The host galaxies of luminous radio-quiet quasars

We present the results of a deep K -band imaging study which reveals the host galaxies around a sample of luminous radio-quiet quasars. The K -band images, obtained at UKIRT, are of sufficient quality to allow accurate modelling of the underlying host galaxy. Initially, the basic structure of the hosts is revealed using a modified clean deconvolution routine optimized for this analysis. Two of the 14 quasars are shown to have host galaxies with violently disturbed morphologies which cannot be modelled by smooth elliptical profiles. For the remainder of our sample, 2D models of the host and nuclear component are fitted to the images using the χ ² statistic to determine goodness of fit. Host galaxies are detected around all of the quasars. The reliability of the modelling is extensively tested, and we find the host luminosity to be well constrained for nine quasars. The derived average K -band absolute K -corrected host galaxy magnitude for these luminous radio-quiet quasars is 〈 M _K 〉=25.15±0.04, slightly more luminous than an L * galaxy. The spread of derived host galaxy luminosities is small, although the spread of nuclear-to-host ratios is not. These host luminosities are shown to be comparable to those derived from samples of quasars of lower total luminosity, and we conclude that there is no correlation between host and nuclear luminosity for these quasars. Nuclear-to-host ratios break the lower limit previously suggested from studies of lower nuclear luminosity quasars and Seyfert galaxies. Morphologies are less certain but, on the scales probed by these images, some hosts appear to be dominated by spheroids while others appear to have disc-dominated profiles.     

Clustering of galaxies around radio quasars at 0.5≤z≤0.8

We have observed the galaxy environments around a sample of 21 radio-loud, steep-spectrum quasars at 0.5≤ z ≤0.82, spanning several orders of magnitude in radio luminosity. The observations also include background control fields used to obtain the excess number of galaxies in each quasar field. The galaxy excess was quantified using the spatial galaxy–quasar correlation amplitude, B _gq, and an Abell-type measurement, N _0.5. A few quasars are found in relatively rich clusters, but on average, they seem to prefer galaxy groups or clusters of approximately Abell class 0. We have combined our sample with literature samples extending down to z ≈0.2 and covering the same range in radio luminosity. By using the Spearman statistic to disentangle redshift and luminosity dependences, we detect a weak, but significant, positive correlation between the richness of the quasar environment and the radio luminosity of the quasar. However, we do not find any epoch dependence in B _gq, as has previously been reported for radio quasars and galaxies. We discuss the radio luminosity–cluster richness link and possible explanations for the weak correlation that is seen.     

A comparative HST imaging study of the host galaxies of radio-quiet quasars, radio-loud quasars and radio galaxies – I

We present the first results from a major HST WFPC2 imaging study aimed at providing the first statistically meaningful comparison of the morphologies, luminosities, scalelengths and colours of the host galaxies of radio-quiet quasars, radio-loud quasars and radio galaxies. We describe the design of this study and present the images that have been obtained for the first half of our 33-source sample. We find that the hosts of all three classes of luminous AGN are massive elliptical galaxies, with scalelengths ≃10 kpc, and R − K colours consistent with mature stellar populations. Most importantly, this is first unambiguous evidence that, just like radio-loud quasars, essentially all radio-quiet quasars brighter than M _R =−24 reside in massive ellipticals. This result removes the possibility that radio 'loudness' is directly linked to host galaxy morphology, but is however in excellent accord with the black hole/spheroid mass correlation recently highlighted by Magorrian et al. We apply the relations given by Magorrian et al. to infer the expected Eddington luminosity of the putative black hole at the centre of each of the spheroidal host galaxies we have uncovered. Comparison with the actual nuclear R -band luminosities suggests that the black holes in most of these galaxies are radiating at a few per cent of the Eddington luminosity; the brightest host galaxies in our low- z sample are capable of hosting quasars with M _R ≃− 28, comparable to the most luminous quasars at z ≃3. Finally, we discuss our host-derived black hole masses in the context of the radio luminosity:black hole mass correlation recently uncovered for nearby galaxies by Franceschini et al., and consider the resulting implications for the physical origin of radio loudness.     

