On the formation of massive galaxies: A simultaneous study of number density, size and intrinsic colour evolution in GOODS

The evolution of number density, size and intrinsic colour is determined for a volume-limited sample of visually classified early-type galaxies selected from the HST/ACS images of the GOODS North and South fields (version 2). The sample comprises 457 galaxies over 320 arcmin2 with stellar masses above 3E10 Msun in the redshift range 0.4<z<1.2. Our data allow a simultaneous study of number density, intrinsic colour distribution and size. We find that the most massive systems (>3E11 Msun) do not show any appreciable change in comoving number density or size in our data. Furthermore, when including the results from 2dFGRS, we find that the number density of massive early-type galaxies is consistent with no evolution between z=1.2 and 0, i.e. over an epoch spanning more than half of the current age of the Universe. Massive galaxies show very homogeneous intrinsic colour distributions, featuring red cores with small scatter. The distribution of half-light radii – when compared to z=0 and z>1 samples – is compatible with the predictions of semi-analytic models relating size evolution to the amount of dissipation during major mergers. However, in a more speculative fashion, the observations can also be interpreted as weak or even no evolution in comoving number density and size between 0.4<z<1.2, thus pushing major mergers of the most massive galaxies towards lower redshifts.

💡 Research Summary

The authors present a comprehensive analysis of the evolution of massive early‑type galaxies (ETGs) by simultaneously examining three fundamental observables: comoving number density, half‑light radius, and intrinsic colour distribution. Using the Hubble Space Telescope Advanced Camera for Surveys (HST/ACS) imaging of the GOODS‑North and GOODS‑South fields (v2), they construct a volume‑limited sample of 457 visually classified ETGs with stellar masses M★ ≥ 3 × 10¹⁰ M⊙ in the redshift interval 0.4 < z < 1.2. The sample is selected through a combination of automated morphological parameters and expert visual inspection, ensuring a clean early‑type classification while minimizing biases associated with colour or luminosity cuts.

Stellar masses are derived from multi‑band (U, B, V, i, z) photometry fitted with Bruzual & Charlot (2003) stellar population synthesis models, and redshifts are taken from spectroscopy where available, otherwise from high‑quality photometric redshifts (Δz/(1+z) ≈ 0.03). Structural parameters are measured with GALFIT, fitting a Sérsic profile (n ≈ 4) to obtain effective radii (Re). Intrinsic colour gradients are quantified by constructing i‑z colour maps and measuring the colour difference between the central 0.5 Re and an outer annulus (1.5–2 Re).

The key results are:

1. Number density – For the most massive systems (M★ > 3 × 10¹¹ M⊙) the comoving number density shows no statistically significant evolution between z ≈ 1.2 and the present day. When combined with low‑redshift measurements from the 2dF Galaxy Redshift Survey, the data are consistent with a constant number density over more than half the age of the Universe. This supports the “early assembly” scenario in which the bulk of massive ETGs are already in place by z ≈ 1.

2. Size evolution – The mass–size relation for the high‑mass subsample (M★ > 3 × 10¹¹ M⊙) is essentially unchanged relative to the local relation; the median Re at z ≈ 1 is within ~5 % of the z = 0 value. Lower‑mass ETGs (3 × 10¹⁰ M⊙ < M★ < 3 × 10¹¹ M⊙) exhibit a modest increase in Re of ~20 % toward lower redshift. This differential behaviour matches semi‑analytic predictions in which major dry (gas‑poor) mergers dominate the growth of the most massive galaxies, limiting size growth, while less massive systems continue to experience dissipative (wet) mergers that expand their radii.

3. Intrinsic colour – Massive ETGs display very homogeneous colour profiles: red cores with shallow gradients (Δ(i‑z) ≈ 0.02 mag) and a small scatter (σ ≈ 0.03 mag). In contrast, lower‑mass ETGs have steeper gradients (Δ(i‑z) ≈ 0.08 mag) and larger scatter (σ ≈ 0.07 mag), indicating residual or recent star formation in their outskirts. The tight colour distribution of the massive sample is consistent with passive evolution after an early, rapid quenching episode.

The authors discuss two interpretative frameworks. The first aligns with standard hierarchical models: massive ETGs form early through highly dissipative mergers, then evolve mainly via dry mergers that preserve number density and only modestly increase size; lower‑mass ETGs continue to grow in size and experience intermittent star formation. The second, more speculative, posits that the observed lack of evolution in both number density and size for massive ETGs across 0.4 < z < 1.2 could imply that the bulk of major merging for these systems occurs at even lower redshifts (z < 0.4). This would shift the epoch of significant mass assembly for the most massive galaxies to the recent Universe.

The study acknowledges limitations: the GOODS fields cover a relatively small area, limiting the ability to probe environmental dependence (cluster vs field); the Sérsic index is fixed to n ≈ 4, which may not capture structural subtleties; and colour gradients derived from only two bands cannot fully disentangle age and metallicity effects without spectroscopic information.

Overall, the paper provides a robust, multi‑dimensional observational constraint on massive galaxy evolution. By demonstrating that the most massive early‑type galaxies have essentially constant comoving number density, unchanged sizes, and uniformly red cores over a substantial fraction of cosmic time, the work strongly supports a scenario in which these systems assembled early and have since evolved passively, with only minor structural modifications driven by dry mergers. The findings serve as a valuable benchmark for semi‑analytic and hydrodynamic simulations aiming to reproduce the observed mass‑size‑colour relations of massive galaxies across cosmic history.