Wong et al. Article -- Galaxy Zoo: Building the Low-mass End of the Red Sequence with Local Post-starburst Galaxies

Title: Galaxy Zoo: building the low-mass end of the red sequence with local post-starburst galaxies

Authors: O. I. Wong, K. Schawinski, S. Kaviraj, K. L. Masters, R. C. Nichol, C. Lintott, W. C. Keel, D. Darg, S. P. Bamford, D. Andreescu, P. Murray, M. J. Raddick, A. Szalay, D. Thomas and J. VandenBerg

First Author’s Institution: CSIRO Astronomy & Space Science, Astronomy Department, Yale University

To read the full article, please click here.

Article Summary + Additional Background Information:

Galaxies were once believed to be isolated and unevolving systems in our universe.  In the past few decades this viewpoint has drastically changed; observations suggest that galaxies are strongly affected by gravitational interactions from the other galaxies in their nearby environment, and these interactions potentially are the main drivers of galaxy evolution.  Studying collisions, tidal interactions, and their effects are exceptionally important for further probing the important processes of galaxy evolution.

Figure 1. A color-luminosity graph indicates the location of the blue cloud, green valley, and red sequence.  The blue cloud is populated by star forming, spiral galaxies (late type) while the red sequence contains non-star forming, passively evolving elliptical galaxies (early type).  Post-starburst galaxies transitioning from the blue cloud to the red sequence inhabit the green valley.  

There are two basic types of galaxies recognized by Hubble’s “tuning fork” classification scheme: blue, star-forming spiral galaxies (late type) and red, quiescent elliptical galaxies (early type).  Spiral galaxies tend to have younger stellar populations emitting higher-energy (bluer) light, while elliptical galaxies are littered with older stars emanating lower-energy (red) light.  Through galaxy collisions, star-forming spiral galaxies in the ‘blue cloud’ are believed to develop into passively evolving elliptical galaxies in the ‘red sequence’, as illustrated in Figure 1.  During this transition galaxies occupy the sparser ‘green valley’.

During galactic interactions, the probability of a star-star collision is on the order of 1 part in a quadrillion.  The dark matter halos of galaxies, which make up about 80% of the galaxy mass, do not interact other than gravitationally.  However, the interstellar gas in galaxies does interact, causing a period of abnormally high star formation called a starburst.  This influx of gas also fuels the supermassive black holes at the center of galaxies, generating an active galactic nuclei (AGN).  The process of two galaxies colliding and merging is extremely slow by terrestrial standards, occurring over hundreds of millions to billions of years.  To study these transitional galaxies, astronomers turn to the Galaxy Zoo project for information about the aftermath of galactic interactions.  This provides a way to understand the evolutionary path from the blue cloud to the red valley.

Galaxies that have recently quenched star formation are called post-quenched or post-starburst galaxies (PSGs).  A recent study of local PSGs using the photometric and spectroscopic data from the Sloan Digital Sky Survey (SDSS) in conjunction with the results from the Galaxy Zoo project led to a better understanding of this transitional period of galaxy evolution.  PSGs are also called ‘E+A’ or ‘K+A’ galaxies, because they are galaxies that have ceased current star formation but still exhibit the spectral signature of recently formed stars (i.e., stars with stellar type 'A', as shown in Figure 2).

Figure 2. Comparison of a typical post-starburst galaxy spectrum with a normal star forming galaxy and a spectrum of a star with stellar type 'A'.

Using the SDSS, one of the largest and most complete samples of local PSGs was assembled, allowing volunteers of the Galaxy Zoo citizen science project to investigate the visual properties of these galaxies.  The galaxies selected were in the nearby universe, with redshifts of 0.02 < z < 0.05, and with a z band magnitude of Mz < -19.5 mag.  The z band was chosen for selection purposes because it is the reddest waveband provided by the SDSS and provides the closest proxy to stellar mass.  This selection reduced the Malmquist bias – which is the preferential detection of intrinsically bright objects.  Since the PSGs are galaxies with recently truncated star formation that still exhibit strong Balmer absorption from young stars, they were identified as having Hα emission line weaker than four times the rms level and Hδ equivalent width wider than 3 angstroms.  Of the 47,573 galaxies in the selected volume 80 matched the criterion of PSGs.  12 of the PSGs selected are presented in Figure 3.

Figure 3. This image shows the SDSS images of 12 PSGs in the sample.  The left-hand panel shows four examples of early-type PSGs, the center panel intermediate-type PSGs, and the right panel late-type PSGs.  

