Riccardo Giovanelli

A large fraction of the mass of rich clusters of galaxies that emits electromagnetic radiation is in the form of a diffuse (density 10-3 atoms cm-3) intracluster medium (ICM), macroscopically at rest in the cluster gravitational potential well. This gas is hot (108 K) and especially conspicuous in x-rays, which are produced by thermal bremsstrahlung.

Individual galaxies in clusters are, of course, radio emitters, as are those outside of clusters; however, the location of cluster galaxies in a high-density environment brings about peculiar characteristics which are usually attributed to their interaction with the ICM or to the exceptional character of the objects dwelling in the cluster cores (e.g., cD galaxies). The radio properties of individual galaxies, including radio galaxies associated with cDs, are described elsewhere in this volume. Here we discuss the characteristics of the radio emission arising from the interaction of radio sources with the ICM, as well as the global radio properties associated with the ICM itself.


A handful of clusters of galaxies are known to harbor diffuse sources of radio emission which cannot be attributed to single galaxies in the cluster. The best studied of these sources is associated with the Coma cluster. Cluster halo sources are concentric with the distribution of galaxies, have very low surface brightness, and are thus difficult to distinguish from a population of distributed discrete, weak sources, or from the outskirts of strong, centrally located radio galaxies. The mean properties of halo sources are derived from a few clusters where clear identification and distinction from other cluster radio sources have been made. Halo sources appear to have steep spectral index (i.e., the radio flux drops with increasing frequency nu with a power larger than -1, typically as nu-1.2), a moderately high radio luminosity (about 1031-1032 erg s-1 Hz-1 at a wavelength of 21 cm or 1041 erg s-1 over the whole radio spectrum) and large size (diameter of about 1 Mpc). Clusters that contain radio halo sources have very high x-ray luminosities; their galaxian population appears compact and dynamically evolved and is characterized by a high velocity dispersion. These clusters have a rich, widely distributed ICM and lack a single dominating central cD galaxy.

Relic sources are also found in clusters. Their properties resemble those of halo sources: They are extended, diffuse, without an optical counterpart, and have steep radio spectra. They are, however, not as extended as halo sources and are not centrally located in the cluster. It is postulated that relic radio sources are ejecta of now-quiescent radio galaxies that have moved away from the scene. It has also been suggested that a cluster halo source is produced by the collective radio emission of a superposition of relic radio sources in the cluster.

The steep nonthermal spectra of these radio sources suggests that the emission process is synchrotron radiation by a cluster population of relativistic electrons in an intracluster magnetic field that needs not be stronger than 1 µG.


A ``standard'' radio galaxy has a compact radio source associated with the active nucleus of an optical galaxy; part of the radio emission is observed in the form of radio lobes, extended regions of emission diametrically opposed with respect to and quite distant from the compact radio source. Narrow jets originating in the central compact source extend out to the lobes and are the conduits by which energy is carried from an active region deep in the core of the central source to the lobes. The energy is carried by a mixture of relativistic and thermal gas, which outlines the jets. In clusters of galaxies, the rapid motion of the central source and the interaction of the gas outflow with the ICM are thought to be responsible for the observed spectacular departures from alingment of the radio lobes. As the angle between the direction from the central source to the lobes progressively departs from 180°, the terminology describing the radio source morphology varies from wide-angle tails (WATs) to narrow-angle tails (NATs) to head-tail sources, the latter of striking cometary appearance. Figure 1 illustrates an archetype of the NAT category, the source associated with the elliptical galaxy NGC 1265 in the Perseus cluster. The conventional model to explain head-tail or NAT sources pictures them as conventional radio galaxies moving at high velocity through the ICM. As plasma beams are quasicontinuously ejected from the active nucleus of the galaxy, they are bent either by direct ram pressure or, if there is a significant interstellar medium in the galaxy, the latter forms a cocoon around the plasma beams and the pressure gradients created by the motion of the galaxy through the ICM ultimately cause the bending of the plasma beams. The ram pressure acting on the ejected gas is P = nicm vg2, where nicm is the ICM density and vg the velocity of the galaxy relative to it. While the basic idea of plasma ejecta interacting with the ICM is still the main driver in explanations of the peculiar cluster radio source morphology, many aspects of the interpretation are still quite uncertain, and are related to the poorly known circumstances associated with the process of ejection, with the difficulty of deprojecting two-dimensional images into three-dimensional representations and with the effects of inhomogeneities in the ICM.

