Galactic Evolution

We all know how much of a fit creationists throw over biological evolution. We've also seen just how much of a fit they throw fits over stellar evolution. But there's change on even bigger scales as well.

From Brahe’s observation of a supernova in 1572, to Arp’s catalogue of odd an interacting galaxies, to Hubble’s observation of an expanding universe, it has become quite apparent in the history of astronomy that we live in a dynamic universe; it evolves. However, the evolution of some systems is easier to trace than that of others. With stars, we can observe clusters, in which only the mass of the various components differs significantly, in order to test our theories of stellar evolution. Unfortunately, for systems as large as galaxies there is no analogue.

One of the main reasons for this is a difference in the way stars evolve as compared to galaxies. Most stars are, for all intents and purposes, isolated systems, separated by vast interstellar distances. Even for the rare stellar systems that are close enough to undergo some sort of transfer, these amounts are typically only fractions of the mass of the objects. Thus, the evolution of stars is governed primarily by internal forces. While galaxies have gas and dust by which they can change their properties by making new stars and building heavier elements, galaxies are never observed in complete isolation. They are always members of some cluster in which interaction is inevitable.

As such, galaxies are not only subject to internal forces, but are also acted upon by external forces when they interact with other galaxies through glancing blows, mergers, or even cannibalization. This is true even in our own galaxy. While our nearest major neighbor may be quite a ways away, the Magellanic Clouds as well as several recently discovered dwarf galaxies swarm around us.

The current theory of galactic formation is that early in the universe, star formation began in smaller systems, which accreted into larger systems as the universe aged (Wiklind, 2007). As such, we would expect massive galaxies to be more prevalent at low redshifts. However, recent studies such as one by Wiklind et al (2007) looking at galaxies in the HDF, have shown that many massive galaxies with aged stellar populations exist at redshifts > 5, indicating that star formation occurred within a few hundred million years of the Big Bang.

For the first time, with larger and more advanced telescopes, are we able to peer back through true cosmological time scales to begin to see how galaxies have evolved as the universe has aged. Powerful new surveys, such as the Cosmological Evolution Survey (COSMOS), are new letting us place constraints on fundamental questions about how galaxies evolve. Questions of the evolution of the number density, when periods of star formation occurred, morphology, and chemical evolution can now be explored.

Mass Accumulation
Since, as previously mentioned, one of the driving forces of galactic evolution is that of accumulation of additional matter, it is of interest to study how this process occurs. This accumulation primarily occurs in two ways. The first is through the accretion of matter from the intergalactic medium (IGM). The second is through mergers with other galaxies that have already formed. The importance of each of these depends on the properties of the local universe at the time. If a great deal of raw matter is still available in the IGM as compared to the number density of galaxies, then the former process will dominate.

Fig 1. Fraction of bright galaxies in close pairs (5-20 kpc) vs (1 + z) for COSMOS field. Vertical error bars are 1σ. Star indicates local fraction. (Kartaltepe, et al., 2007).

One way this question has been approached, is to analyze the number of galaxies in close galaxy pairs at various redshifts. This method was undertaken by Kartaltepe et al (2007). By analyzing 1,749 galaxy pairs from the COSMOS field and comparing the number of paired galaxies to the overall number, they determined that the number of galaxies in close pairs increased significantly throughout the history of the universe (see Fig 1). Their study extended to z ~ 1.2, but they suggested that if the trend were extended to a distance of z = 2, it would indicate that nearly 50% of galaxies were in pairs during that time. However, little has been done in terms of high quality investigations for such limits.

Another interesting study indicating that mergers can strongly influence the properties of galaxies examined the density of galaxies (Trujillo et al, 2007). They found that, at redshifts of ~2, galaxies existed that had a density almost two orders of magnitude higher than any found in the present universe. Because of this, they suggest that such compact galaxies must have merged with others.

But mergers are not the only form of mass intake which galaxies can undergo. Accretion of matter from the IGM also plays an important role. Simulations by Semelin and Combes (2005) have indicated that mass gained via accretion exceeds that of mergers by a factor of 2 to 4. Before z ~ 2, the importance leaned more towards the factor of 4 while more material was still available. After that time, accretion should have become less important. This is somewhat supported by a study done by Netzer and Trakhtenbrot (2007), which looked at the growth time of AGN at z < 0.75 due to accretion and found that the amount of time they should have formed in is older than the observed age of the universe, thus agreeing with the conclusion that the rate of accretion is decreasing towards present time.

Star Formation & Luminosity
The process of adding more material almost certainly induces periods of star formation in galaxies. Accretion passively provides new raw resources; mergers and close encounters provide perturbations necessary to trigger collapses, and can have dramatic consequences, as demonstrated by the M81 group. Where new star formation occurs, so is there an excess of luminous stars, brightening the overall galaxy. Thus, if there is a correlation between the amount of mass and methods of gain through cosmological time, there must also be a relation to luminosity.

Fig 2. Star formation rate at 1,900 Angstroms as a function of resdhift (Bouwens & Illingworth, 2006).

To investigate this, many authors begin by examining what should be expected photometrically from a passively evolving galaxy in which no new star formation is taking place. Observations are then compared to this standard. In this area, studies have indicated that more massive galaxies (typically taken to be M > 10¹¹ M_Sun) show less evidence for luminosity evolution than their lower mass counterparts (Bower, Lucey, & Ellis, 1992). This suggests that the most massive galaxies underwent a large burst of star formation early in the history of the universe, but have not done any significant star formation since that time.

