X-ray Jets: A New Field of Study

Previous   Contents   Next

X-ray Jets: A New Field of Study

Chandra Initiates the Study of X-ray Jets

FIGURE 1: ROSAT (left) and Chandra (right) images of the quasar and jet in 3C 273. Both images are to the same angular scale, showing how the smaller point spread function separates the quasar from the jet, which brightens about 10" from the core. The solid diagonal line in the Chandra image is an artifact due to the ACIS readout streak.FIGURE 2: A composite image of the inner radio lobes and jet of M87. The red scale is from an 8 GHz VLA image and the green is from a 98ks 0.2 to 6keV Chandra image. The general green X-ray background comes from thermal X-ray emission of the hot gas pervading M87 and the Virgo cluster.

With the launch of the Chandra Observatory, the study of X-ray jets has developed as a new area of astronomy. Galactic X-ray jets have been studied in the symbiotic star R Aqr; the Crab, Vela, and several other pulsars; and the subset of X-ray binaries known as “microquasars” (as well as in the previously known SS 433 system). In this article we discuss only the extragalactic jets discovered by Chandra>. These occur in Fanaroff Riley Class I and II (FR I and FR II) radio sources and quasars, covering a wide range of luminosities.

Why has the study of X-ray jets emerged just since the advent of Chandra? Comparison of the ROSAT and Chandra images of the nearby quasar 3C 273, shown to the same scale in Figure 1, illustrates the 100-fold increase in imaging due to the 10-fold increase in linear resolution. Chandra resolves the jet in one dimension, concentrates the photons along the jet, and clearly separates the jet from the quasar core that is 300 times more luminous.

The opening of this field of research was portended shortly after launch when the initial focusing adjustment using what was expected to be an unresolved quasar, PKS0637-752, showed a 10" extension to the west (Schwartz et al. 2000, Chartas et al. 2000). After immediate concerns that this was due to a problem with pointing control or the aspect solution, it was quickly realized that this "stray" emission was associated with a radio jet, and in fact the mirror quality was so good that the jet did not compromise the use of this target to perform the focus adjustment!
What has drawn a cadre of researchers to work on X-ray jets is the situation that in spite of several decades of progress via radio observations, many of the basic questions of extragalactic jets remain unanswered. For example, we do not yet know the identity of the primary medium responsible for transferring energy from the environs of the supermassive black hole to the distant hotspots up to a Mpc away. Contenders are normal p+/e- plasma (either "cold" or relativistic), pair plasma, and Poynting flux. Although we are fairly confident that jets have bulk velocities which are relativistic because of the predominant one-sidedness, we do not know if characteristic velocities β = v/c are of order 0.5 to 0.9, or if values of β greater than 0.99 are common. Other features still hotly debated are how jets are launched, the nature of jet collimation, the internal velocity structure, the extent of entrainment and the mechanisms of deceleration.

What does the X-ray data tell us? The broad band spectral energy distribution (SED) from radio through IR and optical and to the X-rays, can help us deduce the emission mechanisms. Observations of X-rays show us a population of electrons of energy much higher or much lower (depending on the emission mechanism) than those which produce radio emission. In this article we will see how X-rays may be showing us the sites of particle acceleration, allowing observation of the low end of the relativistic electron spectrum, measuring the relativistic bulk velocities of the jets, and may even be serving as beacons to the most distant detectable activity in the early universe.

Extragalactic X-ray Jets

Many of the essential attributes of X-ray jets are illustrated in Figure 2. The non-thermal source is two sided (i.e. the two radio lobes shown in red) whereas the jet is visible on only one side, thus convincing us that relativistic beaming provides the Doppler favoritism to hide the counterjet and enhance the brightness of the jet coming towards us. Almost all of the currently known (more than 50) X-ray emitting jets are one sided. Both the unresolved nucleus of M87 (where the supermassive black hole responsible for the jets is located) and "knot A" (the bright area near the end of the straight part of the jet) are prominent in radio and X-rays. Note that the inner part of the jet has a larger ratio of X-ray to radio brightness than regions further from the nucleus, as shown by the change from white and blue to red. Such behavior is shown by many, though not all, X-ray jets.

X-ray jets are naturally divided into two classes following the radio morphology: those from FRI radio galaxies have projected sizes generally less than 10 kpc, and those from FRII radio galaxies and quasars have X-ray jet lengths usually greater than 10 kpc. The apparent X-ray luminosities also divide with FRI's mostly < 1042 erg s-1, and quasars typically > 1043 erg s-1. Part of this dichotomy can be ascribed to selection effects from finite angular resolution and from possible differences in bulk velocities between the two types of jets. There is an ever-increasing number of jet detections from the inner segments of FRI jets. They all align well with radio features, are mostly one sided, and are of low brightness, making spectral and variability studies difficult.