Figure 1

Figure 1. Total intensity map of the radio source in NGC 1265 in the Perseus cluster. The image was obtained by C. O'Dea and F. Owen at a frequency of 1413 MHz with the Very Large Array radio telescope of the National Radio Astronomy Observatory.


Normal spiral galaxies are characterized by disks containing a fair fraction of their mass in the form of interstellar gas. The neutral component of that gas, primarily hydrogen, extends much farther out from the center of the galaxy than the stellar component. This gas is endowed with a large specific angular momentum, which supports it in equilibrium to the outer reaches of the galaxy's potential well. As a result, this gas is quite vulnerable to external dynamical and thermal perturbations. As a spiral galaxy travels at high speed through the denser regions of a cluster, its interaction with the ICM can sweep a large fraction of its interstellar gas off its outer disk. The mechanics of the interaction are difficult to model; the sweeping efficiency depends on the parameters that determine the ICM ram pressure, the angle between the galaxy's vector of motion and that defining the inclination of its disk, the degree of clumpiness of the interstellar gas, the galaxy's total mass, and the degree of thermal shielding provided by its gaseous corona. Thermal conduction between the cold interstellar gas and the hot ICM can, in fact, play as important a role as ram pressure in the sweeping episode.

Interstellar neutral hydrogen is easily detected in external galaxies by means of its line emission at 21cm. Observations of the HI line have shown that a spiral galaxy moving through the inner parts of a dense cluster core can have most of its interstellar neutral hydrogen removed from the disk. The total neutral hydrogen mass of a normal spiral galaxy ranges between 1 and 30 billion solar masses. A single high-velocity passage through the cluster core can free much of that gas from the galaxy, transferring it to the ICM. A spiral galaxy is called HI-deficient when its HI mass is at least 2-3 times smaller than that found on the average in noncluster spirals of the same type and size. On the other hand, radio observation of the 2.6-mm line of CO in HI-deficient spirals indicate that the molecular gas content of those galaxies is little affected by the sweeping events that deplete its neutral gas reservoir. The explanation of this difference may be in the different degree of clumpiness and radial distribution in the galactic disk of the two components of the interstellar medium: The diffuse, more peripheral HI presents a larger cross-section to ram pressure forces than the clumpy, more centrally located molecular gas and it is less tightly bound to the galaxy than the molecular gas. It has been suggested that the removal of the neutral component of the interstellar medium has a delayed, quenching effect on the star-formation rate in the swept galaxies. This ``interruption of fertility'' does eventually result in an aging of the stellar population, an obliteration of the spiral pattern, and the conversion from spiral to lenticular morphology. It has been proposed that the marked morphological segregation of galaxy types observed in cluster cores, whereby ellipticals and lenticulars dominate the core population, may in part be due to this ``secular,'' environment-driven process.

Disks of spiral galaxies also harbor a diffuse population of relativistic electrons, which are produced continuously by localized stellar sources. The interaction of these electrons with a galactic magnetic field, supported by the interstellar gas, produces synchrotron radiation, providing the main contribution to the radio emission of normal spiral galaxies. The ram pressure produced by the transit of a spiral galaxy through the ICM can produce large-scale displacements of its interstellar medium, which will carry along field and relativistic electrons. Thus, images of radio continuum disks can appear displaced from the stellar disks of galaxies moving through the ICM. Figure 2 illustrates the Point for the galaxy NGC 4438, which is known to be moving at high speed through the core of the Virgo cluster.

Figure 2

Figure 2. Contours of radio emission of the displaced disk of NGC 4438, superimposed on an optical image of the galaxy. Also in the picture (top) is the companion galaxy NGC 4435. The radio observations where made with the Westerbork Synthesis Radio Telescope at 1.4 GHz by C. Kotanyi and R. Ekers.


The ICM radiates mainly in the x-ray energy band, via thermal bremsstrahlung. This process provides the primary means of cooling for the intracluster gas. The cooling time of the ICM, that is, the time necessary for the temperature of the gas to drop significantly from its equilibrium value, can be approximated by

tcool = 8.5 x 1010 [nicm / 10-3 cm-3]-1 [Ticm / 108 K]1/2 yr,

where nicm and Ticm, are the density and temperature of the ICM. In most of the ICM, cool is comparable with or longer than the age of the universe, and therefore no substantial cooling is expected to have occurred. However, in the central parts of the cluster, where the densities are higher, cooling may be important. The cooling gas will then flow towards the center of the cluster and coalesce onto massive galaxies sitting at the bottom of the gravitational potential well. Once in the cooling flow, the gas may rapidly cool to temperatures where hydrogen recombination can occur, and the 21-cm line of the neutral hydrogen may then be detected in emission, or in absorption if the cluster core harbors a strong radio continuum source. The likelihood of the presence of a cooling flow in a cluster is estimated by the inspection of the observed x-ray parameters of the ICM at the cluster core, which yield an indication of nicm and of Ticm. Early searches for cooling flows in the 21-cm line led to negative results. More recent measurements with the Arecibo telescope, however, have provided very encouraging evidence of this interesting phenomenon.


