Forensic reconstruction of galaxy colour evolution and population characterisation

TL;DR

This paper reconstructs galaxy colour evolution over cosmic time using observations and simulations, revealing that galaxy populations are best described by two main groups—blue and red—with little evidence for a third, green group.

Contribution

It introduces a method combining stellar templates and galaxy histories to characterize galaxy colour evolution and population dynamics over time.

Findings

Galaxy populations are best modeled with two groups: blue and red.

Massive galaxies become significantly redder over time.

Simulations and observations show good agreement, with some models predicting overly blue massive galaxies.

Abstract

Mapping the evolution of galaxy colours, from blue star-forming to red passive systems, is fundamental to understand the processes involved in galaxy evolution. To this end, we reconstruct the colour evolution of low-redshift galaxies, combining stellar templates with star formation and metallicity histories of galaxies from the Galaxy And Mass Assembly survey and \shark\ semi-analytic model. We use these colour histories to robustly characterise the evolution of red and blue galaxy populations over cosmic time. Using a Gaussian Mixture Model to characterise the colour distribution at any given epoch and stellar mass, we find both observations and simulations strongly favour a model with only two populations (blue and red), with no evidence for a third "green" population. We map the evolution of mean, weight, and scatter of the blue and red populations as a function of both stellar mass and lookback time. Using our simulated galaxy catalogue as a testbed, we find that we can accurately recover galaxies colour histories up to a lookback time of $\sim6$ Gyr. We find that both populations show little change in the mean colour for low-mass galaxies, while the colours at the massive end become significantly redder with time. The stellar mass above which the galaxy population is predominantly red decreases by 0.3 dex in the last 5 Gyrs. We find a good agreement between observations and simulations, with the largest tension being that massive galaxies from \shark\ are too blue (a known issue with many galaxy evolution models).