The formation of cluster elliptical galaxies as revealed by extensive star formation.
ABSTRACT The most massive galaxies in the present-day Universe are found to lie in the centres of rich clusters. They have old, coeval stellar populations suggesting that the bulk of their stars must have formed at early epochs in spectacular starbursts, which should be luminous phenomena when observed at submillimetre wavelengths. The most popular model of galaxy formation predicts that these galaxies form in proto-clusters at high-density peaks in the early Universe. Such peaks are indicated by massive high-redshift radio galaxies. Here we report deep submillimetre mapping of seven high-redshift radio galaxies and their environments. These data confirm not only the presence of spatially extended regions of massive star-formation activity in the radio galaxies themselves, but also in companion objects previously undetected at any wavelength. The prevalence, orientation, and inferred masses of these submillimetre companion galaxies suggest that we are witnessing the synchronous formation of the most luminous elliptical galaxies found today at the centres of rich clusters of galaxies.
- SourceAvailable from: arxiv.org[show abstract] [hide abstract]
ABSTRACT: We present the results of the first major systematic submillimetre survey of radio galaxies spanning the redshift range 1 < z < 5. The primary aim of this work is to elucidate the star-formation history of this sub-class of elliptical galaxies by tracing the cosmological evolution of dust mass. Using SCUBA on the JCMT we have obtained 850-micron photometry of 47 radio galaxies to a consistent rms depth of 1 mJy, and have detected dust emission in 14 cases. The radio galaxy targets have been selected from a series of low-frequency radio surveys of increasing depth (3CRR, 6CE, etc), in order to allow us to separate the effects of increasing redshift and increasing radio power on submillimetre luminosity. Although the dynamic range of our study is inevitably small, we find clear evidence that the typical submillimetre luminosity (and hence dust mass) of a powerful radio galaxy is a strongly increasing function of redshift; the detection rate rises from 15 per cent at z < 2.5 to 75 per cent at z > 2.5, and the average submillimetre luminosity rises as (1+z)^3 out to z~4. Moreover our extensive sample allows us to argue that this behaviour is not driven by underlying correlations with other radio galaxy properties such as radio power, radio spectral index, or radio source size/age. Although radio selection may introduce other more subtle biases, the redshift distribution of our detected objects is in fact consistent with the most recent estimates of the redshift distribution of comparably bright submillimetre sources discovered in blank field surveys. The evolution of submillimetre luminosity found here for radio galaxies may thus be representative of massive ellipticals in general. Comment: 31 pages - 10 figures in main text, 3 pages of figures in appendix. This revised version has been re-structured, but the analysis and conclusions have not changed. Accepted for publication in MNRAS02/2000;
- [show abstract] [hide abstract]
ABSTRACT: We present the source catalogue for the SCUBA Lens Survey. We summarise the results of extensive multi-wavelength observations of the 15 submillimetre-selected galaxies in the catalogue, from X-rays to radio. We discuss the main observational characteristics of faint submillimetre galaxies as a population, and consider their interpretation within the framework of our understanding of galaxy formation and evolution. Comment: 29 pages, 17 figures, MNRAS in press12/2001;
- [show abstract] [hide abstract]
ABSTRACT: SCUBA, the Submillimetre Common-User Bolometer Array, built by the Royal Observatory Edinburgh for the James Clerk Maxwell Telescope, is the most versatile and powerful of a new generation of submillimetre cameras. It combines a sensitive dual-waveband imaging array with a three-band photometer, and is sky-background limited by the emission from the Mauna Kea atmosphere at all observing wavelengths from 350 microns to 2 mm. The increased sensitivity and array size mean that SCUBA maps close to 10,000 times faster than its single-pixel predecessor (UKT14). SCUBA is a facility instrument, open to the world community of users, and is provided with a high level of user support. We give an overview of the instrument, describe the observing modes and user interface, performance figures on the telescope, and present a sample of the exciting new results that have revolutionised submillimetre astronomy. Comment: 15 pages, 15 figures, accepted for Monthly Notices of RASMonthly Notices of the Royal Astronomical Society 09/1998; · 5.52 Impact Factor
