear regression. Since upcoming sky surveys will make broadband observations only, theoretical modelling of spectra will be essential to estimate physical properties of galaxies. Combined with stellar population synthesis models, our technique will help generate more accurate model spectra and mock catalogues of galaxies to be used to fit data from new surveys. We also show that, by combining PCA coefficients from the pure continuum and the emission lines, a plausible distinction can be made between weak AGNs and quiescent star-forming galaxies. Our method uses a support vector machine, and allows a more refined separation of active and star-forming galaxies than the empirical curve of Kauffmann et al. (2003).