About 74% of the PSGs were neither early or late type galaxies, and were therefore classified as intermediate type.  About 16% and 10% of the PSGs were classified as early and late types, respectively.  This suggests that the PSGs in the sample are an evolutionary stepping-stone from blue, star forming spiral galaxies to red, quiescent elliptical galaxies.  Quantification of merger properties from Galaxy Zoo results concluded that most of the PSG samples did not have signs of an actively merging system, though many of the samples were asymmetrical or disturbed.

Additionally, using the SDSS modelMag tool, the u – r color was determined for these galaxies (where u and r are the bluest and middle band magnitudes used by SDSS, respectively).  The magnitudes were corrected to take out the effects of absorption by methods used in Calzetti et al. (2000), which accounted for both warm dust (T~40-55 K) and cool dust (T~20-23 K).  Most of the PSGs lay in the color range 1.8 < u - r < 2.3, which is the ‘green valley’ between the ‘blue cloud’ (late type galaxies) and the ‘red sequence’ (early type galaxies).

This study also looked at the environment around the local PSGs to correlate galaxy density with evolutionary processes.  The environment of the PSG samples was determined by measuring the number and proximity of galaxies around the point in space where the samples lay.  Half of the PSGs resided in low-density environments, while 26 and 24 percent resided in medium- and high-density environments, respectively.

Since most of the PSG sample consisted of intermediate-type morphologies, further investigation of stellar structure was required to reveal if the PSGs have intermediate-type morphologies due to past interactions or are similar in structure to the early- or late-type galaxies within the same volume.  To determine this, the SDSS fracDev parameter was used.  This gives the fraction of light fitted by a de Vaucouleurs profile, which describes how the surface brightness of an elliptical galaxy varies as a function of the radius from the galactic center.  Using this parameter, the authors found that the structural stellar morphologies of the PSGs in the ‘green valley’ more closely resemble the morphologies of low-mass early-type galaxies, even though star formation has only recently been truncated.

Stellar mass estimates for the galaxies were measured by fitting the five optical wavebands used by the SDSS to star formation history libraries created from stellar models in Maraston (1998, 2005).  A majority of the PSGs in the study had stellar masses below the transition mass that separates low-mass star-forming galaxies from the high-mass passively evolving bulge-dominated galaxies.  No PSGs were found with log stellar masses greater than 11.5 solar masses (i.e., greater than 10^11.5 Msun).  One possible reason for this lack of high-mass PSGs is that the sample was restricted to a very local volume.  These results are consistent with the idea of galaxy formation ‘downsizing’, the theory that more massive galaxies from higher density areas run through their gas quicker and evolve through the PSG phase at higher redshifts than lower mass galaxies.  Figure 4 shows the color versus stellar mass for the PSG sample.

Figure 4. The image above shows the location of the sample PSGs on u-r color vs stellar mass graphs.  The panels of the top row, from left to right, show the location of all the galaxies, early-type galaxies, intermediate-type galaxies, and late-type galaxies in the study, respectively.  The bottom row of panels shows the number fraction of the PSG sample to the galaxy sample of a particular type in a given color-stellar mass bin.  

Current models of galaxy evolution suggest that feedback from AGN could provide the means to quench and truncate the star formation history of a massive galaxy.  Mergers may induce inflows of gas that fuel star formation and the central black hole, while feedback from AGNs quench star formation by reheating cold gas and expelling much of it in AGN-driven winds.  This hypothesis suggests that AGN feedback may play a role in quenching star formation in PSGs.  However, apart from two PSGs in this study that exhibit spectral properties of AGN called LINERs (low-ionization nuclear emission-line regions) no observations of AGN spectral signatures were found in the PSG sample.  These observations coincide with the idea of 'downsizing', in which the buildup of smaller galaxies occurs at later epochs.  The low-z galaxies in this sample were most likely not massive enough to host an AGN and therefore AGN feedback was not the primary quenching mechanism.  For more information on AGN feedback, refer to Schawinski et al. 2007. 

The results of this study show that most local PSGs occupy the ‘green valley’ and are rapidly transitioning to the low-mass end of the ‘red sequence’, with duration of this transitional period on the order of 1 billion years.  The structural morphology of local ‘green valley’ PSGs is very similar to that of low-mass early-type galaxies in the ‘red sequence’, even though star formation has only recently ceased.  This study suggests that these galaxies changed their shape and became bulge-dominated prior to the cessation of star formation, and therefore the transition through the 'green valley' will take approximately as long as it takes for the last batch of recently-formed stars to fade.  These local PSGs show that galactic interactions in recent epochs lead to the growth of the low-mass end of the 'red sequence' and agree with the idea of downsizing. 

