What is an active galactic nucleus (AGN)?

Actually this question is not so easy to answer. There are so many different answers in literature that, when you start studying AGN, you learn really fast to make your own picture.

My picture is perhaps not the most common one, because I do cosmological simulations, where you always need so-called sub-grid models. The aim of these models is to account for physics you can’t resolve in the simulation. But you really need these models because their effects are on scales you do resolve.
In general, an AGN is a supermassive black hole (SMBH) in the centre of a galaxy, which accretes so much gas that it gets active. However, AGN do not only accrete gas, but they also eject energy in the form of AGN feedback as shown in the cartoon (I got the idea for this cartoon from a talk by M. Dotti, 2015). Usually, cosmological simulations account only for one form of feedback, which is thermal feedback, where we heat up the gas surrounding the black hole. This heating is caused through radiation, which is observed as the AGN luminosity, but actually, there is another kind of feedback, which is observed as jets, which transport energy far away from the black hole. In simulations, this means that gas particles are kicked away with highly relativistic velocities.


However, it is still not completely understood what determines how much of the feedback energy goes into radiation and how much into jets. We addressed this question in Steinborn et al. (2015), where we split up the feedback energy into radiation and mechanical outflows.

All in all, my picture is now the following:
1) Always when gas falls into the centre of a galaxy, an accretion disc forms around the central SMBH. The accretion leads to radiative feedback and it spins up the black hole, which in turn increases the efficiency of the radiative feedback.
2) The spinning black hole builds up magnetic field lines which accelerate charged particles perpendicular to the accretion disk and hence, a jet forms. The black hole accretes less and less, because the gas in the accretion disc is simply used up.
3) Probably the jet can still live for a while, when the accretion disc is used up.

AGN Illustration


This was my picture of AGN, but I’m sure, there are many, many others …


Comment on the unified model of AGN

Originally the unified model of AGN says that intrinsically all AGN have the same structure: a torus, an accretion disk, jets, etc. The different observed AGN types can then be explained only by the different viewing angles.
Recently I was on two conferences in which the participants came to very different conclusions: in one the unified model was declared as dead while in the next conference it was declared as alive!
On first sight people seem to have very different views. However, I think that both views are actually based on the same idea:
Of course AGN do not all look the same! However, they can all be described with the same components given by the unified model, if we accept the following:

  • the different components can be more or less dominant or even not present at all! For example, I don’t think that a torus is always necessary to explain an AGN.
  • AGN evolve over time. For example during/after a merger they go through different phases.
  • There can also be other external factors which shape the spectra of AGN like dust lanes near the AGN.

In addition to the viewing angle, these factors can explain the different observed AGN types.
So whether you declare the unified model as dead or alive is just a matter of definition.



Origin and properties of dual and offset AGN in a cosmological simulation at z=2


In our most recent simulation the combination of a large volume and a high resolution, as well as an improved dynamical treatment of black holes (BHs) allows us, for the first time,  to study BH pairs with spatial separations of only a few kpc in a cosmological hydrodynamic simulation. These BH pairs are the consequence of galaxy mergers. Some of them are even dual and offset AGN, which are BH pairs where both BHs or only one BH is active. In my most recent paper I was especially interested in the different origin and properties of of dual AGN, offset AGN and inactive BH pairs.

See Steinborn+16 for more details.