Probing recent star formation with absorption-line strengths in hierarchical models and observations

Stellar population parameters (e.g. age, metallicity, and stellar abundance ratios) derived from spectral line-strengths provide a powerful probe of galaxy properties and formation histories. We implement the machinery for extracting line-strengths and ‘single-stellar-populationequivalent’ (SSP-equivalent) stellar population parameters from synthetic spectra generated by a hierarchical galaxy formation model. Our goals are (1) to test the consistency of these line-strength-derived stellar population parameters with more physically relevant lightand mass-weighted parameters for complex, cosmologically-motivated star-formation histories, (2) to interpret line-strength observations for early-type galaxies within the context of hierarchical structure formation, and (3) to test the galaxy formation models using stellar population parameters derived from some of the best available samples of observed line-strengths. We find that the SSP-equivalent age is related to the light-weighted age in a complicated fashion that reflects the influence of recently-formed stars and is poorly correlated with the mass-weighted age. We find that the tendency for SSP-equivalent ages to be biased young means that ‘archaeological downsizing’ overstates the ‘true’, mass-weighted downsizing in age with mass. We find however that the SSP-equivalent metallicity closely tracks the massand light-weighted metallicities, so that observed mass–metallicity relations for old galaxies closely reflect the underlying trends. We construct mock catalogues of early-type galaxies in a Coma cluster-sized halo and compare them directly to observations of early-type galaxies in the Coma cluster. The similarity of the SSP-equivalent ages in the observational samples and the mock catalogues gives us confidence that the star-formation quenching implemented in the hierarchical galaxy formation model produces roughly the correct amount of recent star formation. Unfortunately, the current observational samples are either too small or have too low signal-to-noise to accurately determine detailed star-formation histories. However, the data show that the model has deficiencies: the SSP-equivalent metallicities are too low and have the wrong slope as a function of velocity dispersion, and the SSP-equivalent ages of the model galaxies may have an incorrect slope as a function of velocity dispersion. These problems are indicative both of the simplified chemical evolution prescription currently implemented in the galaxy formation model and that the star-formation histories resulting from the model are incorrect in detail.