Studying galaxy collisions, starburst galaxies, active galactic nuclei, and post-starburst galaxies is giving a clearer image on how galaxies evolve, and the star formation processes that occur during this transitional phase of galaxy evolution.

Darg et al. Article -- Utilizing Galaxy Zoo to examine properties of merging galaxies

Title: Galaxy Zoo: the properties of merging galaxies in the nearby Universe – local environments, colours, masses, star formation rates, and AGN activity

Authors: D. W. Darg, S. Kaviraj, C. J. Lintott, K. Schawinski, M. Sarzi, S. Bamford, J. Silk, D. Andreescu, P. Murray, R. C. Nichol, M. J. Raddick, A. Slosar, A. S. Szalay, D. Thomas, and J. Vandenberg

First Author’s Institution:  University of Oxford, Department of Physics

To read the full article, please click here.

Article Summary: 

Examining large-scale morphological properties of galaxy mergers proved to be a trying feat until the Galaxy Zoo project was set in motion.  Because of the great variety of configurations of mergers, visually examining images of galaxies is a much better method for identifying and classifying these specimens than using structural parameters.  Using classifications on the Galaxy Zoo interface, one can determine how ‘merger-like’ a Sloan Digital Sky Survey (SDSS) image appears to be based on the percentage of volunteers that flagged the particular image as a merger.  By utilizing the morphological data from Galaxy Zoo, Darg et al. were able to delve into important properties of merging galaxies, such as the structure of progenitor galaxies, internal properties of interacting galaxies, time-scales of merger events, local environments of mergers, star formation histories, and AGN activity. 

In this study, 3003 merging pairs were classified as well as a redshift-matched control sample.  The galaxies lie in the relatively local universe, with redshift range 0.005 < z < 0.1.  Binned redshifts for the merger and control samples are shown in figure 1.  Galaxies were classified both by their morphologies (Elliptical, Spiral, Unclear but probably Elliptical, and Unclear but probably a Spiral) and their merger stages.  Mergers could either be classified as ‘separated’, ‘interacting’, or ‘approaching post-merger’.  Of the 3003 samples, ~84% were classified as interacting, ~6% as separated, and ~10% as approaching post-merger. 

Figure 1. Binned redshift distributions for the merger and control samples.  

Figure 2. Number density of galaxies within 2.0 Mpc
of the merger and control samples.  Rho symbolizes
the Adaptive-Gaussian-environment parameter.  White
background shows galaxies in the field, dark gray shows
galaxies in clusters, and light gray shows the
intermediate regime.  

Of the merging systems visually examined in this study, there were about 3 times as many spirals than ellipticals.  This is interesting given that the ratio of spirals to ellipticals in the global galaxy population is ~1.5.  One issue that Darg et al. inquired about was the reason for this discrepancy.  Does it have to do with the environment in which these mergers take place or differences in the internal properties of these galaxies?  To parameterize the environment of these mergers, a sophisticated measure of the number of galaxies per unit volume called the ‘adaptive-Gaussion-environment parameter’ was used.  This allowed the determination of whether the merging galaxies and the control were located in the low-density field, high-density clusters, or in an intermediate regime.  Figure 2 shows that both mergers and controls peaked in a region dubbed ‘intermediate environments’.  Since mergers were also found to occupy similar if not denser environments than the control (where elliptical galaxies are more prevalent), the role of environment in causing the high spiral-to-elliptical ratio in mergers can be ruled out.  Instead, the prevalence of spirals in mergers likely arises from the longer time-scales of detectability for mergers involving spirals than for mergers involving ellipticals. 

Figure 3. Volume-limited and non-volume limited
distribution of galaxies in color space.  The graphs
on the left show the u-r color versus absolute
magnitude.  The center graphs show the frequency
of ellipticals and spirals compared to the control
sample (EU and SU stand for unsure but probably
elliptical and unsure but probably spiral,
respectively).  The right graphs show the frequency
of all merging samples compared to control.  

At least one of each of the galaxies merging had spectral data, allowing this study to do a color analysis of the samples.  In accord with earlier observations, the merging galaxies had a larger spread of colors than the control sample, supporting the notion that ‘irregular’ morphologies have a greater spread in color than ‘regular’ ones.  A volume limited (where only galaxies with M_r < -20.55 were used) and a non-volume limited color-magnitude diagram for the merger and control samples can be seen in figure 3.  A clear bimodality between the elliptical and spiral regimes can be seen in the binned color plots.  Figure 4 shows the mass-distributions of galaxies in both merger and control samples.  Across almost all environments, the spiral-galaxy stellar mass distributions appear to be the same in the mergers as in the control sample.  Ellipticals mergers on the other hand appear to be slightly more massive than their control counterparts.  When morphologies are not looked at, a very similar mass distribution for merger and control samples is attained.  The fact that mergers favor spirals (which are generally less massive) yet have an overall distribution just as massive as the control sample may indicate that galaxies involved in mergers really are more massive.  

