skip to content

There will be a network outage, Sunday, April 26 from 13:00 EDT - Monday April 27 01:00 EDT. All services, web sites, and systems that require access to the internet will be affected by this work.

ROSAT Guest Observer Facility


D. E. Harris
Smithsonian Astrophysical Observatory,


AGN come in a wide variety of power and morphology at all wavebands. For the present purposes, we take AGN to mean ``non-thermal extragalactic sources''. We review a few topics of AGN physics which can be confronted with X-ray data by mentioning some examples from the biggest and brightest AGN. We then outline a proposal to observe 3CR radio sources with z < 0.3. In preparation for AXAF, the goal is to find sources which are resolved. For the current study, the goal is to discover more details about the gaseous environments of non-thermal sources as well as intrinsic properties such as the emission mechanism for X-rays.


We expect roughly an order of magnitude improvement in both resolution and sensitivity for AXAF compared to the ROSAT HRI. In addition, we will enjoy a much wider band with concomitant spectral resolution. In section 2 we give a few examples of what can be done with the ROSAT HRI for nearby, bright sources; each of these is unique but with AXAF we should be able to study their more distant counterparts; a vital test of what we have learned from, in many cases, only one example.

Since ROSAT no longer has spectral capabilities, we will not explore that aspect of AXAF's capabilities. We will also not discuss the fascinating inferences which can be deduced from timing studies of AGN: they continue to be the subject of many current ROSAT projects and XTE is designed to deal with most timescales.

Types of Physics

The study of Cygnus A with the ROSAT HRI has provided data on several different aspects of powerful radio galaxies and their environment (see figs. 1 and 2).

For the core emission, the ``hidden quasar'' hypothesis requires further testing. From higher energies, a power-law spectral component is found, but its location is not known directly. For the estimated amount of column density and the extrapolation of the observed power law to the ROSAT band, there should be no central component at the intensity observed. Explanations for the observed feature are (a) a density enhancement in the cooling flow which just happens to be within 1/2'' of the radio core; and (b) scattered X-rays from the nucleus which thus should actually appear from components just above and below the ``obscuring torus''. Obviously we need to find further examples and we need spectral resolution to obtain a better estimate of the obscuring matter. AXAF should be able to delineate the spectral cutoff by isolating the emission from just the core component.

For the radio hotspots, a synchrotron self-Compton model fits the current data well, yielding an independent estimate of the average magnetic field strength which is in agreement with the equipartition field for the case of no contribution to the energy density from relativistic protons. If this is substantiated in other sources, it will imply that the jet ``fluid'' is dominated by e+/e- rather than p+/e-. A competing model (the so-called ``Proton Induced Cascade'', PIC) requires much stronger magnetic fields and extremely high energy relativistic protons. A key test will be other sources since while both models work for Cygnus A, in other sources the size of the hotspots together with the B field may not be sufficient to ``hold'' the high energy protons required for the PIC model.

Much has been learned about the interaction of the radio jet and lobes with the ambient hot gas. In particular the lobe appears to form a relatively empty cavity in the gas and the bow shock increases the density and temperature of the ambient gas. These sorts of effects allow us to use pressure balance to estimate the pressures in the ambient gas, in the sheath of shocked gas between the bow shock and the lobe, and the internal pressure of the lobe. By these means, we can garner constraints on the average B field in the lobe, and by comparison with the equipartition field, can estimate contributions to the energy density which arise from components other than the observed relativistic electrons.

For lower luminosity radio galaxies, the jets are often of higher surface brightness than for FRII galaxies. So it is with M87 in the Virgo cluster. In this case, we are able to isolate X-ray emission from radio knots in the jet (see fig. 3), and from morphology, spectra, and time variability, constrain the X-ray emission mechanism: synchrotron is favored; inverse Compton is difficult, and thermal is unlikely, but not completely excluded (Biretta et al. 1991, Neumann et al. 1996).

There are also strong jets in quasars, e.g. 3C 273. The situation here is complicated because the X-ray emission does not mimic the radio and optical morphologies as is the case in M87.

For 4C 08.66, the distorted radio lobes of the northern component (a Q with z=0.62) suggest interaction with a gaseous atmosphere. There are faint optical candidates for a cluster of galaxies and, from an Einstein observation, there is weak X-ray emission. With AXAF, the search for cluster gas associated with quasars can proceed both spatially and spectrally.

For Seyferts, Wilson and co-workers have found a significant number of cases for which the x-ray structure is aligned with the radio and NLR. More examples are needed to test the hot-wind model.

The key aspect for almost all of the above examples is the requirement that the x-ray emission be resolved. To the extent that ROSAT HRI data can locate AGN which are resolved, they will be valuable in allowing AXAF to focus on the more productive AGN in terms of the range of physical problems that can be addressed.

A Suggestion: X-ray Emission from 3CR Sources

Radio galaxies are intimately connected to their environments. Their characteristic radio emission is powered by a jet from the central engine interacting with the surrounding medium. Thus, determining the environments of radio galaxies is an important question in extragalactic astronomy. This has been done using galaxy number counts (e.g. Hill and Lilly 1991), although this method does not always provide a good measure of the surrounding medium. For instance, Cygnus-A apparently lies in a small group of galaxies, but the surrounding X-ray gas is comparable to that of a rich cluster (Arnaud et al. 1984). Thus, X-ray images are required to define the environment. However, radio galaxies themselves are complex and diverse X-ray emitters. ROSAT images of a small sample of Fanaroff and Riley type 1 (FR1) radio galaxies show that their X-ray emission is a combination of nuclear emission and extended hot gas (Worrall and Birkinshaw 1995). To disentangle this complex emission requires high resolution X-ray images.