One problem of immediate interest to X-ray astronomers is the uncertainty as to which non-thermal emission process is responsible for the bulk of X-ray jet emission. Although there is general agreement that synchrotron emission predominates for the low power sources (FRI), the favored, but problematic, process for quasar jets is inverse Compton (IC) emission from relativistic electrons scattering photons from the cosmic microwave background (CMB) (Tavecchio et al. 2000, Celotti et al. 2001). In the following sections, we briefly give examples of nearby FRI jets, and then discuss the higher power radio sources detected by Chandra surveys of radio jets.

X-rays from Low power radio jets


One of the more unexpected and spectacular jet results from Chandra data was a large flare in the knot HST-1 0.8" from the nucleus (Figure 3). The X-ray and optical lightcurves are shown in Figure 4, and details can be found in Harris et al. (2006). The lightcurve shows an overall increase of a factor of 50, and a rapid decay comparable to the rise time. At its peak, HST-1 was brighter than the core and the rest of the jet combined. Once we process optical and radio data already observed, we will be able to evaluate any time delays between bands, separate light travel time ("geometry") effects from electron energy loss time scales, and possibly distinguish between expansion losses and radiation losses and derive a new estimate of the average magnetic field strength. We have been exceedingly lucky to have Chandra's resolution for this flare; with Einstein, ROSAT, or even XMM Newton, this event would most likely have been ascribed to the SMBH neighborhood instead of 60 to 200pc (depending on the projection angle) down the jet. The important lesson is that if one is observing an outburst from a blazer, what is seen may not be coming from a region close to the accretion disk!

Centaurus A

The X-ray jet of Cen A is the closest example of an extragalactic jet and the Chandra resolution of ~1 arcsec provides a spatial resolution of 17 pc, more than a factor of 4 better than for M87. As can be seen from Figure 6, the morphology is more complex than that of M87. Recent publications by Kraft et al. (2002) and Hardcastle et al. (2003) discuss the relation of the X-ray and radio morphology. Some of the radio features in the jet have moved, and some of the X-ray features have changed their intensity.

Quasar Jets

FIGURE 3: A Chandra image of the M87 jet, with radio contours overlaid. This image was constructed from over 20 5ks monitoring observations. The radio contours are from a VLA observation at 8 GHz; contours increase by factors of two. The nucleus is the feature to the lower left, followed by the brightest X-ray feature, HST-1. After the elongated knots D, E, & F is the bright knot A.

Following the discovery of the jet in PKS 0637-752, two groups undertook short exposure (5 to 10 ks) surveys to try to assess the frequency of X-ray jet emission. Both were based on flat spectrum AGN with radio jets, so one would expect that they are biased toward objects pointed in our direction. The team led by Rita Sambruna emphasized the coordinated search for optical emission using the Hubble Space Telescope (Sambruna et al.; 2002, 2004, 2006). A team led by Herman Marshall emphasized coordination with high resolution radio maps from the Australia Telescope Compact Array (Marshall et al. 2005; Gelbord et al. 2005; Schwartz et al. 2006). Both surveys detected jets at a rate of about 60%, and due to the short exposures it is reasonable to expect that the non-detections could be intrisically similar but with a smaller ratio of X-ray to radio luminosity. Where X-ray emission is detected in jets, the radiated X-ray power, νfν, exceeds that of the radio, which is a selection effect due to the X-ray sensitivity of the short observations. Figure 5, prepared by Herman Marshall and Jonathan Gelbord, shows a montage of 20 of the 22 jets detected in their survey of 37 quasars.
A key conclusion of both studies was that the SED did not allow an extrapolation of the radio spectrum through the optical and to the X-ray, even if one includes a high energy break due to loss mechanisms. So a single population of relativistic electrons could not be producing the X-ray emission via synchrotron emission. Yet, if the region emitting the radio synchrotron radiation were anywhere near the minimum energy condition (which is nearly equivalent to equipartition of energy between magnetic fields and relativistic particles), then there were not enough electrons to produce the X-rays via inverse Compton emission off of any possible photon source. The breakthrough idea by Tavecchio et al. (2000) and Celotti et al. (2001) noted that if the jet was in relativistic motion then the CMB photon density would be enhanced in the rest frame of the jet by the square of the bulk Lorentz factor. In this case, the lowest energy X-rays detected by Chandra imply that the electron spectrum extends down to a minimum energy at least as low as Lorentz factors of 50, if the jet bulk motion has Lorentz factor 10, or below 500 if the jet is not in relativistic motion.

FIGURE 4: Lightcurves for the knot HST-1 in the M87 jet. The first Chandra observation was taken in July 2000; the most recent was 5 Jan 2006. Additional UV data from the Hubble Space Telescope will be available later.

FIGURE 5: Twenty of the X-ray jets observed in the surveys by Marshall et al. False color gives the smoothed X-ray data, and the contours are the radio emission in either 1.4, 5, or 8.4 GHz. The straight dashed line shows the direction of the ACIS readout streak. The faintest yellow color corresponds to a single photon. The solid bar at the bottom gives the scale 5" in each panel.

