Radio galaxies and feedback from AGN jets M.J. Hardcastle1 and J.H. Croston2 1Centre for Astrophysics Research, University of Hertfordshire, College Lane, Hatfield AL10 9AB 2School of Physical Sciences, The Open University, Walton Hall, Milton Keynes, MK7 6AA, UK Abstract We review current understanding of the population of radio galaxies and radio-loud quasars from an observational per- spective, focusing on their large-scale structures and dynamics. We discuss the physical conditions in radio galaxies, their fuelling and accretion modes, host galaxies and large-scale environments, and the role(s) they play as engines of feedback in the process of galaxy evolution. Finally we briefly summarise other astrophysical uses of radio galaxy populations, including the study of cosmic magnetism and cosmological applications, and discuss future prospects for advancing our understanding of the physics and feedback behaviour of radio galaxies. Keywords: 1. Introduction Key historical developments in observational RLAGN studies after the first surveys and optical Radio galaxies and radio-loud quasars (collectively identifications included the development in the 1970s radio-loud AGN, or RLAGN in this article) are active of high-resolution interferometers such as the 5-km galaxies characterized by radio emission driven by jets telescope and the NRAO Very Large Array (VLA), on scales from pc to Mpc. The characteristic radio emis- which allowed detailed study of radio structures as well sion is synchrotron emission: that is, it indicates the as optical identifications for the first time; progress in presence of magnetic fields and highly relativistic elec- very long baseline interferometry (VLBI), which has trons and/or positrons. Synchrotron emission may be given increasingly detailed views of the inner parts seen in other wavebands, and this enabled the detection of the jets; the advent of sensitive optical telescopes, of the first radio galaxy jet before the advent of radio as- including the Hubble Space Telescope, which allowed tronomy (Curtis, 1918) but it was only with the capabil- detailed studies of RLAGN host galaxies and environ- ities of radio interferometry (Ryle, 1952) that it became ments in the optical, as well as the study of optical possible to detect and image these objects in detail and synchrotron radiation; a greatly improved understand- in large numbers. As we will discuss in more detail be- ing of the nature of the active nuclei themselves, low, radio observations remain key to an understanding driven by a combination of broad-band photometry and of their origin, dynamics and energetics. spectroscopy; and the development of X-ray telescopes Radio images of some characteristic large-angular- with the sensitivity needed to image both the hot gas scale nearby RLAGN are shown in Fig.1. These show environments of RLAGN and the X-ray synchrotron the large-scale jets and lobes that are the defining feature and inverse-Compton emission from the large-scale of this type of object. The first observations capable of radio structures. Some of the understanding derived showing the radio jets (e.g., Northover, 1973) motivated from those observational advances is discussed in later arXiv:2003.06137v1 [astro-ph.HE] 13 Mar 2020 the development of the now standard ‘beam model’, in sections of this review. However, perhaps the most which collimated outflows from the active nucleus drive important development has been the realization that the extended structures (Longair et al., 1973; Scheuer, the energetic input of RLAGN can have a profound 1974; Blandford and Rees, 1974). Some authors have effect on both the galaxies that they inhabit and their used ‘beam’ to refer to the outflows themselves and ‘jet’ large-scale environment, heating the hot gas that to refer to their observational manifestations, but in this surrounds them and preventing it from cooling and review we use ‘jet’ interchangeably for both, relying on forming stars; this process is an important member of a context to make the distinction clear where it is needed. family of processes that have come to be called ‘AGN Preprint submitted to New Astronomy Reviews March 16, 2020 Figure 1: Radio images of nearby radio galaxies showing a range of morphologies: top row are the Fanaroff-Riley class I source 3C 31 (left) and the Fanaroff-Riley class II source 3C 98; middle row are the wide-angle tail source 3C 465 (left) and narrow-angle tail / head-tail source NGC 6109 (right); and bottom row are double-double radio galaxy 3C 219 (left), and core-restarting radio galaxy 3C 315 (right). Compact ‘cores’ may be seen in all images, well-collimated jets are visible in 3C 31, 3C 98 and 3C 465, and hotspots in 3C 98, 3C 465 and 3C 219. 3C 31 image kindly provided by Robert Laing; 3C 98 image from the online ‘Atlas of DRAGNS’ at http://www.jb.man.ac.uk/atlas/; 3C 465 image courtesy of Emmanuel Bempong-Manful; 3C 219 image from Clarke et al.