Jet spectra in FR I radio galaxies: implications for particle acceleration

Abstract

We describe very accurate imaging of radio spectral index for the inner jets in three FR I radio galaxies. Where the jets first brighten, there is a remarkably small dispersion around a spectral index of 0.62. This is also the region where bright X-ray emission is detected. Further from the nucleus, the spectral index flattens slightly to 0.50 - 0.55 and X-ray emission, although still detectable, is fainter relative to the radio. The brightest X-ray emission from the jets is therefore not associated with the flattest radio spectra, but rather with some particle-acceleration process whose characteristic energy index is 2.24. The change in spectral index occurs roughly where our relativistic jet models require rapid deceleration. Flatter-spectrum edges can be seen where the jets are isolated from significant surrounding diffuse emission and we suggest that these are associated with shear.

Full text

Page 1

arXiv:0801.0154v1 [astro-ph] 30 Dec 2007
Extragalactic Jets: Theory and Observation from Radio to Gamma Ray
ASP Conference Series, Vol. **VOLUME**, **YEAR OF PUBLICATION**
T. A. Rector and D. S. De Young (eds.)
Jet spectra in FRI radio galaxies: implications for particle
acceleration
R.A. Laing
ESO, Karl-Schwarzschild-Straße 2, 85748 Garching-bei-M¨ unchen,
Germany
A.H. Bridle, W.D. Cotton
NRAO, 520 Edgemont Road, Charlottesville, VA 22903-2475, U.S.A.
D.M. Worrall, M. Birkinshaw
Department of Physics, University of Bristol, Tyndall Avenue, Bristol
BS8 1TL, U.K.
Abstract.
the inner jets in three FRI radio galaxies. Where the jets first brighten, there
is a remarkably small dispersion around a spectral index of 0.62. This is also
the region where bright X-ray emission is detected. Further from the nucleus,
the spectral index flattens slightly to 0.50 - 0.55 and X-ray emission, although
still detectable, is fainter relative to the radio. The brightest X-ray emission
from the jets is therefore not associated with the flattest radio spectra, but
rather with some particle-acceleration process whose characteristic energy index
is 2.24.The change in spectral index occurs roughly where our relativistic jet
models require rapid deceleration. Flatter-spectrum edges can be seen where
the jets are isolated from significant surrounding diffuse emission and we suggest
that these are associated with shear.
We describe very accurate imaging of radio spectral index for
1.Radio spectra: 3C31
The detection of X-ray synchrotron emission on kiloparsec scales in the bases of
low-luminosity (FRI) radio jets (e.g. Hardcastle et al. 2002, 2005; Worrall et al.
2007) requires distributed particle acceleration, consistent with the failure of
adiabatic models to reproduce the observed brightness profiles at radio wave-
lengths (Laing & Bridle 2004). The acceleration mechanism is not understood.
Here, we summarize the results of recent work on this problem, using accurate
radio spectral mapping (this section) and detailed radio – X-ray comparisons
(Section 2.).
In the course of our jet-modelling programme (Laing & Bridle 2007, and
references therein), we have accumulated high-resolution, multi-frequency radio
images of the bases of FRI jets and have used these to derive accurate spectral-
index distributions. We have studied NGC315 (Laing et al. 2006a) and 3C296
(Laing et al. 2006b); a third example, the nearby radio galaxy 3C31 (z = 0.0169;
Laing et al., in preparation), is shown in Fig. 1. There is no evidence for any
significant deviation from power-law spectra in 3C31 within 70arcsec of the nu-
1

Page 2

2
Figure 1.
(normalized by total intensity) at a resolution of 5.5arcsec. (b) and (c) Spec-
tral index, α from weighted least-squares, power-law fits to the total intensity.
(b) 5-frequency fit between 1365 and 4985MHz at a resolution of 5.5arcsec
FWHM. (c) 6-frequency fit between 1365 and 8440MHz at a resolution of
1.5arcsec FWHM for the inset area in panel (b).
Radio images of 3C31. (a) Sobel-filtered, mean L-band image
cleus. Out to ≈7arcsec in both jets, the spectral index at 1.5-arcsec resolution
is slightly steeper (?α? = 0.62)1than the average for the inner jets. From 7 –
50arcsec in both jets, the mean spectral index is in the range 0.55 – 0.57. Further
from the nucleus, there is a gradual spectral steepening. There are also slight,
but significant variations in spectral index across both jets within ≈30arcsec of
the nucleus in the sense that their West edges tend to have flatter spectra (?α? =
0.52 – 0.54; Fig. 1c). There is a clear spectral separation between the South jet
and the surrounding emission, matching the separation of these regions defined
by the sharpest brightness gradients (Figs 1a and b). This is particularly clear
where the jet first enters the diffuse emission. The spectral identity of the jet is
evidently maintained even after it bends abruptly about 2arcmin South of the
nucleus and remains until it terminates in a region of high brightness gradient.
The outer South jet in 3C31 is therefore a clear example of the type of spectral
structure noted in other FRI sources (Katz-Stone & Rudnick 1997; Laing et al.
2006b) wherein a flatter-spectrum ‘jet’ with a distinct spectral identity is super-
posed on a ‘sheath’ of steeper-spectrum emission.
1S(ν) ∝ ν−α