arXiv:astro-ph/0309495v1 17 Sep 2003
Version: 2 Febuary 2008
The formation of cluster elliptical galaxies as
revealed by extensive star formation
J. A. Stevens∗, R. J. Ivison∗, J. S. Dunlop†, Ian Smail‡, W. J. Percival†, D. H. Hughes§
H. J. A. R¨ ottgering?, W. J. M. van Breugel¶ & M. Reuland¶
∗Astronomy Technology Centre, Royal Observatory, Blackford Hill, Edinburgh
EH9 3HJ, UK
† Institute for Astronomy, University of Edinburgh, Blackford Hill,
Edinburgh EH9 3HJ
‡ Institute for Computational Cosmology, University of Durham,
South Road, Durham DH1 3LE
§ Instituto Nacional de Astrofisica, Optica y Electronica, Apartado Postal 51 y 216, 72000
? Leiden Observatory, PO Box 9513, 2300 Leiden, The Netherlands
¶ Institute of Geophysics and Planetary Physics, Lawrence Livermore National Laboratory,
PO Box 808, Livermore, CA94459, USA
The most massive galaxies in the present-day Universe are found to lie in the
centres of rich clusters. They have old, coeval stellar populations suggesting
that the bulk of their stars must have formed at early epochs in spectacular
starbursts1– luminous phenomena at submillimetre wavelengths2. The most
popular model of galaxy formation predicts that these galaxies form in proto-
clusters at high-density peaks in the early Universe3. Such peaks are signposted
by massive high-redshift radio galaxies4. Here we report deep submillimetre
mapping of seven high-redshift radio galaxies and their environments. These
data confirm not only the presence of spatially extended massive star-formation
activity in the radio galaxies themselves, but also in companion objects previ-
ously undetected at any wavelength. The prevalence, orientation, and inferred
masses of these submillimetre companion galaxies suggest that we are witness-
ing the synchronous formation of the most luminous elliptical galaxies found
today at the centres of rich galaxy clusters.
– 1 –
Whilst existing submillimetre studies of high-redshift radio galaxies (hereafter HzRGs)
have shown that their star-formation rates are large enough to build a massive galaxy in
< 1 Gyr2,5,6,7they have provided no information on the spatial extent of this emission
or on the star-formation activity in their environments. We have therefore mapped a
sample of seven objects with redshifts ranging from 2.2 to 4.3 at a wavelength of 850µm
with the Submillimetre Common-User Bolometer Array (SCUBA)8on the James Clerk
Maxwell Telescope (JCMT). The targets were chosen from those sources found to be
submillimetre bright in the previous SCUBA surveys of HzRGs6,7. Our new maps illustrate
the distribution of dust-reradiated emission in and around the HzRGs on scales from 5′′
to 160′′, or 30kpc to 1Mpc. We illustrate the seven submillimetre maps from this survey
in Figure 1; the orientation of the radio jets of each HzRG is represented by tick marks on
One of the most striking aspects of the submillimetre maps is that the dust emission
from the central radio galaxy is resolved in at least five of the seven sources – even with
the coarse beam of the JCMT. In Figure 2 and Table 1 we present simple two-dimensional
Gaussian fits to the data which, while not giving a true reflection of the physical situation,
at least provide a quantitative measure of the spatial extent of the dust emission. This
emission is sometimes in the form of several partially-resolved or merged clumps (typified
by 8C1909+722), sometimes in an apparently smoother distribution (e.g. 4C60.07), and
is more extended than the radio emission in most cases. The extent of the dust emission
ranges from 50 to 250 kpc, a physically interesting size because (1) the corresponding
half-light radii (30–150 kpc) are equivalent to those of brightest cluster galaxies in the
local Universe9, and (2) gas-dynamical simulations of major galaxy mergers predict that
star formation should peak when the galaxies are separated on approximately this scale.10
Higher resolution, high signal-to-noise millimetre/submillimetre imaging observations are
required to investigate these possibilities.
