TITLE: Interplay among cooling, AGN feedback and anisotropic conduction in the cool cores of galaxy clusters
AUTHORS: H.-Y. Karen Yang and Christopher S. Reynolds
FIRST AUTHOR INSTITUTION: University of Maryland
STATUS: Submitted to the Astrophysical Journal
Galaxy clusters are fascinating places. The confluence of the threads of the cosmic web (the large scale distribution of galaxies), they are the largest structures in the universe held together by their own gravity. Most of the mass comes from invisible dark matter, the gravitational glue that binds all large structures in the universe together, and which here creates a deep gravitational well in which the cluster’s galaxies are held. But the cluster’s bright mass actually mostly isn’t in galaxies: it’s in diffuse gas that sits between them – the intra-cluster medium (ICM).
The ICM is a pretty strange substance. It’s an extremely diffuse plasma at a temperature of 10-100 million K, and it emits X-ray radiation which freely streams out of the cluster. That means that energy is leaving; the cluster is cooling. As the gas cools, it should become denser and drop further into the gravitational well of the cluster. At the bottom of the well we expect to find the cluster’s brightest galaxy (the “Brightest Cluster Galaxy” or BCG), usually a large elliptical galaxy formed from the merging of several smaller galaxies that found their way to the core of the cluster. Star formation in this galaxy is kept alive by the regular infusion of gas that cools sufficiently to fall onto it. That’s where the problems start: this star formation is much less vigorous than a quick calculation based on how much energy the ICM is losing would suggest. In other words, the flow of cooling gas is being restricted.
This “cooling flow” problem has occupied the attention of astrophysicists for more than two decades and is now more-or-less solved. To solve it, you need a heating source powerful enough to balance the cooling. But it can’t be just any heat source; it needs to react to the flow of cool gas by getting stronger. If you don’t have this negative feedback, you’re stuck: your heat source either does too much (arrests the flow of cool gas entirely, leading to no star formation at all) or doesn’t do enough (the gas continues to pile up, becoming more dense and cooling even faster, leading to runaway cooling after all). The best candidate? The supermassive black hole we expect to find at the centre of the BCG. An accreting supermassive black hole (an Active Galactic Nucleus, or AGN) is more than capable of launching jets with sufficient energy to reheat the ICM, and we can actually see the effect of this in X-ray images of the ICM.
Figure 1 is taken from Sanders & Fabian (2007). What you’re seeing here are shocks and cavities left behind in the ICM as the AGN jets ploughed through it. The ripples are actually sound waves in the ICM; the AGN is playing a kind of ‘cosmic symphony’ in the cluster core.
So, we think we have the answer, but the details of how this release of energy actually heats the ICM and arrests the cooling flow remain unclear. That’s where today’s paper comes in. The authors run a simulation of a cool-core cluster with a plethora of different physical effects included: in particular, they are careful to include the effects of magnetic fields, making this a magnetohydrodynamic (MHD) simulation. It sounds complicated, but it just means they add some electromagnetic forces to the equations of motion that govern the system. In the ICM plasma magnetic field lines can either suppress or boost thermal conduction, a process which could mitigate cooling by channelling heat inwards. Thus, modelling the magnetic field is very important.
The AGN feedback is modelled as a pair of jets which inject mass, momentum, and energy into the simulation in the central pixel. It’s not possible to model the detailed physics of the black hole and the cluster simultaneously because their physical scales are so different, so the authors just have to see what the results of feedback look like in the real universe and model those as carefully as possible.
Figure 2 shows some of the results, displaying the simulation at the start and end of one round of AGN feedback. Conduction alone is not a good way to prevent cooling partly because the flow of heat tends to align any magnetic field unfavourably for further conduction. However, the AGN jets change all that by disrupting the ICM and introducing turbulence. This allows a stable flow of heat into the core, adding to the energy supplied by the AGN directly. The contributions to heating are roughly half and half, and a cooling catastrophe is averted.
The authors conclude the thermal conduction has an important role in preventing over-cooling but caution that some of their assumptions are optimistic; the true contribution may be lower. Future observations may provide better constraints, enabling firmer conclusions to be drawn.