Acknowledgements

1
Introduction. Questions of cosmic evolution
require observations of sources with the largest
look-back times to compare with those of more
recent epochs. At redshifts of 4 to 5 we start to
see clues to the origin and formation of these
sources and their central engines, as well as how
their properties change over cosmic time. At high
redshift the environment of a radio source is
quite different from that of a low redshift
source. The density of the intergalactic medium
scales with redshift as (1 z)3 and the energy
density of the Cosmic Microwave Background scales
as (1 z)4. For a z 5 quasar compared to a z
1 quasar, the IGM density is 27 times higher (in
simple models), and the energy density of the CMB
is 81 times higher. These should have profound
and testable consequences for the morphology and
x-ray emission of high redshift sources. A high
redshift quasar sample. We have constructed a
sample of high redshift quasars in the simplest
possible way. It includes every quasar in the
NASA/IPAC Extragalactic Database (NED) with a
redshift greater than 2.5, and a radio flux
density greater than 70 mJy at 1.4 GHz or 4.9
GHz. The radio fluxes were taken from NVSS at 1.4
GHz, and from the 87GB survey at 4.9 GHz. The
sample is also restricted to the area covered by
both FIRST (Becker, White Helfand 1995) and
SDSS (Abazajian et al 2003) and contains 134
quasars with redshifts between 2.5 and 4.72.
There are therefore excellent optical data on
every source, as well as wide field radio images.
We have been able to image 36 sources deeply
enough from archival VLA data, and three were
observed in program AC755. We show those results
here. The remaining sources are the subject of a
current VLA proposal. This is not a "complete
sample" in the usual sense, but it is certainly a
representative sample. The selection criteria are
now well-defined enough that it will be possible
to construct smaller sub-samples that are
properly complete, or whose selection effects can
be modeled in Monte Carlo simulations (c.f.
Wardle Aaron 1997).
The most important selection effects are due to
beaming and to the increasing rest frame emission
frequency at increasing redshift (e.g. Cohen
1988), and the inevitable youthfulness of high
redshift radio sources (Blundell Rawlings
1999). These effects must be accounted for
carefully when comparing the radio structures of
high redshift and low redshift samples. Scientif
ic goals. (1) The first goal is simply to image
these sources with arcsecond resolution and
reasonably high (10001) dynamic range, in order
to take a fairly deep look at the high redshift
radio universe. The images will be made publicly
available as quickly as possible so that other
astronomers can use the information to plan
follow-up observations of particular sources with
Chandra and HST. Here, timely dissemination is
crucial, because both space observatories have
limited lifetimes at this point. (2) The most
viable mechanism for the x-ray emission from many
kiloparsec scale radio jets is inverse Compton
scattering of the photons of the Cosmic Microwave
Background radiation (the IC/CMB model e.g.
Tavecchio et al 2000). Because the energy density
of the CMB increases with redshift as (1z)4, the
ratio of the x-ray luminosity to the radio
luminosity of the jet also scales as (1z)4,
making high redshift radio jets potentially
powerful x-ray sources. The model also requires
moderate beaming (i.e. at least mildly
relativistic speeds on kiloparsec scales, and
that the observer is within the beaming cone).
The best candidate jets are therefore
flat-spectrum core-dominated sources (beamed in
our direction) that reveal arcsecond-scale jets
on moderately high dynamic range imaging. We
estimate that we will find several tens of such
jets in this sample, and the best cases will be
proposed as Chandra and then HST targets. The HST
observations are required because it is a low
optical flux that can rule out a simple
synchrotron spectrum stretching from radio to
x-ray wavelengths (e.g. Sambruna et al 2004).
This is a strong test of the IC/CMB mechanism,
and it will also yield valuable information on
the speeds of jets on kiloparsec scales. The
radio/X-ray jets in two z4 quasars have been
studied in this way (Cheung 2004 Cheung et al.
2006).