A second striking feature of our SCUBA maps is that several of the fields also contain
serendipitous detections of new, luminous submillimetre galaxies. In Table 1 we list the
properties of those companion galaxies with peak signal-to-noise ratio > 4. Note that
several of these companions are, like the HzRGs, resolved at 850µm. The source density
of companions is also higher than found in ‘blank-field’ surveys; current submillimetre
number counts11indicate that we should find, on average, about one random submillimetre
source per SCUBA field at a flux level of S850> 5–6 mJy compared to our detection of
approximately twice this number (Table 1). Furthermore, the 4C 41.17 and 8C 1909+722
fields each contain a companion source with S850> 20 mJy. By comparison, the biggest
blank field survey conducted to date12, which has an area ∼ 7 times that of our own,
contains no robust sources with S850> 15 mJy.
Arguably the most surprising feature of the maps is the observed aligments between
the HzRG radio axis and (1) the submillimetre emission (2) the brightest submillimetre
‘companions’. The significance of these alignments is presented and quantified in Figure 3.
These results can be interpreted as follows. First, while one might be tempted to conclude
that the first alignment effect is indicative of jet-induced star formation similar to the
radio-UV alignment effect previously reported in HzRGs13,14, the fact that in all but two
cases, the radio source is also aligned with submillimetre source on scales well beyond the
– 2 –
size of the radio emission indicates that this is unlikely to be the complete explanation.
Moreover, this model clearly cannot explain why the submillimetre emission from the
HzRGs appears to be aligned with positions of the brightest companion sources in their
vicinity (in at least 4 out of 7 cases).
There is one model that can explain both alignment effects. We propose that the
brightest submillimetre companions trace the large-scale structure around the HzRG and
that these directions thus contain the densest cross-sections of gas. Next, by selecting some
of the very brightest known radio sources at this epoch we will then have (inadvertently,
but probably inevitably) selected sources which happen to have produced jets aligned
with the densest regions of gas, thereby producing very effective working surfaces and the
brightest hot spots. Such a selection effect has been suggested before to explain apparent
large-scale optical-radio alignment effects at lower redshift.15,4
Of course, an important corollary to this interpretation is that the brightest com-
panion sources seen in the submillimetre maps must lie at the same redshift as the radio
galaxies, occupying the same large-scale structure. However, the fact that extension of this
analysis to the second brightest apparent companion object does not reveal a significant
alignment effect shows that not all of the objects seen in these images need lie at the same
redshift as, and be physically associated with the radio galaxies. This is consistent with
our earlier estimate of the rate of contamination in these fields by unrelated submillimetre
sources. Thus the analysis of these possible alignment effects indicates that the submillime-
tre emission from the radio galaxies themselves, and that of the brightest companions, is
tracing the large scale structure around radio galaxies at z ≃ 3. We note that for 53W002,
a HzRG not included in our sample, it has been shown with optical spectroscopy that the
brightest submillimetre companion is indeed at the same redshift as the radio galaxy.16A
similar conclusion can be inferred for PKS 1138 − 262 where the brightest submillimetre
companion is coincident with one of 5 actuve galactic nuclei (AGN) forming a radio-aligned
filament in the plane of the sky, and which has the same redshift as the HzRG based on
narrow-band imaging of redshifted Lyα emission.17
Since we have shown that these fields contain overdensities of submillimetre sources,
it is of interest to ask what is the typical mass of these galaxies. We can do this by
assuming that they lie at the same redshift as the HzRGs. The dust masses of the HzRGs,
calculated from their 850-µm fluxes18are given in Table 1. Using standard assumptions
we can convert these dust masses into total gas masses, and hence estimate the masses of
their associated dark matter haloes (see table caption for details). We estimate that they
reside in dark matter halos with masses in excess of 1012M⊙. Since the submillimetre flux
densities of the companions are similar to those of the HzRGs we can infer that they also
reside in dark matter halos with masses in excess of 1012M⊙. We therefore conclude that
these regions can contain more than one very massive and gas-rich galaxy of > 1012M⊙.
Can we reconcile this result with the hierarchical models of galaxy and structure formation?