In other cases, significant evolution is frequently noted. A study by Dahlen et al (2007) investigated the star formation and luminosity functions for lower mass galaxies. They concluded that star formation rates have been increasing towards present time and that specifically, in the spectral regimes they examined, this led to an increase of nearly 1 magnitude since z ~ 1.73. Their survey did not extend past this redshift, but other studies have indicated that there may well have been a peak in the star formation rate near z ~ 2. This was the conclusion reached by Bowens & Illingworth (2006) and is illustrated in Figure 2.

Morphology

Fig 3. Size - redshift relation for disk galaxies selected by absolute magnitude. Blue dots show the median value in each redshift bin used. The solid line shows the best-fitting size evolution (1+z)^1+m, where m = 1.1. Also shown are theoretical curves if sizes evolve as r is proportional to H(z)^-1 (dashed line) and r proportional to H(z)^-2/3 (dotted line). (Dahlen et al, 2007).

Another consequence of the evolution of galaxies is that morphologies will change as galaxies evolve. While morphologies can dramatically change due to interactions mangling structure, morphology can also be driven by more quiescent processes, such as the location of star formation. Such an investigation was also carried out by Dahlen et al (2007). They discovered that the number of galaxies with structure dominated by their bulges drops, approaching present time, from an average of ~10% of galaxies at z ~ 0.5 to ~30% at a redshift of 2.2. They also find that the overall size of galaxies has been increasing towards the present, as illustrated in Figure 3.

Meanwhile, in the arena of morphological evolution, there seems to be a more questionable relation to mass and limits of evolution. According to Conselice et al (2007), galaxies with masses of 10¹¹ M_Sun are observed to have a consistent fraction of ellipticals (~70-90%). Other studies, such as that of Cresci et al (2006), have also indicated that massive galaxy morphology may be more stable over long periods. However, for galaxies above 10^11.5 M_Sun they discovered that there has been a ~20% increase in frequency of such galaxies since z = 1.2. Additionally, there has also been an increase in the frequency of the 1010 MSun spiral galaxies since the same time, where peculiar galaxies of that sort have decreased in commonality. These findings are shown in Figure 4.

Fig 4. Frequencies of various types of galaxies vs. redshift for two binnings of mass. (Conselice et al, 2007).

It should also be noted that there is a strong correlation between galaxy morphology and the number of other galaxies in near proximity. Galaxies in clusters are significantly more likely to be of the “early type” (elliptical and lenticular) than a typical field galaxy (Smith et al, 2005). As with most other properties, we may ask whether this correlation also evolves. The findings of Smith et al (2005) suggested that for most clusters, the likelihood of a particular galaxy being early type is roughly constant over the past 7 Gyr. Only the densest clusters they studied showed strong evidence of any evolution of frequency.

Chemical Evolution
As stars are formed and die, they inherently enrich their host galaxies. From my review of the literature, it appears that most of the investigation into this topic has come in the form of modeling and very little has been applied in the way of constraints. One of the few studies that does make this attempt is that of Fritze, et al (2002). They applied observations of Damped Lyman α Absorber (DLA) galaxies to various models for the chemical evolution of spiral galaxies. They concluded that these early DLA galaxies followed the general trend set forth by models, suggesting that they may well be progenitors of spiral galaxies we see in the universe today.

Conclusions
In this post we have investigated four major galactic properties as a function of redshift in order to infer the manner by which galaxies evolve with the universe. It was shown that the number of galaxies in close pairs has been decreasing as the universe has aged. Star formation has also been decreasing in recent cosmological times, after apparently having a peak near a redshift of z ~ 2. The issue of morphology tends to be somewhat more difficult to untangle, as many factors seem to have an effect on this property. In general, galaxies around 10¹¹ M_Sun seem to be somewhat resistant to morphological evolution, although more massive galaxies seem to be susceptible. While chemical evolution also undoubtedly occurs, little seems to be available in the literature as to how this has related to redshift for various types of systems.

In general, evolution is an important a force in the universe at large as it is for life on Earth. It sculpts galaxies and makes them shine.

---

-Bouwens, R. J., Illingworth, G. D., 2006, Nature, 443, 189.
-Bower, R. G., Lucey, J. R., Ellis, R. S., 1992, MNRAS, 254, 601.
-Conselice, C. J., 2007, MNRAS, 381, 962.
-Cresci, G., et al, 2006, preprint, arXiv:astro-ph/06072212v2.
-Dahlen, et al., 2007, ApJ, 654, 172.
-Fritze, U., Alvensleben, V., Linder, U., Fricke, K. J., 2002, Cosmic chemical evolution. Proceedings of the 187th Symposium of the International Astronomical Union, held at Kyoto, Japan, 26-30 August 1997, 147.
-Kartaltepe, J. S., et al. 2007, ApJ, 172, 320.
-Netzer, H., Trakhtenbrot, B., 2007, ApJ, 654, 754.
-Smith, G. P., et al, 2005, ApJ, 620, 78.
-Trujillo, I., et al, 2007, MNRAS, 382, 109.
-Wiklind, T., et al, 2007, preprint, arXiv:0710.0406v1