Kaviraj et al. Article -- Ultraviolet Analysis of Post-Starburst Galaxies and Quenching Mechanisms

Title: UV properties of E+A galaxies: constraints on feedback-driven quenching of star formation

Authors: S. Kaviraj, L. A. Kirkby, J. Silk and M. Sarzi

Authors’ Institutions: University of Oxford Department of Physics and University of Hertfordshire Centre for Astrophysics Research

To read the full article, please click here. 

Article Summary: 

An image of galaxy NGC 3801 combining light from across the spectrum, ranging from ultraviolet to radio.  NASA's GALEX and other instruments caught this galaxy in the act of quenching its cold, gaseous fuel for new stars - possibly marking the transition from a star-forming spiral galaxy to a quiescent elliptical galaxy.  According to theory, star formation will soon be quenched by the shock waves from two powerful jets shooting our of NGC 3801's central supermassive black hole, as seen in the radio emission colored green.  Image courtesy of NASA/JPL-Caltech.  

The study of Post-Starburst Galaxies (PSGs) and the mechanisms that quenched their star formation provides key insights into understanding the processes that shape galaxy evolution.  PSGs offer a look at a valuable evolutionary link between gas-rich star-forming galaxies and gas-poor quiescent galaxies.  A study by Kaviraj et al. in 2007 carried out the first large-scale examination of PSGs with ultraviolet (UV) photometry.  Due to the sensitivity of the UV to young stars, this study was accurately able to reconstruct the star formation histories of 38 PSGs in the nearby Universe by combining optical and UV data from the Sloan Digital Sky Survey (SDSS) and Galaxy Evolution Explorer (GALEX) surveys. 

PSGs, also known as ‘E+A’ Galaxies, show strong Balmer absorption lines that are characteristic of recent star formation but lack the forbidden [OII] and H-alpha emission that are present during ongoing star formation.  This indicates that these galaxies have recently had a strong episode of star formation that was abruptly quenched.  Understanding the processes that ‘quench’ these galaxies is an important step to understanding this transitional period of galaxy evolution. 

At intermediate redshifts (z ~ 0.5), PSGs were found to be primarily in clusters of galaxies, as opposed to in smaller groups or in the field.  For this reason, these galaxies were believed to result from cluster-specific mechanisms such as galaxy harassment or ram-pressure stripping (the stripping of galactic gas as galaxies travel through the cluster).  However, local observations indicate that PSGs are much more common in the field.  This indicates that other channels likely exist in the production of PSGs.  Many PSGs exhibit morphological disturbances, which may mean that their evolution is linked, at least partially, to mergers and interactions.  Simulations support this hypothesis, indicating that gas-rich mergers are capable of triggering strong star formation episodes. 

The criteria used in the selection of this study’s PSG sample were similar to earlier studies: H-delta (EW) > +5.0 Å, H-alpha (EW) > -3.0 Å, and [OII] (EW) > -2.5 Å, where a positive or negative sign denotes absorption or emission lines, respectively, and EW stands for equivalent width.  To ensure accuracy, the sample was restricted to the redshift range 0 < z < 0.2, a signal-to-noise ratio greater than 10, and galaxies with evidence of an Active Galactic Nuclei (AGN) were removed. This was because the scattered light from the AGN could contaminate the UV continuum. The authors also checked the morphologies of the sample using the SDSS fracDev tool. They found that they have spheroidal morphologies, which provides support for the idea that PSGs are precursors of early-type galaxies.

Figure 1. Position of E+A galaxies used in the study (filled blue circles) compared to a sample of early-type galaxies from SDSS DR5 in (NUV – r) versus (g – r) color space.  

The seven photometric filters used were the five SDSS bands (u, g, r, i, z) and the two GALEX filters in the far-ultraviolet (FUV) and near-ultraviolet (NUV).  Figure 1 shows the position of the PSG sample in (NUV – r) versus (g – r) color space, compared to a sample of early-type galaxies.