By design we have selected highly-biased regions of the high-redshift universe by imag-
ing around some of the most luminous radio galaxies at these epochs, which are expected
to host some of the most massive black holes. We can attempt to quantify the impact
of this bias by exploring the predictions of numerical simulations of the collapse of Cold
Dark Matter haloes within a Λ-dominated universe. The details and results of these sim-
– 3 –
ulations are given in Figure 4. These simulations do indeed appear to be in accord with
the observations, i.e. both data and theory suggest that the submillimetre companions
revealed by our SCUBA imaging are associated with dark-matter halos more massive than
1012M⊙. This conclusion is, however, subject to an assumption about the duty cycles of
the submillimetre luminous phase; if true it would imply that the vast majority of such
massive halos are actively forming stars at the epochs sampled by these images. This is
not unreasonable, especially since we have deliberately targetted fields in which at least
one massive object (i.e. the HzRG) is actively engaged in intense star formation.
Pursuing this comparison one step further, we can explore the predicted and observed
distribution of companion objects in this high-mass range. In fact, while the average
number of such companions in the simulated SCUBA images centred on the HzRG is 1,
this average arises from a skewed distribution. As shown in Figure 4, of the 40 high-mass
haloes investigated, 50% have no such massive companions, with the average of 1 resulting
from the fact that, of the remaining half, 40% have 2 companions or more. We have
over-plotted in Figure 4 the corresponding histogram of companion incidence for the 7
HzRGs imaged in this study (after statistically correcting for field contamination). This
comparison at least illustrates that the distribution of companion incidence in our data is
consistent with the prediction of the simulations, i.e. typically we see either no companion,
or ≥ 2 companions.
The average number of companions, the distribution of companion incidence, and the
inferred baryonic gas masses of the companions are all most consistent with the interpre-
tation that the SCUBA sources uncovered in this study are the progenitors of massive
present-day cluster ellipticals. In this case, since elliptical galaxies are known to contain
massive black holes with mass proportional to that of the spheroid,19it is reasonable to as-
sume that the black hole and stellar mass grow coevally from the same gas reservoir. This
in turn suggests that the companion objects should contain buried AGN.20,21First results
show a high rate of correspondence between the submillimetre companions and luminous
X-ray sources,16,22suggesting that this is indeed the case. We are presently pursuing high-
resolution follow-up observations of these fields at millimetre and optical/near-infrared
wavelengths. The former will yield information on source structure while the latter will
reveal counterparts for spectroscopic redshift determination on 10-m class telescopes.
– 4 –
1. Ellis, R. S. et al. The homogeneity of spheroidal populations in distant clusters.
Astrophys. J. 483, 582–596 (1997).
2. Dunlop, J. S. et al. Detection of a large mass of dust in a radio galaxy at redshift
z = 3.8. Nature 370, 347–349 (1994).
3. Kauffmann, G. The age of elliptical galaxies and bulges in a merger model. Mon. Not.
R. Astron. Soc. 281, 487–492 (1996).
4. West M. J. Anisotropic mergers at high redshifts: the formation of cD galaxies and
powerful radio sources. Mon. Not. R. Astron. Soc. 268, 79–102 (1994).
5. Hughes, D. H., Dunlop, J. S. & Rawlings, S. High-redshift radio galaxies and quasars at
submillimetre wavelengths: assessing their evolutionary status. Mon. Not. R. Astron.
Soc. 289, 766–782 (1997).
6. Archibald, E. N. et al. A submillimetre survey of the star-formation history of radio
galaxies. Mon. Not. R. Astron. Soc. 323, 417–444 (2001).
7. Reuland, M., R¨ ottgering, H. & van Breugel, W. SCUBA observations of high redshift
radio galaxies IN: ‘Radio Galaxies: Past, present and future.’ Elsevier Science, in
8. Holland, W. S. et al. SCUBA: a common-user submillimetre camera operating on the
James Clerk Maxwell Telescope. Mon. Not. R. Astron. Soc. 303, 659–672 (1999).
9. Graham, A., Lauer T. R., Colless M., & Postman M. Brightest cluster galaxy profile
shapes. Astrophys. J. 465, 534–547 (1996).
10. Mihos, J. C., & Hernquist, L. Gasdynamics and starbursts in major mergers. Astro-
phys. J. 464, 641–663 (1996).
11. Smail, I., Ivison, R. J. Blain, A. W. & Kneib, J.-P. The nature of faint submillimetre-
selected galaxies. Mon. Not. R. Astron. Soc. 331, 495–520 (2002).
