text stringlengths 128 2.05k |
|---|
With the development of evolutionary models for radio sources came the idea of a dual-population model. The initial version of this dichotomy |
(Longair, 1966 was in terms of low and high luminosities. But with high-frequency surveys, and the large number of flat/inverted spectrum sources revealed in them, an alternative classification emerged, based exclusively on the source spectra: sources with a spectral index [MATH] (where [MATH] ), corresponding to optic... |
(Schmidt, 1976 ; Masson & Wall, 1977 . However, Dunlop & Peacock ( 1990 studied the radio luminosity function (density of sources with a given luminosity per unit of co-moving volume) of these two classes of radio sources, and came to the conclusion that both populations were undergoing a similar evolution, implying th... |
The Fanaroff-Riley (FR) classification (Fanaroff & Riley, 1974 , originally a sub-classification for steep-spectrum objects, was employed to provide a further categorization of radio sources. This classification divides radio sources into two classes of double-lobed sources based on the appearance of their jets. The FR... |
(Rawlings et al., 1989 ; Baum & Heckman, 1989 Fanaroff & Riley ( 1974 found these two classes to be divided in radio power, with a break luminosity [MATH] , with FRII sources lying above this limit. Subsequently Owen & Ledlow ( 1994 showed that the break was a function of both radio and optical luminosity. |
During the 1980s the ’unification’ hypothesis emerged to describe how viewing aspect could relate RQSOs, (Radio Quasi-Stellar Objects of either flat or steep-spectrum) to FRII radio galaxies |
(e.g. Peacock, 1987 ; Scheuer, 1987 ; Barthel, 1989 . However, the scheme did not include lower-luminosity AGNs such as FRI galaxies and BL Lac objects. The unifying connection between these was introduced by |
Marcha & Browne ( 1995 . The unified model of AGN proposed by Wall & Jackson ( 1997 and Jackson & Wall ( 1999 assumes that the cosmic evolution of radio loud AGN is based on a division of the radio sources into a low-luminosity ( [MATH] component corresponding to FRIs, and a high-luminosity component corresponding to F... |
Initially, in modelling the space density of radio AGN, it was assumed for simplicity that the low-luminosity radio galaxies including FRI sources showed no cosmic evolution (Wall, Pearson & Longair, 1980 ; Jackson & Wall, 1999 , the strong cosmic evolution confined only to the higher luminosities and the FRII galaxies... |
Snellen & Best ( 2001 Willott et al. ( 2001 Sadler et al. ( 2007 and Rigby, Best & Snellen ( 2007 , found significant evolution for low power sources – but mild evolution in comparison with that of the high-luminosity sources. Rigby, Best & Snellen ( 2007 argued that if both FRIs and FRIIs have similar evolution, the d... |
The FRI/FRII dual-population scheme has encountered several problems. One of these concerns the correspondence between FRI galaxies and BL Lac objects. Urry & Padovani ( 1995 noted that some BL Lac objects have non-FRI-like morphologies and that the density of FRI sources is too low to account for the entire BL Lac pop... |
demonstrate that only about one third of low-luminosity core dominated radio sources - which are supposedly the beamed counterpart of FRI sources - are conventional BL Lac objects. Most of the remaining sources have optical classification such as Seyfert objects or elliptical galaxies. |
A related issue concerns the existence of FRI RQSOs. Until recently, these objects were thought not to exist, leading to the hypothesis that FRI and FRII central engines were of different nature (Baum et al., 1995 and that the torus opening in FRI sources was too small to observe a quasar nucleus |
(Falcke et al., 1995 . However, the discovery of an FRI QSO, E1821+643, by Blundell & Rawlings ( 2001 overthrew those assumptions. More recent VLA observations (Heywood, Blundell & Rawlings, 2007 uncovered another 4 sources of this type. |
Finally, if sources with different FR classes undergo different evolution, this might imply that their fundamental characteristics, such as the black hole spin or jet composition, are different too. However, the existence of hybrid sources, which display both FRI and FRII morphological characteristics (Capetti et al., ... |
argued that the FR dichotomy is based purely on the interaction between the jets and the environmental medium, and not on intrinsic properties of the central engine. This view is also supported by |
Gopal-Krishna & Wiita ( 2000 and Gawronski et al. ( 2006 . However, Wang et al. ( 1992 suggested that some AGN engines could be capable of ejecting jets of unequal power, resolving the problem of hybrid sources. Gopal-Krishna & Wiita ( 2000 |
found no evidence for such a process in their sample. Other schemes have been suggested to resolve these difficulties. Kaiser & Best ( 2007 explained the FR dichotomy by postulating that all sources start as FRII objects and used an analytical model in which the evolution of the radio sources is governed by energy loss... |
Willott et al. ( 2001 used an approach to a dual-population unified scheme based on the luminosity of sources instead of morphology. Optical spectra of FRII sources are heterogeneous and they can display both strong and weak low-excitation emission lines |
