Puffing up elliptical galaxies October 3, 2009
Posted by sarah in: new astronomy, science , trackback
Elliptical galaxies are the boring uncles of the galaxy family: they’re amorphous blobby things, ubiquitous in the Universe, that contain a fairly uniform population of old, red stars. Without the interstellar gas and dust that is needed to harbour pretty sites of star formation, they are supremely unphotogenic. But they have far more going on beneath their featureless surface: the complex dynamics inside many ellipticals show evidence of a turbulent past and, with many of the most massive known galaxies in our local Universe being ellipticals, they clearly play an important role in galaxies’ evolution.
Studies seem to suggest that high-redshift elliptical galaxies are more compact than their present-day counterparts (figure from Glazebrook, 2009)
Observational surveys of elliptical galaxies at high redshift have in recent years revealed a further interesting fact: ellipticals at high redshift appear to be much smaller in size than those in our local Universe, but have about the same mass and density of stars. In a recent ApJ paper, Ivana Damjanov of the University of Toronto and collaborators describe how a sample of elliptical galaxies at redshifts 1 to 2 looked 2-3 times smaller than those in the local Universe. The first surprise lies in that they evolve at all between redshift 2 and 0. In our current understanding of galaxy formation and evolution, ellipticals are the “red and dead” endpoints of evolution.
Moreover, if we reasonably assume that the local ellipticals are simply the descendants of high-redshift ones, how come these “red nuggets” evolve into the larger galaxies we see today, without adding substantial mass or stars? This has been a hotly debated issue in the last couple of years and the subject of many recent papers (I’ve provided some references below).
Galaxy Wars
The highly compact elliptical galaxies seen at high redshifts are almost non-existent in our present-day neighbourhood. Whatever process causes the galaxies to grow in size must act indisciminately on all shapes and sizes. Their demographic of stars doesn’t change either: ellipticals contain a pretty uniform population of old red stars that appear to have formed well before their morphological evolution in the last 10 billion years.
From observations we know that mergers between galaxies occur. We’re pretty sure they play in important part in the long-term evolution of galaxies, and the collision of two galaxies can certainly result in a larger end product. But the conditions under which this can result in a larger elliptical galaxy without adding a lot of mass are very specific.
Neither galaxy can contain much gas, as the gravitational turmoil of a galaxy collision in the presence of large amounts of interstellar gas would trigger a burst of star formation. Today’s ellipticals show no evidence of such young stars. The merger would also have to take place between galaxies of unequal sizes, as an equal-mass merger would result in a much larger final star count. Finally, as each of these merger events result just a small increase in size, a typical elliptical would have to go through many mergers, all of them minor and gas-poor, to puff them up to the sizes we see today. Although much is unknown about the frequency of galaxy collisions, the specific constraints of this scenario compared with the observed ubiquity of the process make it seem rather unlikely.
Hidden internal strife
Maybe galaxies can grow in size all on their own, without interacting with others. An effective way of making a galaxy expand is by removing some of its mass, so that the drop in gravitational force binding it together causes the stars to drift apart over time. Last year a team of Chinese and Italian astronomers proposed that a short burst of quasar activity early on in the galaxies’ lifetime can blow out a large fraction of their mass, causing them to expand and settle into a larger size.
Ellipticals are known to host black holes in their centres, and we’ve seen such quasar-driven outflows occur in galaxies at high redshift, thought to be powered by accretion of material onto a central black hole. But can such a violent episode in a galaxy’s lifetime really occur without leaving any noticable scars? And does each quasar event blow out just the right amount of mass to cause the observed puffing up? If too little mass is stripped, the expansion would be too small; if too much, the galaxy is simply destroyed.
Or could we be creating a problem where really there isn’t one? Several recent papers by separate groups of astronomers examined the stellar density profiles of samples of elliptical galaxies at high and low redshifts, and suggest that perhaps the galaxies don’t need to puff up at all. Maybe the stars in their outer regions are already there, only they’re too faint to see at such large distances. A recent paper by Italian astronomer Mancini and collaborators studying a few of the most massive high-redshift ellipticals observed with Hubble did not look smaller at all than our present day neighbours. This beautifully simplistic solution appears to be gaining support in the community.
Dark forces at work?
One recent development in this discussion, described in the paper that sparked my dig into this issue, invokes the stuff we can’t see to explain the ellipticals’ apparent bloating. Japanese astronomer Tomonori Totani of the University of Kyoto raises the possibility that dark compact objects, such as primordial black holes, could reside in the galaxies’ haloes and transfer energy to the outer stars, causing them to spread out and increase the host galaxy’s apparent size.
This is an appealing suggestion, were it not for dark compact objects being rather unlikely dark matter candidates. While our understanding of dark matter is still pretty basic, several observational surveys have posed some strong constraints on what it could be made of (see e.g. this paper for a discussion). Even if dark compact objects, such as primordial black holes do fit the bill, they aren’t likely dominant constituents of DM.