12. Scott, S. E. et al. The SCUBA 8-mJy survey - I. Submillimetre maps, sources and
number counts. Mon. Not. R. Astron. Soc. 331, 817–838 (2002).
13. Chambers, K. C., Miley, G. K. & van Breugel, W. Alignment of radio and optical
orientations in high-redshift radio galaxies. Nature 329, 604–606 (1987).
14. McCarthy, P. J., van Breugel, W., Spinrad, H. & Djorgovski, S. A correlation between
the radio and optical morphologies of distant 3CR radio galaxies. Astrophys. J. 321,
15. Eales, S. A. A new theory for the alignment effect. Astrophys. J. 397, 49–54 (1992).
16. Smail, I. et al. A SCUBA galaxy in the protocluster around 52W002 at z=2.4. As-
trophys. J. 583, 551–558 (2003).
17. Pentericci, L. et al. A Chandra study of X-ray sources in the field of the z=2.16 radio
galaxy MRC 1138-262 Astron. & Astrophys. 396, 109–115 (2002).
18. Hildebrand, R. H. The determination of cloud masses and dust characteristics from
submillimetre thermal emission. Qua. J. R. Astron. Soc. 24, 267–282 (1983).
19. Magorrian, J. et al. The demography of massive dark objects in galactic centers.
Astron. J. 115, 2285–2305 (1998).
– 5 –
20. Kauffmann, G. & Haehnelt, M. A unified model for the evolution of galaxies and
quasars. Mon. Not. R. Astron. Soc. 311, 576–588 (2000).
21. Page, M. J., Stevens, J. A., Mittaz, J. P. D. & Carrera, F. J. Submillimetre evidence
for the coeval growth of massive black holes and galaxy bulges. Science 294, 2516–
22. Smail, I. et al. Chandra detections of SCUBA galaxies around High-z radio sources.
Astrophys. J. submitted (2003).
23. De Breuck, C. et al. CO emission and associated HI absorption from a massive gas
reservoir surrounding the z = 3 radio galaxy B3J2330+393. Astron. & Astrophys.
401, 911–925 (2003).
24. Dunlop, J. S. et al. Quasars, their host galaxies, and their central black holes. Mon.
Not. R. Astron. Soc. 340, 1095–1135 (2003).
25. Gorenstein, P. Empirical relation between interstellar X-ray absorption and optical
extinction. Astrophys. J. 198, 95–101 (1975).
26. Ivison, R. J. et al. An excess of submillimeter sources near 4C41.17: a candidate
proto-cluster at z = 3.8? Astrophys. J. 542, 27–34 (2000).
27. Omont, A. et al. Molecular gas and dust around a radio-quiet quasar at redshift 4.69.
Nature 382, 428–431 (1996).
28. Papadopoulos, P. P. et al. CO(4–3) and dust emission in two powerful high-z radio
galaxies, and CO lines at high redshifts. Astrophys. J. 528, 626–636 (2000).
29. Percival, W. J., Scott, D., Peacock, J. A. & Dunlop, J. S. The clustering of halo
mergers. Mon. Not. R. Astron. Soc. 338, L31–L35 (2003).
30. Reuland, M. et al. An obscured radio galaxy at high redshift. Astrophys. J. 582,
The JCMT is run by the Joint Astronomy Centre on behalf of the Particle Physics & As-
tronomy Research Council (PPARC). JAS, JSD, IRS and WJP would like to acknowledge
funding from PPARC, the Royal Society and the Leverhulme Trust. The work of WvB
and MR was performed under the auspices of the U.S. Department of Energy, National
Nuclear Security Administration by the University of California, Lawrence Livermore Na-
Correspondence and requests for materials should be addressed to Jason Stevens (firstname.lastname@example.org).