Page 3

3
Figure 2.
smoothed Chandra 0.8 – 5keV data (grey-scale) for NGC315 (Worrall et al.
2007). (a) Radio resolution 1.5arcsec. The area covered by panel (b) is shown
by the white box. (b) radio resolution 0.4 arcsec.
Overlays of VLA 5GHz images (contours) on adaptively
2.Radio – X-ray comparison: NGC315
Comparison of deep, high-resolution radio and X-ray images also provides clues
to particle-acceleration processes. Fig. 2 shows overlays of 5-GHz radio emission
at resolutions of 1.5 and 0.4arcsec FWHM on deep Chandra X-ray images for
the main jet in NGC315 (z = 0.01648; Worrall et al. 2007). The radio and X-
ray spectral indices integrated over the inner jet are 0.61 and 1.2, respectively,
consistent with synchrotron emission from a single population of relativistic
electrons in both wavebands. The radio jet is relatively faint and unresolved
in width out to about 4arcsec from the core, after which it brightens abruptly.
The most prominent feature of the radio brightness distribution between 4 and
12arcsec (Fig. 2b) is an oscillatory filament, whose nature is discussed in detail
by Worrall et al. (2007). The first X-ray enhancement which can unambiguously
associated with the jet is 3.6arcsec from the nucleus, where the radio emission
is still faint. From 5 – 7.5arcsec, there is particularly good morphological corre-
spondence between radio and X-ray images, with diffuse emission over the full
width of the jet as well as localized X-ray peaks following the radio ridge-line.
At larger distances, the general correspondence is still reasonable, but there
are differences of detail. The X-ray emission can be traced out to ≈30arcsec
(Fig. 2a), but becomes weaker with respect to the radio at ≈20arcsec. This is
most clearly illustrated by the profiles of X-ray and radio emission along the jet
at matched resolutions in in Fig. 3; an equivalent plot for 3C31 (Laing & Bridle
2004) is shown for comparison.
3. Discussion
These observations contribute to the developing picture of spectral variations
in FRI sources. Bright X-ray emission is detected close to the nucleus, in the
faint, well-collimated jet bases that precede the sudden radio brightening (e.g.
Fig. 3). There is approximate morphological correspondence between features in

Page 4

4
Figure 3.
the brighter jets of NGC315 (Worrall et al. 2007) and 3C31 (Laing & Bridle
2004). The locations of the brightening points (where the rest-frame radio
emission increases rapidly) and the start of rapid deceleration, as modelled
by Canvin et al. (2005) and Laing & Bridle (2002) are also indicated.
Profiles of radio (curve) and X-ray (points) flux density along
the radio and X-ray brightness distributions after the former brightens, although
there are differences on small scales (e.g. Fig. 2b). In contrast, there are no
systematic transverse variations in the X-ray/radio ratio within ≈30arcsec of
the nucleus in NGC315 (the best resolved case; Worrall et al. 2007). Particle
acceleration appears to be distributed throughout the jet volume, rather than
being exclusively associated with discrete knots or with the boundary. The ratio
of X-ray to radio emission decreases where our kinematic models show that the
jets start to decelerate from speeds of 0.8 – 0.9c (Fig. 3; Laing & Bridle 2002;
Hardcastle et al. 2002; Canvin et al. 2005; Worrall et al. 2007).
Where the jets first brighten and before they decelerate, there is a remark-
ably small dispersion around a radio spectral index of α = 0.62 in the three
sources we have studied in detail, as well as 3C66B (Fig. 1; Hardcastle et al.
2001; Laing et al. 2006a,b).The average is dominated by emission immedi-
ately after the point at which the jets first brighten. This is also the region
from which X-ray emission is detected from the main jets in all four sources
(Fig. 3; Hardcastle et al. 2001, 2005). The spectral index of the fainter emis-
sion close to the nucleus in 3C449 (Katz-Stone & Rudnick 1997), PKS1333−33
(Killeen, Bicknell & Ekers 1986) and 3C66B (Hardcastle et al. 2001) appears to
be slightly steeper than α = 0.62 although the uncertainties are larger. Further
from the nucleus, the spectra flatten slightly to α = 0.50 – 0.55, contrary to
any naive expectation from models in which electrons are accelerated at the
brightening point and suffer synchrotron losses as they propagate. X-ray emis-
sion is still detected from these regions, but at a lower level relative to the radio
(Fig. 3). The brightest X-ray emission from the jets is therefore not associ-
ated with the flattest radio spectra, but rather with some particle acceleration
process whose characteristic energy index is 2α + 1 = 2.24. A related result is
that an asymptotic low-frequency spectral index of 0.55 is common in FRI jets

