1999MNRAS.303..641A

We analyse the spatially resolved colours of distant galaxies of known redshift in the Hubble Deep Field, using a new technique based on matching resolved four-band colour data to the predictions of evolutionary synthesis models. Given some simplifying assumptions, we demonstrate how our technique is capable of probing the evolutionary history of high-redshift systems, noting the specific advantage of observing galaxies at an epoch closer to the time of their formation. We quantify the relative age, dispersion in age, on-going star formation rate and star formation history of distinct components. We explicitly test for the presence of dust and quantify its effect on our conclusions. To demonstrate the potential of the method, we study the spirals and ellipticals in the near-complete sample of 32 I₈₁₄<21.9 mag galaxies with z∼0.5 studied by Bouwens, Broadhurst & Silk. The dispersion of the internal colours of a sample of 0.4<z<1 early-type field galaxies in the HDF indicates that ∼ 40 per cent (4/11) show evidence of star formation which must have occurred within the past third of their ages at the epoch of observation. This result contrasts with that derived for HST-selected ellipticals in distant rich clusters, and is largely independent of assumptions with regard to metallicity. For a sample of well-defined spirals, we similarly exploit the dispersion in colour to analyse the relative histories of bulge and disc stars, in order to resolve the current controversy regarding the ages of galactic bulges. Dust and metallicity gradients are ruled out as major contributors to the colour dispersions that we observe in these systems. The median ages of bulge stars are found to be significantly higher than those in galactic discs, and they exhibit markedly different star formation histories. This result is inconsistent with a secular growth of bulges from disc instabilities, but is consistent with gradual disc formation by accretion of gas on to bulges, as predicted by hierarchical theories. We extend our technique in order to discuss the star formation history of the entire Bouwens et al. sample in the context of earlier studies concerned with global star formation histories. Finally, we consider how to extend our method using near-infrared data and to deeper samples.