Age-Divided Mean Stellar Populations from Full Spectrum Fitting as the Simplified Star Formation and Chemical Evolution History of a Galaxy: Methodology and Reliability

We introduce a practical methodology for investigating the star formation and chemical evolution history of a galaxy: age-divided mean stellar populations (ADPs) from full spectrum fitting. In this method, the mass-weighted mean stellar populations and mass fractions (f_mass) of young and old stellar components in a galaxy are separately estimated, which are divided with an age cut (selected to be 10^9.5 yr ~ 3.2 Gyr in this paper). To examine the statistical reliability of ADPs, we generate 10,000 artificial galaxy spectra, each of which consists of five random simple stellar population components. Using the Penalized PiXel-Fitting (pPXF) package, we conduct full spectrum fitting to the artificial spectra with noise as a function of wavelength, imitating the real noise of Sydney-Australian Astronomical Observatory Multi-object Integral field spectrograph (SAMI) galaxies. As a result, the \Delta (= output - input) of age and metallicity appears to significantly depend on not only signal-to-noise ratio (S/N), but also luminosity fractions (f_lum) of young and old components. At given S/N and f_lum, \Delta of young components tends to be larger than \Delta of old components; e.g., \sigma(\Delta [M/H]) ~ 0.40 versus 0.23 at S/N = 30 and f_lum = 50 per cent. The age-metallicity degeneracy appears to be insignificant, but \Delta log(age/yr) shows an obvious correlation with \Delta f_mass for young stellar components (R ~ 0.6). The impact of dust attenuation and emission lines appears to be mostly insignificant. We discuss how this methodology can be applied to spectroscopic studies of the formation histories of galaxies, with a few examples of SAMI galaxies.
Comment: 26 pages, 24 figures, accepted for publication in MNRAS