Counting galaxy mergers you can’t see

[tweetmeme only_single=false service=wp.me source=allinthegutter]

I’m going to pick up where I left off a while ago, when we talked about galaxy evolution. I have a staggering backlog of papers to read on my desk, most of which have the words “merger history”, “mass assembly” or “galaxy pairs”. All of these expressions are more or less equivalent, and they relate to the one of the processes we believe regulates galaxy growth – merging. The other one is star formation – by which I mean the process of turning cold gas into stars – but we’ll talk about that some other time.

We know that galaxies merge. We do not only see it  (go to Galaxy Zoo’s Mergers site to revel in pretty mergers images, and also to help astronomers get some real science out of them), but it is also a prediction of our current model of structure formation. I’ll cover that model and prediction in another post, but for now let me just say that we are in a position in which measuring the rate at which galaxies of different mass or luminosity merge in the Universe is becoming very important as a way to constrain our models of galaxy and structure formation. In other words, it’s time to get quantitative.

So what we want to know is, on average, how many galaxies merge per unit volume, per unit time, as the Universe evolves. If you sit down and think about this for a moment or two, you’ll quickly come to the conclusion that simply counting galaxies that are merging (which you can identify by looking at the images) is one way to go. But this is only possible relatively near by – as we go to higher redshift it becomes increasingly hard to get good enough images. Still, a number of people have been working hard at measuring this, and pushing this sort of analysis forward.

Another way to go, is to simply count galaxy pairs that are closer than a given physical distance. You can assume that if galaxies are too close then gravity will win at some point, and the galaxies will merge. The upshot is that you don’t need really good images to actually see interacting galaxies and you can take this to higher redshift. The downside is that you need to make assumptions about what this physical separation should be and, perhaps more importantly, how long it will take them to merge – the dynamical timescale. Another disadvantage comes from the fact that you miss pairs of galaxies in which one of the two is very faint – so you are limited to counting pairs of bright galaxies. The jargon for the merging of two galaxies of similar mass or luminosity is “major merging”. Some neat pieces of work have come out of this, and have measured the major-merger rate of luminous galaxies to a respectable redshift. The last one I read (but by no means the only, nor the last!) was by  Roberto de Propris et al. (2010) who did this up to a redshift of 0.55, but there are measurements of this quantity which span the last 8 Gyrs of the lifetime of the Universe or so (equivalent to a redshift of 1).

A few weeks back however, I read another paper which took a different and rather interesting approach to the subject. This is the work of Sugata Kaviraj et al. (2010), and their idea is as follows. The types of measurement like the ones I described in the above paragraph give you a number of how many major mergers there are at some point in the Universe. These mergers, however, leave a signature in the shapes of the galaxy for a certain time – they look disturbed (i.e., not smooth), until they final relax into one larger, smoother, and stable galaxy. However, this means that you should be able to predict how many galaxies of a given mass, on average, should look disturbed at any point in time by assuming a measured rate of mergers in the past.

And so they did. They took a whole load of high-resolution images from the Hubble Space Telescope and looked for signs of disturbed elliptical galaxies. What they found (amongst other neat things that I don’t have time to go into), is that there are too many of these disturbed galaxies if we assume that the other rates are correct. But hang on in there for a minute – the other rates are limited to major mergers because we can’t see the minor mergers when looking at pairs. So Sugata Kaviraj and collaborators postulate that the excess is due to these minor mergers – we can’t see them happening at high redshift, but we can see their effect at lower redshift. Moreover, they also observe these minor mergers to be significantly more dominant than major mergers since redshift of one, suggesting that galaxies have been growing from accreting smaller (fainter) galaxies in the recent Universe, but this was potentially very different at high redshift.

Other people have found this sort of behaviour in some way or another (including me!), but I was happy to see a rather neat way to.. well.. see (and measure) the unseen.

R. De Propris, S. P. Driver, M. M. Colless, M. J. Drinkwater, N. P. Ross, J. Bland-Hawthorn, D. G. York, & K. Pimbblet (2010). An upper limit to the dry merger rate at ~ 0.55 ApJ arXiv: 1001.0566v1

Sugata Kaviraj, Kok-Meng Tan, Richard S. Ellis, & Joseph Silk (2010). The principal driver of star formation in early-type galaxies at late epochs: the case for minor mergers MNRAS (submitted) arXiv: 1001.2141v1