Extended Radio Sources in
Clusters of Galaxies
Elizabeth Blanton
University of Virginia
Collaborators
Craig Sarazin (UVa)
Brian McNamara (Ohio U.)
Michael Wise (MIT)
Noam Soker (U. Haifa at Oranim)
David Helfand (Columbia U.)
Michael Gregg, Robert Becker (UC Davis / LLNL)
Radio Sources in Clusters
• Radio sources and the X-ray emitting ICM have a
profound effect on each other, as seen with Chandra.
Radio sources blow bubbles in the ICM and the ICM
confines and distorts the radio lobes.
• 70% of cooling flow clusters contain central cD galaxies
with associated radio sources, and 20% of non-cooling
flow clusters have radio-bright central galaxies.
• The fact that radio sources are distorted by dense ICM can
be exploited to find high-z clusters using radio sources as
tracers.
Cooling Flows
• When the cooling time of gas t cool  T1/2/n (with
T=temp. and n=density) is shorter than the Hubble time,
or the time since the last major merger of the system, a
cooling flow will be set up
• Excess peak in central surface brightness above a b-model
profile
2 3β 1/ 2

 r 
5 M
dLcool 
kdT
I X (r)  1    
2 μm
  rc  
• Spatially resolved temperature gradients measured from
X-ray spectra
The Cooling Flow “Problem”
• Where does the cooling gas go?
• Central cD galaxies in cooling flows do emit blue light and
exhibit massive star formation, however the star formation
accounts for only ~ 1--10% of the expected gas derived
from the X-ray predictions (as measured from Einstein,
ROSAT, and ASCA)
• Both Chandra and XMM-Newton have revealed an apparent
lack of the expected quantities of cooler gas below about
kT < 1- 2 keV (~107 K)
• Radio sources are possible heaters (but not by strong
shocks).
Abell 2052
(Radio Contours, Burns [1990])
Chandra ACIS-S3, 36 ksec, z=0.0348, 1’’ = 1 kpc
(Blanton et al. [2001,2002])
Abell 262
(Radio Contours, Parma et al. [1986])
Chandra ACIS-S3, 28 ksec, z=0.0163, 1’’ = 0.5 kpc
(Blanton et al. [2003])
Source of Pressure in Bubbles?
• Pressure measured in X-ray shells is an order of mag.
higher than the equipartition pressure measured from the
radio source.
• Magnetic field may be larger than equipartition value?
• Contribution from low-energy relativistic electrons?
• Very hot, diffuse thermal gas?
• Chandra spectral limits on temp. of thermal component are
not very restrictive. Limit for A2052 from lack of Faraday
depolarization is kT > 20 keV. May be able to observe
very hot gas directly with XMM-Newton.
Buoyant Bubbles
A2597 (McNamara et al. 2001)
Perseus (Fabian et al. 2000)
• Buoyant bubbles have been seen in Perseus, A2597, and possibly A262.
• The density inside the radio cavities is much lower than the ambient gas,
so we expect the bubbles formed by radio sources to be buoyant.
• These bubbles may transport very hot gas and magnetic fields into the
ICM.
Total Spectrum / A2052
• Cooling flow model
• kTlow < 0.55 keV (0.43 +0.12/-0.43 keV)
• kThigh= 2.95 +0.069/-0.021 keV
  34 8 M /yr (approximately 3x lower than values
• M
-8
sun
measured from Einstein, ROSAT, and ASCA).
Can Radio Source Heating Offset
the Cooling Flow (A2052)?
• Total energy output of the radio source, including
the work done on compressing the gas is 5/2 PV =
11059 ergs
•
With kT = 3 keV and

5 M
 = 42 Msun/yr,
Lcool 
kT M
2 μm
Lcool = 3 x 1043 erg/s.
• If the repetition rate of the radio source is 108 yr
(as with Perseus and A2597), E/t for the radio
source is 3 x 1043 erg/s.
Radio Sources as Tracers of
Distant Clusters
• Since the appearance of radio sources is
affected by interaction with the ICM
(confinement, distortion), we can use radio
sources to locate distant clusters of galaxies
that would be difficult to find in optical (b.c.
of projection effects) or X-ray (b.c. of flux
limits) surveys.
Some types of radio sources are more
often found in clusters than others
FR II
(most often not in clusters)
FR I
FR I / WAT
(most often in clusters)
Wide-angle tails (WATs)
•Usually associated with cD or
central E galaxies.
•Usually not in cooling flows.
•May need a cluster merger to get
the ram pressure from the ICM to
bend the radio lobes.
•Should be good pointers to
clusters.
(Gomez et al. [1997], radio grayscale on
ROSAT contours of A562)
FIRST Bent Doubles
•384 sources visually selected from the
VLA FIRST survey
•Optically followed up (imaging and
spectroscopy)
•Low-z complete sample showed that
50% of BDs are found in clusters, as
revealed in the optical and X-ray
•High-z sample had an 80% success rate
using radio and optical images
•Highest-z object was observed in
optical and NIR, with spectra taken at
Keck.
Blanton et al. (2001)
z = 0.96 Cluster
Radio overlaid onto K-band
(KPNO 2.1m, IRIM) .
Radio exp. = 3 min. (VLA).
Optical R-band image.
Exp. = 2 hrs. (MDM 1.3m)
Blanton et al. (2003)
z = 0.96 Cluster
•10 galaxies spectroscopically
confirmed at z=0.96 with the Keck
II and LRIS.
•Velocity dispersion:
 ||  530190
90 km/s
•Expected LX,bol ~ couple 1044 erg/s
Keck II, K-band snapshot,
central ~290 kpc
•Scheduled with Chandra and
XMM-Newton
Conclusions
• Cluster central radio sources and X-ray-emitting ICM
profoundly affect one another: the ICM confines and distorts
radio lobes while the radio lobes blow bubbles in the ICM.
• Central radio sources are at least part of the solution to the
problem with the standard cooling flow model (with additional
contributions from, i.e., inhomogeneous abundances and
conduction). Energy is released into the ICM and dispersed
through buoyancy; strong shocks are not seen.
• Since distorted radio sources often attain their morphologies
from interaction with the ICM, they can be used as tracers for
distant clusters and groups of galaxies for studies of cosmology
and galaxy evolution.