Page 5

5
over larger areas than we consider here (Young et al. 2005). Flatter-spectrum
edges can be seen where the jets are isolated from significant surrounding diffuse
emission, most clearly in NGC315 (Laing et al. 2006a). Our kinematic models
(Laing & Bridle 2002; Canvin et al. 2005; Laing et al. 2006b) show that all of
the jets have substantial transverse velocity gradients and it is plausible that
the process that produces the flatter spectrum is associated with high shear
(Stawarz & Ostrowski 2002). In 3C31, the flatter-spectrum regions (Fig. 1c)
occur predominantly on the outer edges of bends, perhaps consistent with this
idea.
As well as a smooth steepening of the jet spectrum at larger distances from
the nucleus, as would be expected from synchrotron and adiabatic losses affecting
a homogeneous electron population, multiple spectral components are observed
(Fig. 1b). Jets appear to retain their identities even after entering regions of
diffuse emission and are clearly identifiable by their flatter spectra. They are
usually also separated from the surrounding emission by sharp brightness gra-
dients (Fig. 1a). This spine/sheath separation is observed in FRI sources with
bridges of emission extending back towards the nucleus (e.g. 3C296; Laing et al.
2006b) as well as tailed sources like 3C31 (Fig. 1). Although there is an overall
trend for the spectrum of the diffuse emission to steepen towards the nucleus in
bridges and away from it in tails (Parma et al. 1999), the variations in individual
objects are complex. The termination regions of jets in tailed FRI sources are
perhaps best regarded as bubbles which are continually fed with fresh relativistic
plasma by the jets and which in turn leak material into the tails. Their spectral
steepening would then be governed by a combination of continuous injection,
adiabatic, synchrotron and inverse Compton energy losses and escape.
The National Radio Astronomy Observatory is a fa-
cility of the National Science Foundation operated under cooperative agreement
by Associated Universities, Inc.
Acknowledgments.
References
Canvin, J.R., Laing, R.A., Bridle, A.H., Cotton, W.D., 2005, MNRAS, 363. 1223
Hardcastle, M.J., Birkinshaw, M., Worrall, D.M., 2001, MNRAS, 326 1499
Hardcastle, M.J., Worrall, D.M., Birkinshaw, M., Laing, R.A., Bridle, A.H., 2002, MN-
RAS, 334, 182
Hardcastle M.J., Worrall D.M., Birkinshaw M., Laing R.A., Bridle A.H., 2005, MNRAS,
358, 843
Katz-Stone, D.M., Rudnick, L., 1997, ApJ, 488, 146
Killeen N.E.B., Bicknell G.V., Ekers R.D., 1986, ApJ, 302, 306
Laing, R.A., Bridle, A.H., 2007, these proceedings
Laing, R.A., Bridle, A.H., 2002, MNRAS, 336, 328
Laing, R.A., Bridle, A.H., 2004, MNRAS, 348, 1459
Laing, R.A., Canvin, J.R., Cotton, W.D., Bridle, A.H., 2006a, MNRAS, 368, 48
Laing, R.A., Canvin, J.R., Bridle, A.H., Hardcastle, M.J., 2006b, MNRAS, 372, 510
Parma, P., Murgia, M., Morganti, R., Capetti, A., de Ruiter, H.R., Fanti, R., 1999,
A&A, 344, 7
Stawarz, L., Ostrowski, M., 2002, ApJ, 578, 763
Worrall, D. M., Birkinshaw, M., Laing, R. A., Cotton, W. D., Bridle, A. H., 2007,
MNRAS, 380, 2
Young A., Rudnick L., Katz D., Delaney T., Kassim N.E., Makishima K., 2005, ApJ,
626, 748