Figure 2 shows the PSGs approximate ages, mass fractions (amount of stellar mass formed during the starburst compared to the mass of the galaxy), time-scales, and star formation rates (SFRs). They derive the SFR by dividing the stellar mass formed during the starburst by the estimated time-scale of the starburst.  While low-luminosity PSGs have implied SFRs less than 50 solar masses per year, high-luminosity PSGs exhibit SFRs greater than 300 and even as high as 2000 solar masses per year.  These SFRs are comparable to those found in Luminous Infrared Galaxies (LIRGs) and Ultra-Luminous Infrared Galaxies (ULIRGs) at low redshifts, indicating that massive LIRGs could potentially be the progenitors of massive PSGs. 

Figure 2. The top plot shows the age of the burst versus the mass fraction.  The middle plot shows the binned timescale of the starburst.  The lower plot shows the implied star formation rate versus z-band magnitude.

One of the most important aspects of starburst galaxies is the quenching mechanisms that truncate their bursts.  PSGs experience ‘negative feedback’ that causes the star formation rate to slow down.  In this study it was found that for galaxies below a mass of 10¹⁰ solar masses, the quenching efficiency decreased with an increasing galactic mass.  However, for galaxies with masses greater than ~10¹⁰ solar masses, this trend was reversed; quenching efficiency increased with an increasing galactic mass.  Figure 3 shows the relationship between galaxy mass and quenching efficiency, with a clear change at ~10¹⁰ stellar mass.  This observation suggests that there are two primary sources for negative feedback: supernovae and AGN.  In the absence of AGN, supernovae would be the primary source of negative feedback.  As galaxies become more massive, the depth of the potential well increases, making it more difficult for supernovae to eject gas from the system.  However, for galaxies greater than ~10¹⁰ solar masses, AGN begin to appear and become the dominant source of negative feedback.  Since the mass of the black hole scales with the central velocity dispersion, it is expected that AGN feedback will become more effective as the galaxy mass increases.  Because galaxies with ongoing AGN activity were excluded from the sample, it is plausible that AGN feedback processes simultaneously quench both star formation and AGN activity.  Through quantitative analysis, Kaviraj et al. were able to support these qualitative predictions. 

Figure 3. A log-log plot of galaxy mass vs time-scale ratio.  The time-scale ratio is the ratio of the time-scale of the burst to the dynamical time-scale of the galaxy, which describes the 'natural' time-scale over which processes such as star formation would take place if left unhindered.  Note that time-scale ratio and quenching efficiency are inversely correlated, so an decreasing slope is analogous to a increasing quenching efficiency.  

One last investigation in this study was to probe the migration time from gas-rich star-forming galaxies to gas-poor quiescent galaxies, also known as moving from the ‘blue cloud’ to the ‘red sequence’ (see Figure 1 in the Wong et al. article summary).  Migration times were estimated by ‘ageing’ the best-fitting star formation model of each PSG.  Most galaxies complete their migration time within 2 Gyr, with a median migration time of ~1.5 Gyr.  Figure 4 presents the migration tracks of PSGs in color-color and color-magnitude space. 

Figure 4. Migration tracks for E+A galaxies in the (NUV – r) versus (g – r) color space (top panel) and the (NUV – r) versus M(z) color-magnitude space (bottom panel).  Ages (in Gyr) along the track are shown color-coded.  

In conclusion, by combining the optical and UV data, this study was able to reconstruct the time-scales, mass fractions, SFRs, migration times, and quenching mechanisms in this sample of PSGs.  This study suggests that supernovae are the primary quenching mechanism for galaxies under 10¹⁰ solar masses, and AGN become the primary source of negative feedback for galaxies over ~10¹⁰ solar masses.  When supernovae are the primary source, quenching efficiency decreases with galaxy mass because the increasing depth of the potential well makes it more difficult to eject gas from the system.  As AGN become the dominant source of negative feedback, quenching efficiency increases with galaxy mass, due to the AGN luminosity scaling with the mass of the black hole.  The study of PSGs helps us understand the processes that shape galaxy evolution. Future comparative studies of PSGs at low and high redshifts could help provide insight into the processes that dictate galaxy evolution over cosmic time.