– 6 –
Table 1. Target fields.
nameR.A. Dec.(mJy) (mJy)(M⊙)(′′/◦)(′′/◦)
8C 1909+72219 08 23.30 +72 20 10.43.54
1. 19 08 27.47 +72 19 28.0
2.19 08 29.31 +72 20 49.6
3.19 08 16.12+72 20 24.0
8C 1435+63514 36 37.33+63 19 13.14.261
1.14 36 32.46+63 20 02.5
B3 J2330+3927 23 30 24.91+39 27 11.23.086
1.23 30 19.14 +39 27 03.0
2. 23 30 20.52+39 26 57.7
PKS 1138-26211 40 48.25 -26 29 10.12.156
1.11 40 53.38 -26 29 11.9
2. 11 40 45.80-26 29 56.6
3. 11 40 45.61-26 29 06.6
4C 60.0705 12 54.80+60 30 51.73.788
1. 05 12 46.52+60 30 35.8
4C 41.17 06 50 52.15+41 30 30.83.792
1. 06 50 51.52+41 30 01.5
2.06 50 49.25+41 30 01.5
WN J0305+352503 05 47.42+35 25 13.42@64
aCoordinates of the HzRG companion sources are measured from the submillimetre maps. They are
accurate to 4−5 arcseconds.
bSpeakis the flux density per beam calculated at the peak of the dust emission. Stotalis the flux density
of total dust emission calculated in an aperture. A calibration uncertainty of ∼ 10% is not included
in these values. For some sources, the higher total flux densities reflect the extended nature of the dust
emission, consistent with the finding that interferometric millimetre flux densities are often smaller than
those calculated from single dish measurements.28,23
cDust mass calculated from the peak fluxes, and assuming a dust emissivity index, β=2.0; a dust tem-
perature, Td=40 K typical of HzRGs; a mass absorption coefficient, kd(850µm)=0.076 m2kg−1; and an
Ωm=0.3, ΩΛ=0.7, h =0.7 cosmology. Our selection criteria result in a small spread in the calculated
dust masses; an average value is ∼ 8 × 108M⊙. We can convert this dust mass into a conservative
gas mass (molecular and atomic) assuming a standard Galactic gas-to-dust mass ratio of 20025giving
1.6×1011M⊙. We can confirm this estimate using the measured CO luminosities of three of the HzRGs,
which can then be turned into molecular gas, and then total gas, mass using the standard Galactic con-
version factor between CO luminosity and H2mass and a ratio of atomic-to-molecular gas of ∼ 2. In
this manner, gas masses of 1.3 − 2.2 × 1011M⊙have been estimated for 4C 60.07, 8C 1909+722 and
B3 J2330+3927.28,23Hence the gas masses from the two methods are in reasonable agreement, and a
representative estimate for the total gas mass in a HzRG with a submillimetre flux density of 10–15 mJy
is thus ≃ 2 × 1011M⊙. While this number is arguably uncertain by a factor of a few, it is very hard to
argue that it can be inflated by an order of magnitude.5Assuming an order of magnitude more mass is in
– 7 –
the form of dark matter we estimate the halo masses to be > 1012M⊙.
dFrom a 2-dimensional Gaussian fit to the 850µm emission, deconvolved from the beam. All P.A.s are
measured east of north. For the companion galaxies we give sizes for the brightest (highest S/N) sources
only. A quoted size of zero arcseconds means that the source was not resolved in that direction.
eQuoted radio sizes are hotspot separations measured at 5 GHz.
fWNJ 0305+35 lacks a robust spectroscopic redshift, but is thought to lie at z = 3 ± 1.30
– 8 –
Figure 1. Continuum emission from dust in and around seven high-redshift radio galaxies. The data for
4C 41.17 have been published previously.26Data were obtained using the SCUBA8submillimetre camera
on the James Clerk Maxwell Telescope (JCMT) in its dual-wavelength mapping mode during 1998–2001
in the top 20 percentile of Mauna Kea weather conditions. The beam, after smoothing to a FWHM of
14.4′′at 850µm, is shown bottom right in the form of a map of the blazar, 3C 345, with contours at −40,
−30, −10, +10, +30, +50, +70, +90% of the central peak. The JCMT secondary mirror was chopped
and nodded by 30′′in Right Ascension (R.A.), resulting in the −0.5/+1.0/−0.5 beam profile. After all
overheads, 30–40 ks were spent integrating on each of the fields, a total of ∼100 hr before overheads, split
down into 1280- s integrations separated by checks on focus, pointing accuracy and atmospheric opacity.