(1992); NGC 6109 and 3C 315 from unpublished LOFAR data. 2 feedback’. This understanding of the importance of in Fig.1 are of order 1 nT for a powerful source, lead- RLAGN in galaxy formation and evolution, derived ing to energy densities of order a few ×10−13 J m−1 and both from X-ray observations and from numerical total energies of order 1054 J or more — which would modelling of the formation and evolution of galaxies, require the direct conversion to energy of millions of has moved RLAGN studies into the mainstream of solar masses of matter. If there is any departure from extragalactic astrophysics. In this chapter we will the minimum-energy assumptions, these numbers will therefore also discuss how observations and models of be larger — and possibly very much larger if, e.g., there RLAGN constrain the ‘feedback’ processes that may be are large departures from equipartition or if the energy operating. density in the lobes is dominated by non-radiating par- Throughout the review we adopt the convention that ticles. γ represents the (random) Lorentz factor of an individ- It can be seen that estimates of the energetics of the ual electron and Γ represents the bulk Lorentz factor radio-emitting structures depend strongly on the char- due to directed motion. Luminosities and physical sizes acteristic magnetic field strength. This cannot be es- quoted are based on a standard concordance cosmology timated directly from observations of synchrotron in- −1 −1 with H0 = 70 km s Mpc . tensity. Synchrotron emission is strongly polarized — the fractional polarization can be ∼ 70% for a uniform- field region with a power-law spectrum, and even higher 2. Observational approaches where the spectrum is exponentially cutting off — but the polarization does not tell us about the field strength In this section we provide an overview of the observa- either, although it does give an emission-weighted es- tional methods that provide us with constraints on radio timate of the magnetic field direction along a particu- galaxy physics. lar line of sight, if Faraday rotation effects may be ne- glected (see below). For optically thin radio emission, 2.1. Radio the only way of directly estimating the magnetic field strength is to use additional observations, for example The fact that the radio emission is synchrotron emis- observations of inverse-Compton emission, discussed sion was realised early on from its polarization and below (Section 2.4). spectrum (Baade, 1956; Burbidge, 1956) and, together Faraday rotation is an effect caused by the propaga- with the optical identification of these objects with rel- tion of electromagnetic radiation through a magnetised, atively distant galaxies (see Section 2.2), turned out to ionized medium. The polarization angle rotates due to imply very large energies stored in the extended struc- a difference in propagation speed for the two circularly tures. The details of the radiation mechanisms can be polarized components of the electromagnetic wave. The found in e.g., Longair(2010) or Rybicki and Lightman change in angle is dependent on frequency, and the ro- (1979). The point that we wish to emphasise here is that tation measure — the strength of the rotation effect — the energy density in the radiating electrons and field, depends on the magnetic field strength and electron den- U, can be written in terms of the volume emissivity J(ν) sity of the intervening material (e.g. Cioffi and Jones, and the magnetic field strength B: for a power-law dis- 1980). For a single line of sight through a Faraday- tribution of electron energies with energy index p, we active medium towards a background polarized source find 2 the measured polarization angle χ is given by (Burn, − p+1 B U = kJ(ν)B 2 + (1) 1966): 2µ0 2 χ = χ0 + φλ (2) where k is a constant incorporating physical constants, where χ is the intrinsic polarization angle and φ is the observing frequency, and the integral over electron 0 given by energies. Clearly eq.1 has a minimum at some value Z d of B, and by solving for the minimum and computing φ = K neB · dS (3) the minimum energy density, we can get both an esti- 0 mate of a characteristic field strength and a lower limit in which K is a constant with value (in SI units) 2:63 × on the energy responsible for a given region of a ra- 10−13 T−1. Clearly in general different lines of sight, dio source. The minimum-energy condition turns out even within a given telescope beam, will have differ- to be close to the equipartition energy, Ue = UB, but the ent values of φ. However, observationally, it is of- important conclusion is that the minimum-energy field ten the case that the rotation measure RM = dχ/dλ2 strengths for 100-kpc-scale lobes such as those shown shows smooth behaviour across a source, and in this 3 case rotation measure observations can be used to es- scales might be arcminutes, corresponding to hundreds timate magnetic field strength external to the source of kpc to Mpc.