We propose to image all 3CR radio sources with |b| > 20 and 0.05 < z < 0.3. The 3CR catalog is flux-limited and essentially completely identified with measured redshifts (Spinrad et al. 1985, 1996). It is the standard for most statistical studies as well as observations by new instruments : HST is currently obtaining ``snapshot'' exposures on every extragalactic 3CR source. We will determine core X-ray fluxes from the HRI images and correlate these with optical and radio parameters as previously attempted (e.g. Fabbiano et al. 1984), with the confidence that our fluxes are uncontaminated by extended X-ray emission. We also will map the surrounding hot gas. Recently Barthel and Arnaud (1996) found that the presence of X-ray gas around a radio source increases the efficiency of the conversion of AGN power into total radio luminosity. Together with the suggestion from Crawford and Fabian (1995) that such X-ray gas is more common at higher redshifts, it is clear that to interpret radio luminosity functions and their evolution requires high resolution imaging observations. On a more speculative note, we also will investigate the suggestion by Nulsen and Fabian (1996) that cluster cooling flows play a fundamental role in the formation of the giant elliptical galaxies which host bright radio sources.

The proposed X-ray images also will serve as a ``finding list'' for future ROSAT deep images and AXAF observations. With a factor of ten improvement over the ROSAT HRI in sensitivity and spatial resolution, and even greater improvement in energy range and spectral resolution, AXAF can study many AGN to the detail that ROSAT achieved on Cygnus A (Harris et al. 1994, Carilli et al. 1994) and M87 (Neumann et al. 1996).

Feasibility of 3CR Observations

Although some radio emitters (e.g. quasars) can be very luminous in X-rays, most present epoch radio galaxies have typical X-ray luminosities of to a few  ergs s. At a redshift of 0.3, a source with a power law spectrum, typical galactic absorption, and a luminosity of 4 ergs s, has an X-ray flux of  ergs cm s. In a 20 ks HRI observation, this source provides 50 counts, sufficient for detection and to determine if the source is spatially extended. For low redshift 3CR sources at z=0.05, our limiting sensitivity is  ergs s.

Structure larger than the core of the HRI PRF will be resolved. At z=0.05, 5  corresponds to 7 kpc, and at z=0.3, to 28 kpc. Since the scale of hot gas is many tens of kpc's for galactic halos and hundreds of kpc's for an ICM, we can differentiate nuclear emission from surrounding hot gas. Although 10% of the AGN emission is scattered into a halo by the HRI, the relatively weak core X-ray emission in radio galaxies allows detection of extended X-ray gas from a surrounding group or cluster, whose luminosities typically exceed  ergs s. Thus a hot ICM, if present, will be detectable around each source.


C.L. Carilli, K. Arnaud, C. Jones, and W. Forman contributed to this report.


Arnaud, K.A. et al. 1984, MNRAS 211, 981.

Barthel, P.B., and Arnaud, K.A., 1996, Nature, submitted.

Biretta, J.A., Stern, C.P., and Harris, D.E.. 1991, AJ 101, 1632.

Carilli, C.L., Perley, R.A., and Harris, D.E. 1994, MNRAS 270, 173.

Crawford, C. and Fabian, A.C. 1995, MNRAS 273, 827.

Fabbiano, G. et al. 1984, ApJ, 277, 115.

Harris, D.E. Carilli, C.L., and Perley, R.A. 1994, Nature, 367, 713.

Hill, G.J. and Lilly, S.J. 1991, ApJ, 367, 1.

Neumann, M., Meisenheimer, K., and Roser, H.J. 1996, to be submitted to Astron. Astrophys

Nulsen, P., and Fabian, A., 1996, MNRAS, 277, 561.

Spinrad et al. 1985, PASP 97, 932.

Spinrad, H. 1996, private communication.

Worrall, D.M. and Birkinshaw, M. 1995, Roentgenstrahlung in the Universe, ed. Zimmermann et al., MPE.

Figure 1: Cygnus A with the ROSAT HRI after subtraction of a symmetrical King distribution. Visible are the radio hotspots, believed to be synchrotron self-Compton emission; the core component coincident with the radio nuclear source; the ``inner-double'' which is thought to be thermal emission from ambient gas between the bow shock and the radio lobe; and the cavities, which are the dark areas adjacent to the core where the X-ray emission is below that which would have been observed if the radio lobes were not present.

Figure 2: Same as fig 1, except the X-ray greyscale is reversed and radio contours are overlaid from a VLA 327 MHz map. Note how the ``inner double'' (X-ray bright features) appear to ``pinch'' the eastern lobe.

Figure 3: A simulated image of the AXAF observation of the M87 jet. The four brightest features are (from left to right): the core, knot D, knot A, and knot B.

Return to the Title Page
About this document ...

This document was generated using the LaTeX2HTML translator Version 96.1 (Feb 5, 1996) Copyright © 1993, 1994, 1995, 1996, Nikos Drakos, Computer Based Learning Unit, University of Leeds.

The command line arguments were:
latex2html -split 0 harris.tex.

The translation was initiated by Michael Arida on Fri Jun 7 11:20:43 EDT 1996

Michael Arida
Fri Jun 7 11:20:43 EDT 1996