All images have been cleaned with the smoothed beam.26Submillimetre contours are shown at −3, 3, 4,
5, 6, 8, 10 ×σ, where σ includes the contribution due to sources below the detection threshold and to the
chopping and nodding procedure and ranges from 1 to 2 mJy beam−1. The galaxy with the most compact
appearance, 4C 41.17, actually subtends ∼7.5′′FWHM when deconvolved from the beam. The tick marks
around the central radio galaxies show the direction of the kiloparsec scale radio jets.
Figure 2. Submillimetre extent of the radio galaxies. Radial profiles of the dust emission compared
with that of the beam (solid line). The x-axis shows the radius from the centre of the radio galaxy dust
emission (arcseconds) and the y-axis shows the normalised flux density. An accurate beam profile or point
spread function (PSF) was measured with a ∼ 1 hr mapping observation of the blazar 3C 345 which
had an 850µm flux density of > 2 Jy at the time. The image (shown in Figure 1) has a full width
half maximum (FWHM) of 14.4′′. All subsequent size measurements are deconvolved from the PSF (see
Table 1). The reality of the extended nature of the dust emission is established from observations of
the quasar, BR 1202−0725, using the same technique and integration times used for the radio galaxies
to mimic any systematic errors. BR 1202−0725 serves as a particularly conservative comparison object
— its dust continuum was resolved27by the IRAM interferometer into two components separated by 4′′
(P.A. 120±10◦). As expected, it appears relatively compact (6.5′′× 5.5′′at P.A. 6◦) in our submillimetre
maps, and its radial profile is much more similar to that of the beam than to those of the HzRGs. We
conclude that our observing procedure does not produce spurious morphologies, that the extended nature
of the submillimetre emission from the radio galaxies is real and occurs on scales that are typically ≫4′′.
Figure 3. Observed alignment effects. The histograms show the offset in degrees between the kilo-
parsec scale radio structure of the HzRG and a) the position angle of the submillimetre source, b) the
position angle defined by the projected vector joining the radio source to the brightest, spatially distinct
submillimetre source, and c) the position angle defined by the projected vector joining the radio source
to the second brightest, spatially distinct submillimetre source. In four cases these submillimetre sources
are the companions listed in Table 1. In the remaining three cases one or both of the companions has a
signal-to-noise ratio of 2 to 4; while these sources are not listed in Table 1 we use them in the analysis for
completeness. We have applied the Kolmogorov Smirnov test to assess whether the apparent alignment
effect seen in the first two of these plots is indeed statistically significant. Despite the small number statis-
tics, we find that both of these apparent alignment effects are significant at the 2-σ level (as compared
to a random distribution; p = 0.04 and p = 0.05 respectively) whereas the corresponding result for the
third distribution is not significant (p = 0.3).
Figure 4. Comparison with hierarchical models. The predicted distribution of number of companion
sources per SCUBA field (after subtracting a background of 1 source per map for field contamination) from
our Λ–CDM simulations shown against the observed distribution (shaded region). For the simulations we
– 9 –
assume that the radio emission from the HzRG is powered by black holes with mass ≃ 109M⊙24, and
therefore these lie at the heart of dark matter haloes with masses of ≃ 1013M⊙19. Allowing for intrinsic
scatter and uncertainties in this estimate, we therefore searched for virialized dark matter haloes with
masses in the range 12.5 < log10(Mhalo)/M⊙< 13.5 at z ≃ 3 within an N-body simulation of a
region of a Λ–CDM universe, covering a comoving volume of (100h−1Mpc)3. Parameters of the simulation
were matched to the “concordance model” with Ωm= 0.3, ΩΛ= 0.7, h = 0.7; further details of the
simulation are given elsewhere (ΛCDM100a)29. Within this simulation we identified 40 haloes which
could be the counterparts of the HzRGs, and then investigated the mass distribution of other virialized
dark-matter haloes within a sphere of projected diameter 3-arcmin (approximately the SCUBA field of
view) around each of these. Within these regions we found typically 10 haloes with M > 1011M⊙, but
on average only 1 halo with M > 1012M⊙.
– 10 –
– 11 –
8C 1909+7228C 1435+635
– 12 –
– 13 –