Tuesday, September 13, 2011

AGN Activity Increased by Galaxy Mergers?

Sometimes I forget the amazing things we take as given. Some months ago, a friend of mine asked if scientists had ever figured out if black holes exist. I replied by saying:

"Well, DUH! Of course they exist! They exist at the centers of every known galaxy. That's something everybody knows, right?".

If you really think about it, though, black holes, and their existence, are astounding. They are so astounding, in fact, that the man who predicted their existence was convinced they were too weird to exist in nature. That man was Albert Einstein.

Today, it's pretty conclusive that black holes are ubiquitous throughout the universe. They range in size from about the mass of our sun to gargantuan masses of billions of times as large (although there seems to be a curious deficiency in intermediate mass black holes - which would connect the smaller, stellar mass ones with the supermassive ones at galactic centers).

Establishing that black holes do exist, and are quite common, I'm going to go further and say that astronomers have seen two main types - those that are actively accreting gas, and those that lie dormant. Our Milky Way's black hole is a dormant type, and the ones currently feeding are called AGNs (Active Galactic Nuclei).The pretty picture below (stolen shamelessly from somewhere on the internet), shows a representation of an AGN.

Though there are different types of AGNs, they all are thought to stem from one physical model (more or less, there are still fights about specifics). In short, the rotating supermassive black hole is surrounded by a heated accretion disk. The continuum radiation from the infall of gas into the black hole photoionizes the surrounding gas, which creates emission lines we can see on Earth. In addition, a large disk of clumpy, optically thick clouds (or, clouds too thick to see through in some wavelengths, mainly optical in this case) surround the accretion disk, which can obscure some AGNs if the AGN is oriented in such a way that the clouds are between us and the black hole. Some AGNs also host twin collimated jets. These outflows emerge in opposite directions from the central black hole, and travel at velocities close to the speed of light. Heated bubbles from these jets can reach massive sizes, and how these jets impact galactic evolution (for it seems very likely they must have some effect), is not well known.

This all brings me to a paper I presented at Science Coffee last Thursday: The Impact of Galaxy Interactions on AGN Activity in zCOSMOS by Silverman et al. This study looked at pairs of interacting galaxies using Chandra X-ray data and the spectroscopic redshifts from zCOSMOS. The main point of this study was to see if galaxy mergers could increase the chances of AGN activity. The main thought is that if a gas-rich galaxy began the process of merging, the gas transferred from that galaxy could provide the fuel to feed the supermassive black hole.

Previous studies have been inconclusive about the effect of galaxy mergers on the probability of AGN activity. These studies (well, the ones mentioned) used Hubble Space Telescope (HST) observations to detect AGN activity, and they tended to conclude that they saw no significant increase in AGNs with merging galaxies. However, Hubble observes in optical wavelengths. Because of this, the thick clouds around the accretion disk could hide AGN activity, depending on the orientation of the galaxy. In an effort to get a more complete sample of AGNs, this new study uses X-rays from Chandra to detect AGNs, which are able to pass through the clouds.

To determine if galaxies in their survey sample were interacting, the authors looked at interacting pairs of galaxies between 0.25<z<1.05, distances of less than 75 kpc apart, and velocity differences of less than 500 km/s. This, of course, does not include galaxies that have already merged. The figure below shows some of their interacting pairs. The left panels show Chandra X-ray data (AGN detections), and the right panels show HST images. The red crosses denote AGNs, and the yellow crosses denote centers of normal galaxies. (Sorry this is so small. This is the largest size I can get on the blog).

They found that AGN activity was more common for interacting systems by about a factor of 2 over AGN activity in galaxies not interacting. In addition, for just the interacting pairs, they found a higher fraction of AGNs for lower physical separations and lower velocity differences. This seems to support the idea that more closely interacting galaxies were more likely to host an AGN. They even found a slight increase in AGN probability for interacting galaxies that were more comparable in size (though their sample was not statistically large enough to make that a very strong conclusion).

I think this study does make a point that galaxy mergers may have an effect on the probability of a galaxy hosting an AGN. How big is that effect? I don't know. Like the authors, I do agree that galaxy mergers can't be a dominant factor in deciding if a galaxy is active or not (about 80% of the AGN sample were in galaxies not interacting). However, they do seem more likely to spur AGN activity, and because of that, shouldn't be discounted entirely. 


  1. Haha! I think that was me. I remember posing the question to you and your boyfriend. I still need to see a bunny sent through the one in the middle of our galaxy...and where the heck are all the geth and quarians?

  2. This comment has been removed by the author.

  3. [Ack, changed my display name and it wouldn't update]
    This survey at z~2 comparing rest-frame optical morphologies of (X-ray selected) AGN hosts and quiescent galaxies seems to suggest that mergers aren't very important: http://arxiv.org/abs/1109.2588
    (We talked about this at our astro-ph coffee today, and I remembered your post!)

  4. Oh, I saw that paper too. I also mentioned it in my Science Coffee, ha ha!

    The only thing I have to say about it is that looking at objects at z=2 is really ambitious, since these galaxies are much dimmer at that higher redshift. Their galaxy images, as a result, don't look as good. Some look like fuzzy blobs to me, so I find it hard to believe they were able to classify everything accurately - as in, distinguish an elliptical from a disturbed galaxy.