The evolution of star formation in quasar host galaxies

We have used far-infrared data from IRAS , Infrared Space Observatory ( ISO ), Spitzer Wide-Area Infrared Extragalactic (SWIRE), Submillimetre Common User Bolometer Array (SCUBA) and Max-Planck Millimetre Bolometer (MAMBO) to constrain statistically the mean far-infrared luminosities of quasars. Our quasar compilation at redshifts  0 < z < 6.5  and I -band luminosities  −20 < I _AB < −32  is the first to distinguish evolution from quasar luminosity dependence in such a study. We carefully cross-calibrate IRAS against Spitzer and ISO , finding evidence that IRAS 100-μm fluxes at <1 Jy are overestimated by ∼30 per cent. We find evidence for a correlation between star formation in quasar hosts and the quasar optical luminosities, varying as star formation rate (SFR)  ∝ L ^0.44±0.07_opt  at any fixed redshift below   z = 2  . We also find evidence for evolution of the mean SFR in quasar host galaxies, scaling as  (1 + z )^1.6±0.3  at   z < 2  for any fixed quasar I -band absolute magnitude fainter than −28. We find no evidence for any correlation between SFR and black hole mass at  0.5 < z < 4  . Our data are consistent with feedback from black hole accretion regulating stellar mass assembly at all redshifts.     

HST Planetary Camera images of quasar host galaxies

We present Hubble Space Telescope ( HST ) images of seven low-redshift quasars (six taken with the Planetary Camera, one with the Wide Field Camera). These complete the sample of 14 quasars observed by the Faint Object Camera Investigation Definition Team (FOC IDT). Following subtraction of the quasar nuclear light, host galaxies can be seen in all seven cases. A combination of the optical morphology and luminosity profiles of the residual host galaxies and the results of 2D cross-correlation model fitting implies that five of the objects have elliptical host galaxies and two have disc host galaxies. The luminosities vary from slightly fainter than L ^* to about 1.3 mag brighter than L ^*.   We discuss the properties of the complete sample of 14 quasars. Nine of the objects appear to have elliptical host galaxies (all six of the radio-loud quasars in the sample as well as three radio-quiet quasars). Two further radio-quiet quasars appear to lie in disc galaxies. The other three objects (radio-quiet, ultraluminous infrared quasars) all lie in violently interacting systems. The sample as a whole has an average luminosity about 0.8 mag brighter than L ^*, although the radio-loud objects have hosts on average 0.7 mag brighter than the radio-quiet objects.   We compare our results with those from HST imaging of quasars by other authors. Taken together, our observations are in broad agreement with those of Bahcall et al. Radio-loud quasars appear to lie in luminous elliptical galaxies whereas radio-quiet quasars are found to lie in either elliptical or spiral hosts. Host galaxy luminosities (of radio-quiet and radio-loud quasars) are much brighter than would be expected if they followed a Schechter luminosity function.     

Three quasars from a survey of strong 25-μm emitters

We have carried out a spectroscopic survey of 750 sources that are strong 25-μm emitters from the IRAS Faint Source data base. Many of these sources are previously unknown active galactic nuclei including new IRAS quasars, three of which we describe here: F21382−2659, Z06367−6845 and Z05558−5008. They are all radio and X-ray quiet, and compared to the known IRAS quasars they have similar 25-μm luminosities, L (25 μm), but lower values of L (25 μm)/ L ( B ). Their F (25 μm)/ F (60 μm) IRAS colours lie in the range 0.33 to 1.08, indicating the presence of relatively warm dust, presumably in a dusty torus surrounding the central source, and with temperatures similar to those of the known IRAS quasars. The quasar with the warmest dust, F21382−2659, exhibits broad (full width at half-maximum ∼4000 km s⁻¹) asymmetric Balmer lines with H γ having an opposite asymmetry to the other broad lines; also H β (only) is double-peaked. Fe  ii is very weak in F21382−2659 but strong in the other two quasars, and the anticorrelation between Fe  ii and [O  iii ] holds as anticipated. Two of the quasars are unpolarized: although F21382−2659 is optically polarized (2.1 per cent at 4950 Å), we argue that this provides little insight into the orientation of its dust torus relative to the line of sight.     