Figure 4. Mass distributions for galaxies in all
environments (top row), galaxies in the field (second
row), galaxies in intermediate regimes (third row) and
galaxies in dense clusters (bottom).  

Figure 5 shows the entire sample of merger-pairs in a mass-color-morphology graph.  Both color and morphology of the galaxies scale strongly with mass.  An interesting find is that there is a near absence of ellipticals with masses below 3 x 10¹⁰ solar masses, raising the question as to what happens to two low-mass spiral galaxies when they merge.  This may be due to low-mass galaxies retaining a sufficient amount of gas to reform a disc after a major merger event (gas content along with conservation of angular momentum is what leads to a flattened-out disc shape in galaxies).  More massive galaxies may be prone to more catastrophic angular momentum loss during a merger event, and the remaining gas supplies may plunge into the central core and transfer the angular momentum required for disc morphology into the stellar dispersion of the remnant bulge. 

Figure 5. A mass-color-morphology diagram.  The top three plots show the average stellar mass for (from left to right) spiral-spiral mergers, spiral-elliptical mergers, and elliptical-elliptical mergers.  The main plot shows the u-r color of each sample, the stellar mass of each galaxy (with the more massive galaxy's value on the x-axis) and the type of galaxies that are taking part in the merger event (given by a circle, asterisk, or triangle data point).  

Due to the importance of feedback mechanisms to a gas retention model, the study next examined active galactic nuclei (AGN) and star formation signatures of the mergers.  By examining the measured fluxes of emission lines in the samples, this study was able to determine the most likely sources of these emissions and separate their galaxies into 4 categories: star-forming, mixed (both star formation and AGN activity), AGN (either Seyfert nuclei or LINERs), and quiescent (or ‘weak emission-line’).  Figure 6 shows the locations of these four types of mergers in mass-color space.  In this plot, galaxies characterized by star-formation occupied the low-mass region, AGN occupied the intermediate-mass region, and quiescent types occupied the high-mass region.  This suggests that the fuel supply of high-mass galaxies has been exhausted (as to not fuel star formation or AGN activity) and low-mass galaxies may have insufficient mass to power AGN.  Alternatively, AGN signatures in low-mass galaxies may also be obscured by high gas content and high star formation rates (SFRs).  By comparing spectral signatures to a control sample, the study determined that mergers significantly enhance SFRs in spiral galaxies only, whereas ellipticals live up to being ‘red and dead’ and their SFRs not as affected by major mergers.  Using H-alpha emission strength, estimations of the SFRs (of galaxies that fell in the star-forming category) were measured to be ~5.2 solar masses per year, which was about twice the value of a control sample of non-merging star-forming galaxies.  The highest SFR of the merging galaxies was ~ 95 solar masses per year. 

Figure 6. Color-stellar mass relation for galaxies of differing spectral types.  Each plot highlights the samples that fall into each respective category.  

The use of galaxy zoo morphology classifications allowed this study to analyze the effects of mergers on different types of galaxies.  By estimating the environment around mergers, the prevalence of spiral galaxies in merger events was found to not be due to the density of the environments in which mergers occur.  Therefore, internal properties of galaxies may be the reason for the high number of spirals in mergers; spirals have large gas reservoirs that may result in longer time-scales of merger events, whereas when two elliptical galaxies merge one would expect them to produce comparatively faint tidal tails and little star formation, thus making them harder to detect.  Since detectability of mergers relates to their timescales and timescales relate to internal properties of galaxies, addressing the colors, stellar masses, and spectral emission of the samples is of importance.  This study found that colors of merging galaxies scale strongly with mass and morphology, and are spread over a larger area than control galaxies.  Ellipticals are rare below a mass of ~ 3 x 10¹⁰ solar masses, which may be due to low-mass spiral mergers surviving the event and having enough gas to reform their disc.  Moving to the feedback mechanisms of the merging samples, Darg et al. found that mergers induce intense star-formation only when they involve spiral galaxies, and AGN activity was not present in low-mass mergers.  In star-forming mergers, the SFRs were ~2 times greater than that of a control sample of star-forming galaxies.  This study also found that specific SFRs (star-formation rates per unit stellar mass), scale down with stellar mass, possibly due to gas supplies being continually drained as galaxies accumulate mass.  The results generally imply that mergers affect spirals much more than ellipticals, which in turn affects the time-scales of detectability for merger events.