Recent observations show that the cooling flows in the central regions of galaxy clusters are highly suppressed. Observed active galactic nuclei (AGN

We investigate the effects of AGN feedback on the cosmological evolution of an isolated elliptical galaxy by performing two-dimensional high-resolution hydrodynamical numerical simulations. The inner boundary of the simulation is chosen so that the Bondi radius is resolved. Compared to previous works, the two accretion modes—namely, hot and cold, which correspond to different accretion rates and have different radiation and wind outputs—are carefully discriminated, and the feedback effects by radiation and wind in each mode are taken into account. The most updated AGN physics, including the descriptions of radiation and wind from the hot accretion flows and wind from cold accretion disks, are adopted. Physical processes like star formation and SNe Ia and II are taken into account. We study the AGN light curve, typical AGN lifetime, growth of the black hole mass, AGN duty cycle, star formation, and X-ray surface brightness of the galaxy. We compare our simulation results with observations and find general consistency. Comparisons with previous simulation works find significant differences, indicating the importance of AGN physics. The respective roles of radiation and wind feedback are examined, and it is found that they are different for different problems of interest, such as AGN luminosity and star formation. We find that it is hard to neglect any of them, so we suggest using the names “cold feedback mode” and “hot feedback mode” to replace the currently used ones.

The observed bimodality in radio luminosity in galaxy clusters is puzzling: cosmic rays (CRs) accelerated by structure formation shocks are expected to indiscriminately produce bright radio haloes in every cluster. We investigate the possibility that CR streaming in the intracluster medium (ICM) can ‘switch off’ hadronically induced radio and gamma-ray emission. For self-confined CRs, this depends on the source of magnetohydrodynamic wave damping: if only non-linear Landau damping operates, then CRs stream on the slow Alfvénic time-scale, but if turbulent wave damping operates, super-Alfvénic streaming is possible. As turbulence increases, it promotes outward streaming more than it enables inward turbulent advection. Curiously, the CR flux is independent of ∇f (as long as it is non-zero) and depends only on plasma parameters; this enables radio haloes with flat inferred CR profiles to turn off. We perform 1D time-dependent calculations of a radio mini-halo (Perseus) and giant radio halo (Coma) and find that both diminish in radio luminosity by an order of magnitude in several hundred Myr, given plausible estimates for the magnetic field in the outskirts of the cluster. Due to the energy dependence of CR streaming, spectral curvature develops, and radio haloes turn off more slowly at low frequencies – properties consistent with observations. Similarly, CR streaming rapidly turns off gamma-ray emission at the high energies probed by Cherenkov telescopes, but not at the low energies probed by Fermi. CR mediated wave heating of the ICM is unaffected, as it is dominated by ∼GeV CRs which stream Alfvénically.

It is widely accepted that feedback from active galactic nuclei (AGN) plays a key role in the evolution of gas in groups and clusters of galaxies. Unequivocal evidence comes from quasi-spherical X-ray cavities observed near cluster centers having sizes ranging from a few to tens of kpc, some containing radio emission. Cavities apparently evolve from the interaction of AGN jets with the intracluster medium (ICM). However, in numerical simulations it has been difficult to create such fat cavities from narrow jets. Ultra-hot thermal jets dominated by kinetic energy typically penetrate deep into the ICM, forming radially elongated cavities at large radii unlike those observed. Here, we study very light jets dominated energetically by relativistic cosmic rays (CRs) with axisymmetric hydrodynamic simulations, investigating the jet evolution both when they are active and when they are later turned off. We find that, when the thermal gas density in a CR-dominated jet is sufficiently low, the jet has a correspondingly low inertia, and thus decelerates quickly in the ICM. Furthermore, CR pressure causes the jet to expand laterally, encounter and displace more decelerating ICM gas, naturally producing fat cavities near cluster centers similar to those observed. Our calculations of cavity formation imply that AGN jets responsible for creating fat X-ray cavities (radio bubbles) are very light, and dominated by CRs. This scenario is consistent with radio observations of Fanaroff-Riley I jets that appear to decelerate rapidly, produce strong synchrotron emission and expand typically at distances of a few kpc from the central AGN.