1–5 μm imaging of 3CRR galaxies: the K–z relation and the geometry of the torus

It has been claimed by Taylor et al. that the low-redshift end of the K – z relation for radio galaxies is too bright by about half a magnitude owing to contributions from the obscured quasar nuclei. Such a result has major implications for the use of the K -band Hubble diagram in understanding the cosmological evolution of radio galaxies. In this paper we present 1–5-μm imaging data of a nearly complete sample of low-redshift radio galaxies; this approach allows us to determine accurately the strengths of any unresolved nuclear components in the galaxies. We detect nuclear sources in five targets, whose broad-band colours are consistent with reddened quasar spectra. In all the five cases the ratio of the inferred intrinsic near-infrared luminosity to the narrow-line luminosity is typical of quasars. We find a correlation between the inferred nuclear extinction and core-to-lobe ratio, which places constraints on the geometry of the torus. We find evidence for a shift of the K – z relation to fainter magnitudes, but by a much smaller amount (∼0.1 mag) than determined by Taylor et al. Under the assumption that the nuclear sources in radio galaxies have the same intrinsic near-infrared spectra as quasars, our multiwavelength images allow us to limit any possible shift to less than 0.3 mag.     

Infrared photometry of z ∼ 1 3C quasars

We present JHKL ' photometry of a complete sample of steep-spectrum radio-loud quasars from the revised 3CR catalogue in the redshift range 0.65 z <1.20. After correcting for contributions from emission lines and the host galaxies, we investigate their spectral energy distributions (SEDs) around 1 μm. About 75 per cent of the quasars are tightly grouped in the plane of optical spectral index, α _opt, versus near-infrared spectral index, α _IR, with the median value of α _opt close to the canonical value, and the median α _IR slightly flatter. We conclude that the fraction of moderately obscured, red quasars decreases with increasing radio power, in accordance with the 'receding torus' model which can also explain the relatively flat median near-infrared spectra of the 3CR quasars. Two of the red quasars have inverted infrared spectral indices, and we suggest that their unusual SEDs might result from a combination of dust-scattered and transmitted quasar light.     

Modelling the cosmological co-evolution of supermassive black holes and galaxies – II. The clustering of quasars and their dark environment

We use semi-analytic modelling on top of the Millennium simulation to study the joint formation of galaxies and their embedded supermassive black holes. Our goal is to test scenarios in which black hole accretion and quasar activity are triggered by galaxy mergers, and to constrain different models for the light curves associated with individual quasar events. In the present work, we focus on studying the spatial distribution of simulated quasars. At all luminosities, we find that the simulated quasar two-point correlation function is fit well by a single power law in the range  0.5 ≲ r ≲ 20  h ⁻¹ Mpc  , but its normalization is a strong function of redshift. When we select only quasars with luminosities within the range typically accessible by today's quasar surveys, their clustering strength depends only weakly on luminosity, in agreement with observations. This holds independently of the assumed light-curve model, since bright quasars are black holes accreting close to the Eddington limit, and are hosted by dark matter haloes with a narrow mass range of a few  10¹²  h ⁻¹ M_⊙  . Therefore, the clustering of bright quasars cannot be used to disentangle light-curve models, but such a discrimination would become possible if the observational samples can be pushed to significantly fainter limits. Overall, our clustering results for the simulated quasar population agree rather well with observations, lending support to the conjecture that galaxy mergers could be the main physical process responsible for triggering black hole accretion and quasar activity.     