With those assumptions and interpretations, one can calculate the rest frame magnetic field and the Doppler factor from spatially resolved elements along the jets. The numerical accuracy is limited by the many assumptions, including setting the Doppler factor equal to the Lorentz factor, but still restricts quantities to an allowed range of about a factor of two. Typically one finds magnetic fields in the range 5 to 25 micro-Gauss, and Doppler factors from 3 to 15. One can go on to calculate the maximum angle by which the jets could deviate from our line of sight to be in the range 4° to 20°, and hence from their measured angles on the sky we deduce their intrinsic lengths extend at least up to the order of a few million light years.

FIGURE 6: The inner portion of the jet in Centaurus A. The X-ray map was supplied by R. Kraft and is adaptively smoothed. The nuclear X-ray component is not visible because this map was made with an energy filter of 0.5-2 keV and the column density towards the core is ~10
23 cm-2. Overlayed are radio contours from the Australian Telescope Compact Array at 5 GHz, with contour levels increasing by factors of two.

Once we know the magnetic field and relativistic velocity, we can estimate the kinetic energy flux being transported out to those millions of light years. The calculation gives fluxes of at least 1045.5 to 1047.5 ergs s-1. These are often comparable to or in excess of the bolometric, isotropic radiation from the quasar itself. Such large energy transport, which must originate at the quasar black hole center, has important implications for the accretion mechanism and therefore the rate at which black holes might form in the early universe. Also, when such jets occur in clusters of galaxies they provide more than enough energy to power the cavities which have been revealed by Chandra observations, and to stop the cooling flow of cluster gas.

Implications of the IC mechanism

Is inverse Compton scattering on the CMB really the mechanism of X-ray jet emission? Although it is the simplest modification to producing the entire SED via synchrotron emission from a single particle population, adding only one parameter, many authors have shown that more complex sub-structure or additional electron populations, could be constructed to emit the X-rays. Indeed, in the case of the nearest object, Cen A, and the brightest quasar jet, 3C 273, there is clear evidence of spatial inhomogeneity and of independent electron populations.

Perhaps the most serious mystery for any mechanism is how the relative ratio of X-ray to radio emission is often constant, within a factor less than 2, along tens to hundreds of kpc of the jet. X-ray synchrotron emitting electrons have lifetimes at most tens of years, so there would have to be a delicate re-acceleration balance. Compton X-ray and radio synchrotron emitting electrons differ in energy by only two orders of magnitude, so in the inverse Compton CMB scenario it may be much more natural to maintain a contant emission ratio. However, that would still require the assumption that the magnetic field remains relatively constant along the jet. Basically we know that if the bulk Lorentz factor of the jet is 3 to 15, the emission of X-rays is likely to be via IC/CMB since the CMB energy density will exceed that of the magnetic field energy density. Furthermore we know the jets in bulk relativistic motion from observations of one-sidedness and superluminal expansion in the AGN cores. So it may boil down to determining exactly what is the distribution of bulk Lorentz factors displayed by quasars and where does deceleration of the jet occur.

It is most exciting to pursue the IC/CMB hypothesis because of the remarkable prediction that such an X-ray jet would maintain a nearly constant surface brightness no matter how far distant it is. This is because the IC emission depends on the energy density of the CMB photons, and that density increases with redshift z as (1+z)4, offsetting exactly (except for a bandwidth, or "K-correction", term) the (1+z)-4 cosmological diminution of surface brightness. In fact, this guarantees that at some redshift, which may be modest values of 2 to 4 for jets already observed, the IC/CMB mechanism must come to dominate. Therefore, current observational capabilities would allow us to detect X-ray jets at whatever large redshift they first form!

Dan Schwartz and Dan Harris


Celotti, A., Ghisellini, G., & Chiaberge, M. 2001, MNRAS 321, L1
Chartas, G. et al. 2000, ApJ 542, 655
Gelbord, J. M. et al. 2005, ApJL, 632, L75
Georganopoulos, M. & Kazanas, D. 2004, ApJ 604, L81
Hardcastle, M. J. et al. 2003, ApJ, 593, 169
Harris, D. E. et al. 2006, ApJ, accepted, astro-ph/0511755
Kraft, R. P. et al. 2002, ApJ, 569, 54
Marshall, H. L. et al. 2005, ApJS, 156, 13
Sambruna, R. M. et al., 2002, ApJ, 571, 206
Sambruna, R. M. et al., 2004, ApJ, 608, 698
Sambruna, R. et al., 2006, ApJ, accepted, astro-ph/0511459
Schwartz, D. A. et al. 2000, ApJ 540, L69
Schwartz, D. A. et al. 2006, ApJ 640 in press, astro-ph/0601632
Tavecchio, F., Maraschi, L., Sambruna, R.M., & Urry, C.M. 2000, ApJL 544, L23