Red Galaxies at Night, Astronomers’ Delight!

Title: The Dawn of the Red: Star formation histories of group galaxies over the past 5 billion years 

Authors: Sean L. McGee, Michael L. Balogh, David J. Wilman, Richard G. Bower, John S. Mulchaey, Laura C. Parker, and Augustus Oemler

First Author’s Institution: University of Waterloo, Ontario, Canada

Status: Accepted to MNRAS, open access on arXiv

Galaxies are a true wonder of the universe. Unimaginably vast, they can contain up hundreds of billions of stars. The location of a galaxy is an important factor in its overall evolution, as this process can be influenced by its surroundings. A key quantity that can measure the effect of a galaxy’s environment is the star formation rate (SFR).  Among other things, the SFR gives an insight into how active the galaxy is. Curiously, the overall star formation rate of galaxies in the universe has decreased over time, with peak star formation having already occurred in the early universe. Even more perplexing is how this general reduction has been shown to apply to galaxies across (almost) all stellar masses. Today’s work is tasked with determining whether this general reduction applies across all environments.

Identifying Individual Galaxies with Friends

In order to study the large scale influence of environmental effects on galaxy SFRs, it is crucial to obtain survey samples over large distances and with as wide a field of view as possible. This work used the results of two surveys: the high-redshift Group Environment and Evolution Collaboration (GEEC) (redshift referring to how far away a galaxy is, the higher the redshift, the more distant the object), and the low-redshift Sloan Digital Sky Survey (SDSS).

The SFRs were obtained by fitting spectral energy distributions (SEDs) to each sample, which mapped the energy emitted by the object over different wavelengths. The GEEC survey obtained spectroscopic redshifts for 6000 galaxies over a total sky area of 1.5 square degrees (an area on the sky smaller than that of an outstretched thumb) with around 200 unique galaxy groups identified using a friends-of-friends algorithm (a type of percolation algorithm that processes separations between pairs of samples). This work also used the results of Data Release 6 of the SDSS survey (over 790,000 samples across about 1/6 of the sky). For this second set of samples, group finding algorithms were again used for this second set of samples in order to separate them into group galaxies (those in clusters or otherwise close together) or field galaxies (those that exist in isolation).

The SED fitting was achieved by first using a set of template SEDs from galaxies with known parameters, then fitting this template to the photometry (properties of the light) of the observed samples. The goal is to obtain the galaxy parameters that best fit the observed SED given the template SED. In order to create a model template SED, one must first have a model of the stellar population. Given some initial mass function (a function that describes the initial distribution of stellar masses), and a model that can compute the spectral properties of that population over time, it is possible to estimate fluxes for each photometric band. The SFR can thus be derived from the fitted SED. To compare results, this work used the specific SFR (SSFR) – the ratio of the SFR divided by the stellar mass – to allow samples with different masses to be more easily compared.

Active and Passive Galaxies

Figure 1 shows the results from the SDSS samples. The dotted red line marks the division between active “blue” star forming galaxies and passive “red” non-star forming galaxies. It is well known that there is a strong correlation between stellar mass and SFR, but the SSFR is even more useful in that it encodes the star formation history of the galaxy. As a rule of thumb, the SSFR can be converted into a birth rate b by multiplying the SSFR by the age of the Universe (in years). If this birth rate is greater than 1, then the galaxy is forming stars more quickly than it has in the past (and vice versa). The set of galaxy samples at around log10 SSFR = −10 in Figure 1 appears to have a well defined slope. This area has been dubbed the main-sequence of star-forming galaxies, because the birth rates of these galaxies are around b ~ 1, suggesting that they have been forming stars at this rate for most of their lifetime.  The clumps of samples below the red line correspond to quiescent galaxies; those that are no longer forming stars. Interestingly, the shape of the plot for the field and group galaxies is near identical. What’s more, this shape is also reflected in the GEEC samples (Figure 2).

Figure 1: (Figure 5 in the paper) SSFRs for the group (left) and field (right) galaxies of the SDSS sample as functions of stellar mass. The red line divides these samples into active (above) and passive (below) galaxies.
Figure 2: (Figure 6 in the paper) SSFRs for the group (left) and field (right) galaxies of the GEEC sample as functions of stellar mass.

The main result of this work lies in Figure 3, where the mean SSFRs of star forming galaxies (corresponding to the main sequences in Figures 1 and 2) are displayed for each survey and environmental group. Immediately it is clear that the high redshift GEEC samples have consistently higher mean SSFRs than the low redshift SDSS samples.  Thus there is a reduction in SFR over all stellar masses as well as across all environments (Figure 3).

Figure 3: (Figure 7 in the paper) The mean SSFR of star-forming galaxies in the group and field of both the SDSS and GEEC surveys

Red Galaxies at Morning…

One could argue that this is the sort of paper that yields more questions than answers. Although it has been shown that the decline in SFR is independent of environment, the fraction of star-forming galaxies still has a strong environmental dependence (see Figure 4). This demonstrates how environments still play an important role in the evolution of galaxies.

This work has shown that that star-forming galaxies in all environments show a systematic, mass-independent lowering of SFR from z = 0.4 (around 4 billion years ago) to z = 0.08 (about 1 billion years ago), Curiously, they also found that the fraction of passive galaxies (those that have stopped forming stars) is higher in group galaxies than in field galaxies.  Accretion models have been proposed in an attempt to explain why this is the case. The most promising candidate is a process known as strangulation, where hot gas surrounding a galaxy is stripped when the galaxy becomes a satellite of a massive dark matter halo. Despite this, many models produce more so-called “green” galaxies (i.e those in transition from active to passive) than have currently been observed. 

There is still more work to be done in order to reconcile these differences. Until then, the secrets behind why groups have higher passive fractions, and why the overall SFR of galaxies have been dropping over time, are still yet to be discovered.

About Mitchell Cavanagh

Mitchell is a PhD student in astrophysics at the University of Western Australia. His research is focused on the applications of machine learning to the study of galaxy formation and evolution. Outside of research, he is an avid bookworm and enjoys gaming, languages and code jams.

Leave a Reply