The European Large Area ISO Survey – IV. The preliminary 90-μm luminosity function

We present the luminosity function of 90-μm-selected galaxies from the European Large Area ISO Survey (ELAIS), extending to z =0.3. Their luminosities are in the range 10⁹65⁻² L /L_⊙<10¹², i.e. non-ultraluminous. From our sample of 37 reliably detected galaxies in the ELAIS S1 region from the Efstathiou et al. S ₉₀100 mJy data base, we have found optical, 15-μm or 1.4-GHz identifications for 24 (65 per cent). We have obtained 2dF and UK Schmidt FLAIR spectroscopy of 89 per cent of identifications to rigid multivariate flux limits. We construct a luminosity function assuming that (i) our spectroscopic subset is an unbiased sparse sample, and (ii) there are no galaxies that would not be represented in our spectroscopic sample at any redshift. We argue that we can be confident of both assumptions. We find that the luminosity function is well described by the local 100-μm luminosity function of Rowan-Robinson, Helou & Walker. Assuming this local normalization, we derive luminosity evolution of (1+ z )^2.45±0.85 (95 per cent confidence). We argue that star formation dominates the bolometric luminosities of these galaxies, and we derive comoving star formation rates in broad agreement with the Flores et al. and Rowan-Robinson et al. mid-infrared-based estimates.     

High-redshift obscured quasars: radio emission at sub-kiloparsec scales

The radio properties of 11 obscured 'radio-intermediate' quasars at redshifts   z ≳ 2  have been investigated using the European Very-Long-Baseline-Interferometry Network (EVN) at 1.66 GHz. A sensitivity of  ∼25 μJy per 17 × 14 mas² beam  was achieved, and in seven out of 11 sources unresolved radio emission was securely detected. The detected radio emission of each source accounts for ∼30–100 per cent of the total source flux density. The physical extent of this emission is ≲150 pc, and the derived properties indicate that this emission originates from an active galactic nucleus (AGN). The missing flux density is difficult to account for by star formation alone, so radio components associated with jets of physical size ≳150 pc and ≲40 kpc are likely to be present in most of the sources. Amongst the observed sample steep, flat, gigahertz-peaked and compact-steep spectrum sources are all present. Hence, as well as extended and compact jets, examples of beamed jets are also inferred, suggesting that in these sources, the obscuration must be due to dust in the host galaxy, rather than the torus invoked by the unified schemes. Comparing the total to core (≲150 pc) radio luminosities of this sample with different types of AGN suggests that this sample of   z ≳ 2  radio-intermediate obscured quasars shows radio properties that are more similar to those of the high-radio-luminosity end of the low-redshift radio-quiet quasar population than those of Fanaroff–Riley type I (FR I) radio galaxies. This conclusion may reflect intrinsic differences, but could be strongly influenced by the increasing effect of inverse-Compton cooling of extended radio jets at high redshift.     

A unified model for the evolution of galaxies and quasars

We incorporate a simple scheme for the growth of supermassive black holes into semi-analytic models that follow the formation and evolution of galaxies in a cold dark matter-dominated Universe. We assume that supermassive black holes are formed and fuelled during major mergers. If two galaxies of comparable mass merge, their central black holes coalesce and a few per cent of the gas in the merger remnant is accreted by the new black hole over a time-scale of a few times 10⁷ yr. With these simple assumptions, our model not only fits many aspects of the observed evolution of galaxies, but also reproduces quantitatively the observed relation between bulge luminosity and black hole mass in nearby galaxies, the strong evolution of the quasar population with redshift, and the relation between the luminosities of nearby quasars and those of their host galaxies. The strong decline in the number density of quasars from z ∼2 to z =0 is a result of the combination of three effects: (i) a decrease in the merging rate; (ii) a decrease in the amount of cold gas available to fuel black holes, and (iii) an increase in the time-scale for gas accretion. The predicted decline in the total content of cold gas in galaxies is consistent with that inferred from observations of damped Ly α systems. Our results strongly suggest that the evolution of supermassive black holes, quasars and starburst galaxies is inextricably linked to the hierarchical build-up of galaxies.     