(3) In an often quoted paper, Barthel and Miley
(1988) showed that high redshift (z gt 1.5)
quasars were smaller, and had a more bent,
distorted radio structure than quasars at lower
redshifts. They attributed this to the effects of
a denser, clumpier intergalactic medium at higher
redshifts. But in a series of papers Neff and her
collaborators (Neff Hutchings 1990, and
references therein) found these effects far less
pronounced than did Barthel and Miley, and
attributed much of them to a dependence on
luminosity rather than redshift. The Barthel
Miley high redshift sample contained 80 quasars
with z gt 1.5, but only 6 had redshifts gt 2.5. The
Neff et al high redshift sample contained 58
sources with z gt 2.0, but only 10 had z gt 3.0.
We note that our whole sample of 134 sources has
z gt 2.5, and 55 of them are at z gt 3.0. This
makes disentangling luminosity dependence from
redshift dependence much easier, and the large
number of very high redshift sources in our
sample should make all epoch-dependent properties
much more apparent. The Barthel Miley and the
Neff et al high redshift samples in effect
become intermediate redshift samples to compare
our results against. (4) While a higher IGM
density at higher redshift is expected to lead to
smaller overall linear sizes, it will not by
itself cause bends and distortions. Those suggest
any or all of the following a clumpy IGM,
significant host galaxy velocities with respect
to the IGM, and changing axes of the central
engines. This is the natural scenario of galaxy
and cluster formation through mergers at early
epochs. Bent and distorted radio structure may
therefore be one of the best signatures of very
young systems that are still in the process of
formation (see Overzier, Miley Ford 2007, for
such a system at z 2.2, and Overzier et al
2008, for another at z 4.1). We suspect that
our observations will yield many such systems,
which will then be excellent candidates for
follow-up observations with HST and Chandra to
study the emission line gas, the hot IGM and the
properties of the other objects in the field.
(5) It is now thought that the formation and
evolution of supermassive black holes (SMBHs) in
AGN is intimately linked to the formation and
evolution of their host galaxies (e.g. Richstone
et al 1998, Ho 2004, and references therein). The
highest redshift radio sources in our sample are
inevitably extremely young (Blundell Rawlings
1999), and our observations may allow us to
investigate the evolution of SMBHs through the
radio sources they make. Reasonable estimates of
the black hole mass can be made using (for our
redshift range) the continuum luminosity and the
CIV (l1549) line width (Vestergaard Peterson
2006). This adds an extra dimension to
investigating the evolution of source properties
at the earliest times.
THE RADIO STRUCTURES OF VERY HIGH REDSHIFT
QUASARS John Wardle1, Doug Gobeille1 and Teddy
Cheung2 1 Physics Department, Brandeis
University, Waltham, MA 02454 2 NASA/GSFC
Greenbelt, Maryland 20771 Abstract. We are making
VLA observations of all known radio-loud quasars
with z gt2.5 in the area common to both FIRST and
SDSS, and with radio fluxes gt 70 mJy at either
1.4 or 5 GHz. The sample contains 134 objects
ranging in redshift up to z 4.7. When combined
with previous lower redshift samples we have a
unique dataset for studying the evolution of
radio loud AGN and their environment back to the
earliest epochs. We will present VLA images
showing the variety of structures exhibited by
the highest redshift sources, and also give some
preliminary results concerning the density and
clumpiness of the intergalactic medium at these
early epochs. A parallel goal of these
observations is to find new high redshift jets
for observing with Chandra to test the IC/CMB
mechanism.

• Acknowledgements
• We are very grateful to Jennifer Carson, who
  compiled the original list of all 307 radio
  sources with z gt 2.5 from NED.
• This research has made use of the NASA/IPAC
  Extragalactic Database (NED) which is operated by
  the Jet Propulsion Laboratory, California
  Institute of Technology, under contract with the
  National Aeronautics and Space Administration.
• It has also made extensive use of the VLA
  archives of the National Radio Astronomy
  Observatory. The National Radio Astronomy
  Observatory is a facility of the National Science
  Foundation operated under cooperative agreement
  by Associated Universities, Inc.
• This work has been supported by grant AST-0607453
  from the Nation Science Foundation.
• References
• Abazajian, K. et al 2003, Astron. J. 126, 2081
  
• Barthel, P. D. Miley, G. K. 1988 Nature 333,
  319
• Becker, R. H., White, R. L., Helfand, D. J.