The ICM is not completely transparent to radio waves. In fact, the optical depth for scattering of microwave photons of a gas with electron density ne is

tau = integ sigmaT ne dl,

where the integration is performed along the line of sight through the ICM and sigmaT is the Thompson electron scattering cross-section, which is on the order of 10-28 m2. For a typical cluster, tau is between 0.001 and 0.01. Thus, a small fraction of the photons from any radio source behind a cluster will be scattered off the line of sight by the cluster's ICM. The cosmic microwave background radiation (MBR) is a nearly isotropic bath of radio photons well described by a blackbody temperature of 2.7 K, a ``fossil'' relic of the Big Bang. Because of its nearly isotropic character, MBR photons can be scattered into the line of sight from any direction. Because they are much ``cooler'' than the ICM electrons (2.7 versus 108 K), they can gain energy, that is, be ``heated'' by the interaction with the ICM electrons (the process is known as inverse Compton scattering). The net result of the transit of the MBR photons through a cluster is that the blackbody curve that describes their spectrum is slightly shifted. The shift of the spectral energy density of the MBR F (nu) at the frequency nu is described by

DeltaF (nu) = chinu d / dnu {nu4 d / dnu [nu-3 Fbb (nu)]},

where chi = (kTicm tau / me c2), k is the Boltzmann constant, me the electron's mass, c the speed of light, and Fbb is the blackbody curve at 2.7 K. Because the number of MBR photons is conserved in transit through the ICM, the heating of the MBR by this process translates into a decrease of the number of photons on the low-frequency side of the blackbody curve, and an increase on the high-frequency side. Because the measurements are generally done on the low-frequency side (the Rayleigh-Jeans domain) of the blackbody curve, the effect of the cluster on that spectral region is an at first sight paradoxical apparent cooling of the MBR: The MBR flux in the line of sight to the cluster is somewhat lower than in the directions around it. However, if the energy budget is estimated over all frequencies, on both sides of the peak of the energy distribution curve, it is found that the total energy carried by MBR photons after transit through the cluster has increased.

This effect was first proposed by Ya.B.Zeldovich and Rashid A. Sunyaev in 1972. The shift for the densest and hottest clusters can be described as a temperature shift in the Rayleigh-Jeans part of the MBR spectrum of about -0.1 mK. This is a very tiny effect, a DeltaT / Tmbr on the order of -10-4. It has however been successfully measured in a few clusters at frequencies of about 20 GHz. Testing the Sunyaev-Zeldovich effect has important cosmological applications. Among them is the possibility of obtaining an estimate of the Hubble constant in a manner completely independent from that of more traditional methods and the uncertainties associated with local calibrators. Although these measurements are difficult and still relatively inaccurate, the potential of this line of radio astronomical work is exceptionally attractive.

Additional Reading
  1. Fabian, A.C., ed. (1987). Cooling Flows in Clusters and Galaxies. Reidel, Dordrecht.
  2. Haynes, M.P., Giovanelli, R., and Chincarini, G.L. (1984). Enviromental effects on the HI content of galaxies. Ann. Rev. Astron. Ap. 22 445.
  3. O'Dea, C. and Owen, F.N. (1986). Astrophysical implications of the multifrequency VLA observations of NGC 1265. Ap. J. 301 841
  4. O'Dea, C. and Uson, J.M., eds. (1986). Radio Continuum Processes in Cluster of Galaxies. NRAO, Green Bank.
  5. Sarazin, C.L. (1986). X-ray emission from clusters of galaxies. Rev. Mod. Phys. 58 1.
  6. Uson, J.M. and Wilkinson, D.T. (1988). The microwave background radiation. In Galactic and Extragalactic Radio Astronomy, G.L. Verschuur and K.I. Kellermann, eds. Springer-Verlag, Berlin.
  7. See also Background Radiation, Microwave; Clusters of Galaxies; Clusters of Galaxies, X-Ray Observations; Galaxies, Radio Emission; Intracluster Medium.