The far-infrared–submillimetre spectral energy distribution of high-redshift quasars

We combine photometric observations of high-redshift     quasars, obtained at submillimetre to millimetre wavelengths, to obtain a mean far-infrared (rest-frame) spectral energy distribution (SED) of the thermal emission from dust, parametrized by a single temperature ( T ) and power-law emissivity index ( β ). The best-fitting values are     and     . Our method exploits the redshift spread of this set of quasars, which allows us to sample the SED at a larger number of rest wavelengths than is possible for a single object: the wavelength range extends down to ∼60 μm, and therefore samples the turnover in the greybody curve for these temperatures. This parametrization is of use to any studies that extrapolate from a flux at a single wavelength, for example to infer dust masses and far-infrared luminosities.
We interpret the cool, submillimetre component as arising from dust heated by star formation in the host galaxy of the quasar, although we do not exclude the presence of dust heated directly by the active galactic nucleus (AGN). Applying the mean SED to the data, we derive consistent star formation rates ∼1000 M_⊙ yr⁻¹ and dust masses ∼10⁹ M_⊙, and investigate a simple scheme of AGN and host galaxy co-evolution to account for these quantities. The time-scale for formation of the host galaxy is     , and the luminous quasar phase occurs towards the end of this period, just before the reservoir of cold gas is depleted. Given the youth of the Universe at     (1.6 Gyr), the coexistence of a massive black hole and a luminous starburst at high redshifts is a powerful constraint on models of quasar host galaxy formation.     

Simulating wide-field quasar surveys from the optical to near-infrared

A number of deep, wide-field, near-infrared (NIR) surveys employing new infrared cameras on 4-m class telescopes are about to commence. These surveys have the potential to determine the fraction of luminous dust-obscured quasars that may have eluded surveys undertaken at optical wavelengths. In order to understand the new observations it is essential to make accurate predictions of surface densities and number–redshift relations for unobscured quasars in the NIR based on information from surveys at shorter wavelengths. The accuracy of the predictions depends critically on a number of key components. The commonly used single power-law representation for quasar spectral energy distributions (SEDs) is inadequate and the use of an SED incorporating the upturn in continuum flux at  λ∼ 12 000 Å  is essential. The presence of quasar host galaxies is particularly important over the rest-frame wavelength interval  8000 < λ < 16 000 Å  and we provide an empirical determination of the magnitude distribution of host galaxies using a low-redshift sample of quasars from the Sloan Digital Sky Survey Data Release 3 quasar catalogue. A range of models for the dependence of host galaxy luminosity on quasar luminosity is investigated, along with the implications for the NIR surveys. Even adopting a conservative model for the behaviour of host galaxy luminosity the number counts for shallow surveys in the K band increase by a factor of 2. The degree of morphological selection applied to define candidate quasar samples in the NIR is found to be an important factor in determining the fraction of the quasar population included in such samples.     

Implications for unified schemes from submillimetre and far-infrared follow-up of radio-selected samples

We extend our previous analysis which used generalized luminosity functions (GLFs) to predict the number of quasars and galaxies in low-radio-frequency-selected samples as a function of redshift, radio luminosity, narrow-emission-line luminosity and type of unified scheme. Our extended analysis incorporates the observed submillimetre (850-μm) flux densities of radio sources, employs a new method which allows us to deal with non-detections, and focuses on the high-luminosity population. First, we conclude that the submillimetre luminosity L ₈₅₀ of low-frequency-selected radio sources is correlated with the bolometric luminosity L _bol of their quasar nuclei via an approximate scaling relation   L ₈₅₀∝ L ^0.7±0.2_bol  . Secondly, we conclude that there is quantitative evidence for a receding-torus-like physical process for the high-luminosity population within a two-population unified scheme for radio sources; this evidence comes from the fact that radio quasars are brighter in both narrow emission lines and submillimetre luminosity than radio galaxies matched in radio luminosity and redshift. Thirdly, we note that the combination of a receding-torus-like scheme and the assumption that the observed submillimetre emission is dominated by quasar-heated dust yields a scaling relation   L ₈₅₀∝ L ^1/2_bol  which is within the errors of that determined here for radio-selected quasars, and consistent with that inferred for radio-quiet quasars.     