  1995, ApJ, 450, 559
• Blundell, K. M. Rawlings, S. 1999 Nature, 399,
  330
• Celotti, A., Ghisellini, G., Chiaberge, M.
  2001, MNRAS, 321, L1
• Cheung, C.C. 2004, ApJ, 600, L23
• Cheung, C. C., Stawarz, L. Siemiginowska, A.
  2006, ApJ, 650, 679 Cohen, M. H. (1989) in BL Lac
  Objects. Proceedings of a workshop held in Como,
  Italy.

Results This is a progress report on an on-going
project to explore the high redshift radio
universe. About one half of the 134 quasars with
z gt 2.5 have been imaged from data in the VLA
archives, and some general comments can be
made. (1) The range of structures seen among the
resolved sources mimics that seen in lower
redshift samples. We see classical triples,
doubles and jets. (2) The fraction of resolved
sources (41 ) is far lower than among the z lt
2.5 sources in the same area, studied by Barthel
Miley (1988) and by Neff Hutchins (1990). For
1.0 lt z lt 2.5 the fraction of resolved sources is
79 , and for z lt 1 it rises to 90 . This can be
attributed to the increasing emitted frequency at
higher redshifts, making steep spectrum extended
structure fainter relative to the cores. (3) The
great majority of the sources in our sample (84
) are selected by their core flux. Since the
cores are assumed to be beamed, this favors
sources whose jets make a small angle to the line
of sight and whose cores are Doppler boosted
(Cohen 1988, Lister and Marscher 1997). This
sample will have a strongly anisotropic
distribution of orientations, making the core to
lobe flux ratios larger and the projected linear
sizes smaller.
X-rays from kiloparsec scale jets at high
redshift. The detection of x-ray jets at large
redshifts is expected as a natural consequence of
the inverse Compton (IC) emission off the cosmic
microwave background (CMB) model (e.g., Tavecchio
et al. 2000 Celotti, Ghisellini, Chiaberge
2001). This is because the (1z)4 dependence of
the CMB energy density compensates for
cosmological dimming of radiation, so that IC/CMB
X-ray jets should remain detectable out to large
cosmological distances (Schwartz 2002). The model
has been successfully applied to account for
X-ray jets in many other powerful quasars at more
modest redshifts (e.g., Sambruna et al. 2004).
The model requires that the jets are still
relativistic on kiloparsec scales, in order that
the electrons in the jet frame see an adequately
boosted photon source. It is unclear whether this
is consistent with other estimates of jet speeds
on kiloparsec scales (e.g. Wardle Aaron 1997)
or if a more complicated model is required.
Further detections in x-rays of high redshift
jets will be a stringent test of the model.
(4) Superficially the median projected linear
size of the resolved sources at z gt 2.5 is
smaller (50 kpc) than at low redshift (100
kpc). In view of the bias towards small angles to
the line of sight mentioned above, and the
youthfulness effect (Blundell Rawlings 1999),
it would be premature to draw any conclusion at
all. Similar comments can be made about bending
angles for the triple sources. The 13 sources
above include one source with a bend of 90
degrees and others that are remarkably
straight. (5) Perhaps the most important result
is that two of our highest redshift quasars, at
redshifts of 3.89 and 3.82, are large triple
radio sources with total projected linear extents
of 149 kpc and 128 kpc respectively. Evidently,
even by the tender age of 1.6 Gy the universe had
already formed high luminosity radio loud AGN,
containing, presumably, supermassive black holes
of 108 109 solar masses.
Plot of the ratio of the jet X-ray to radio
monochromatic luminosity vs. redshift (taken from
Cheung (2004) and updated). Only jet features
interpreted by the authors as IC/CMB X-ray
emission are plotted. The curves indicate the
expected ratio for given combinations of B and d,
which scale as (1 z)4. For reference, the µG
and case derived for GB 15085714, which used
the additional equipartition constraint, defines
the dotted line that lies between the other two
curves. Light vertical lines connect features
from the same source, i.e. different knots in the
same jet. Red points are for lobe dominated
sources, and are consistent with lower Doppler
factors, as expected.