The apparent puffing up of elliptical galaxies is an intriguing problem that is keeping galaxy evolution folks pretty preoccupied at the moment. Being able to observe to redhsifts 2-3 with more sensitive instruments, such as the newly commissioned ones on board Hubble, will allow them to get a better signal from these faint sources and discern any extended outer regions that may currently be hidden in the background noise. That will confirm (or otherwise!) the increasingly popular hypothesis that the compact nuggets of galaxies at high redshift are simply the only parts of the larger whole that are bright enough for us to see.
Reading
Damjanov, I. et al. (2009). Red nuggets at z~1.5: Compact passive galaxies and the formation of the Kormendy relation The Astrophysical Journal, 695 (1), 101-115 DOI: 10.1088/0004-637X/695/1/101
Bezanson, R. et al. (2009). The relation between compact, quiescent high-redshift galaxies and massive nearby elliptical galaxies: Evidence for hierarchical, inside-out growth The Astrophysical Journal, 697 (2), 1290-1298 DOI: 10.1088/0004-637X/697/2/1290
Fan, L., Lapi, A., De Zotti, G., & Danese, L. (2008). The Dramatic Size Evolution of Elliptical Galaxies and the Quasar Feedback The Astrophysical Journal, 689 (2) DOI: 10.1086/595784
Glazebrook, K. (2009). Galaxy formation: Too small to ignore Nature, 460 (7256), 694-695 DOI: 10.1038/460694a
C. Mancini, E. Daddi, A. Renzini, F. Salmi, H. J. McCracken, A. Cimatti, M. Onodera, M. Salvato, A. M. Koekemoer, H. Aussel, E. Le Floc’h, & C. Willott (2009). High-redshift elliptical galaxies: are they (all) really compact? MNRAS accepted arXiv: 0909.3088v1
Tomonori Totani (2009). Size Evolution of Early-Type Galaxies and Massive Compact Objects as the Dark Matter PASJ arXiv: 0908.3295v1
Image: Nature, 2009
Comments»
Hello SarahAskew,
Great post! This subject really is fascinating. When I first heard about this problem via Julianne Dalcanton at Cosmic Variance (I’d give you the link to her post but I can’t find it) I thought the solution would simply be galaxy collisions/mergers that enlarge the galaxies. But, of course, the story isn’t so simple as the mass of these beasts seems to remain constant.
On the other hand, I don’t understand why unseen stars in the outer regions of these galaxies solves the problem. Hasn’t it been shown* that the stars we do see have high velocity dispersions? This means that the mass contained within these star’s orbits is very high, so we still have massive galaxies that are denser than their present-day siblings.
-Grad Student
*Actually, maybe this is the only case where it has been “shown.”
Van Dokkum et al 2009:
http://arxiv.org/abs/0906.2778
Thanks. This is obviously a potential subject for an entire paper – and I admit this is not my own specialty. I just started reading some papers, found it all a bit intriguing, talked to some people and took it from there. So I’ll refrain from trying to explain the finer details of velocity dispersion measurements.
Essentially, I would keep the following in mind:
- yes, the implied size growth between z~2 to 0 would imply an increase in velocity dispersion at high redshift, and there is evidence of this. Cenarro & Trujillo (2009) show an increase of ~1.3 in sigma between present-day spheroid galaxies and those at z~1.6, based on a stacked spectrum of 13 galaxies. This would support a (minor) merger scenario.
- it does not, however, seem to rule out the explanation that the red nuggets are simply the cores of today’s massive ellipticals. Hopkins (2009) argues that the accretion of low-density material over time can affect the effective radius and the surface brightness profile, but would not have a large effect on sigma.
- the observations presented in van Dokkum’s (2009) paper are really pushing up against the limits of what can be done with today’s instruments. The authors themselves acknowledge that the spectral features used to fit the velocity dispersion have low signal to noise. Moreover, calculating the mass of the galaxy relies on fitting the surface brightness profile. These high-redshift galaxies as imaged with NICMOS on Hubble are barely resolved (see Kriek et al (2009) for images) – so most of what you see in the image is just instrumental point spread function. So these model fits are basically not representative in the galaxies’ central regions, and the resulting mass has significant uncertainty.
- that said, I know some of these astronomers and would certainly trust them to do a thorough job with the data they have. So if we accept that this galaxy has the high sigma they present, it’s still only one galaxy, i.e. not exactly a “representative sample”. For that we’ll have to wait for a new generation of spectrographs with better sensitivities and spatial resolutions that will allow more reliable mass and sigma measurements.
Anyway this is kind of the way I see it. Do you think that makes sense? Sorry for the delay in replying – I wanted to read up a bit more on the subject before saying something incredibly stupid.
Hi, Sarah,
Thanks for all the information in your post about the activity around the issue of how red ellipticals seem to have puffed up since their early days.
I’d like to keep track of this issue, to follow up on my own post (here) on the subject – and thanks for your comment on it.
The whole area of galaxy evolution is getting pretty interesting, especially with all the new technology almost available to do much better studies.