On the lack of cold dust in IRAS P09104+4109 and IRAS F15307+3252: their spectral energy distributions and implications for finding dusty AGNs at high redshift

We present upper limits on the 850-μm and 450-μm fluxes of the warm hyperluminous (bolometric luminosity     galaxies IRAS P09104+4109     and IRAS F15307+3252     , derived from measurements using the SCUBA bolometer array on the James Clerk Maxwell Telescope. Hot luminous infrared sources like these are thought to differ from more normal cold ultraluminous infrared     galaxies in that they derive most of their bolometric luminosities from dusty active galactic nuclei (AGNs) as opposed to starbursts. Such hot, dusty AGNs at high redshift are thought to be responsible for much of the mass accretion of the Universe that is in turn responsible for the formation of the supermassive black holes seen in the centres of local galaxies. The galaxy IRAS P09104+4109 is also unusual in that it is a cD galaxy in the centre of a substantial cooling-flow cluster, not an isolated interacting galaxy like most ultraluminous infrared galaxies. Previously it was known to have large amounts of hot     dust from IRAS observations. We now show that the contribution of cold dust to the bolometric luminosity is less than 3 per cent. Most ultraluminous infrared galaxies possess large amounts of cold dust, and it is now known that some cooling-flow cluster cD galaxies do as well. Yet this object, which is an extreme example of both, does not have enough cold gas to contribute significantly to the bolometric luminosity. We outline physical reasons why this could have happened. We then provide a discussion of strategies for finding hot dusty AGNs, given the limitations on submillimetre surveys implied by this work.     

Optical and near-infrared imaging of faint Gigahertz Peaked Spectrum sources

A sample of 47 faint Gigahertz Peaked Spectrum (GPS) radio sources selected from the Westerbork Northern Sky Survey (WENSS) has been imaged in the optical and near-infrared, resulting in an identification fraction of 87 per cent. The R  −  I R  −  K colours of the faint optical counterparts are as expected for passively evolving elliptical galaxies, assuming that they follow the R -band Hubble diagram as determined for radio-bright GPS galaxies. We find evidence that the radio spectral properties of the GPS quasars are different from those of GPS galaxies. The observed distribution of radio spectral peak frequencies for GPS sources optically identified with bright stellar objects (presumably quasars) is shifted compared with GPS sources identified with faint or extended optical objects (presumably galaxies), in the sense that a GPS quasar is likely to have a higher peak frequency than a GPS galaxy. This means that the true peak frequency distribution is different for the GPS galaxies and quasars, because the sample selection effects are independent of optical identification. The correlation between peak frequency and redshift that has been suggested for bright sources has not been found in this sample; no correlation exists between R magnitude (and therefore redshift) and peak frequency for the GPS galaxies. We therefore believe that the claimed correlation is actually caused by the dependence of the peak frequency on optical host, because the GPS galaxies are generally at lower redshifts than the quasars. The difference in the peak frequency distributions of the GPS galaxies and quasars is further evidence against the hypothesis that they form a single class of object.     

The local far-infrared galaxy colour–luminosity distribution: a reference for BLAST and Herschel/SPIRE submillimetre surveys

We measure the local galaxy far-infrared (FIR) 60 to 100 μm colour–luminosity distribution using an all-sky IRAS survey. This distribution is an important reference for the next generation of FIR–submillimetre surveys that have and will conduct deep extragalactic surveys at 250–500 μm. With the peak in dust-obscured star-forming activity leading to present-day giant ellipticals now believed to occur in submillimetre galaxies near   z ∼ 2.5  , these new FIR–submillimetre surveys will directly sample the spectral energy distributions of these distant objects at rest-frame FIR wavelengths similar to those at which local galaxies were observed by IRAS . We have taken care to correct for the temperature bias and the evolution effects in our IRAS 60-μm-selected sample. We verify that our colour–luminosity distribution is consistent with the measurements of the local FIR luminosity function, before applying it to the higher redshift Universe. We compare our colour–luminosity correlation with recent dust–temperature measurements of submillimetre galaxies and find evidence for pure luminosity evolution of the form  (1 + z )³  . This distribution will be useful for the development of evolutionary models for Balloon-borne Large Aperture Submillimeter Telescope (BLAST) and Spectral and Photometric Imaging Receiver (SPIRE) surveys as it provides a statistical distribution of the rest-frame dust temperatures for galaxies as a function of luminosity.