(Laing, 1994 . Therefore, radio sources can also be grouped based on their emission lines, with one population composed of low-luminosity sources having weak emission lines (containing both FRI and FRII objects), and the other composed of high-luminosity sources with strong emission lines (containing only FRII objects)... |
(Hardcastle et al., 2006 Defining the relation between the different radio sub-populations together with their cosmic evolution is becoming fundamental to our understanding of galaxy formation. |
The current paradigm for galaxy formation follows hierarchical build-up in a Cold Dark Matter (CDM) universe. Nevertheless, serious difficulties arise from this model in its simplest form, as discussed by Bower et al. ( 2006 . It implies that current epoch galaxies must be the largest and bluest and have the highest st... |
(Silk & Rees, 1998 ; Granato et al., 2001 ; Quilis et al., 2001 . Note that there is also some positive AGN feedback (van Breugel et al., 2004 ; Klamer et al., 2004 in which the pressure from the jets compresses the inter-stellar medium and induces star formation. The balance between these processes remains to be under... |
Although FRI and FRII sources show different evolution, they also lie in different luminosity ranges. There is therefore a possibility that both types may show similar evolution for overlapping luminosities (i.e. high-luminosity FRIs and low-luminosity FRIIs). In order to sort out the FR dichotomy and its details, accu... |
The CoNFIG sample is defined as all sources with [MATH] [MATH] 1.3 Jy from the NRAO-VLA Sky Survey (NVSS, Condon et al., 1998 catalogue within the north region of the Faint Images of Radio Sources at Twenty-cm survey (FIRST, White et al., 1997 , a 1.5 sr region defined roughly by [MATH] |
and 7 hr [MATH] 17 hr. The flux density limit of 1.3 Jy was chosen so that the number of sources in the sample was of statistical significance while allowing us to identify the morphology for each source individually. Optical identifications were obtained from the SuperCOSMOS Sky Survey (SSS, Hambly et al., 2001 and re... |
The structure of this paper is as follows. The details of the construction of the CoNFIG sample are explained in § while § describes how the morphologies were determined. Optical identifications and redshift information are discussed in § and § outlines the computation of the morphology-dependent luminosity distributio... |
Throughout this paper, we assume a standard [MATH] CDM cosmology with [MATH] =70 km s -1 Mpc -1 [MATH] =0.3 and [MATH] =0.7. The CoNFIG Sample |
2.1 Radio Surveys 2.1.1 NVSS The NRAO-VLA Sky Survey (NVSS) is a 1.4 GHz continuum survey covering the entire sky north of [MATH] declination (corresponding to an area of 10.3 sr). The completeness limit is about 2.5 mJy with an rms brightness fluctuation of about 0.45 mJy/beam. The survey has yielded over 1.8 million ... |
Since the median angular size of faint extragalactic sources at these flux density levels is [MATH] 10 arcsec (Condon et al., 1998 , most sources are unresolved, and the NVSS flux density measurements are quite accurate. However, the large beam size does not allow one to determine precise structure of sources or to det... |
2.1.2 FIRST The Faint Images of the Radio Sky at Twenty-cm (FIRST) is another 1.4 GHz continuum survey, covering an area of [MATH] 9030 [MATH] over the North Galactic Pole. The completeness limit is about 1 mJy with a typical rms of 0.15 mJy. The survey yielded [MATH] 811,000 sources, implying a surface density of [MAT... |
(White et al., 1997 . This survey complements the NVSS survey well, providing a beam size small enough to resolve the structure of most nearby extended radio sources. |
2.2 Sample Definition The CoNFIG sample is defined as all sources with [MATH] [MATH] 1.3 Jy from the NVSS catalogue within the north region of the FIRST survey (1.5 sr region defined roughly by [MATH] |
and 7 hr [MATH] 17 hr), resulting in a sample of 261 objects. Very large sources resolved in NVSS within this initial sample, such as 3CRR sources (Laing, Riley & Longair, 1983 , need to be considered. In these cases, two or more NVSS sources with [MATH] [MATH] 1.3 Jy are actually components of a much larger resolved s... |
Multi-component sources, where each component has [MATH] [MATH] 1.3 Jy but with a total flux density [MATH] [MATH] 1.3 Jy, also need to be considered. For this purpose, NVSS sources with 0.1 Jy [MATH] [MATH] [MATH] 1.3 Jy were selected and, if any other source in the catalogue was located within 4 arcmin of the listed ... |
Finally, after looking at the NVSS and FIRST contour plots and optical information of all the sources, 7 sources were deleted, either because the source is a globular cluster, or because the source actually consists of independent sources, each with |
[MATH] [MATH] 1.3 Jy. The final sample, referred to as the CoNFIG Sample, consists of 274 sources. Details of the sources can be found in Appendix |
Morphology 3.1 Initial Classification 76 sources in the sample appear in the 3CRR catalogue (Laing, Riley & Longair, 1983 in which the morphology type of each 3CRR source has been determined. These sources are marked by a star next to their 3C name in Appendix |
For other sources, the morphology was determined by looking at the FIRST and NVSS radio contour plots, obtained from the NASA virtual observatory Skyview ( |
). The contour plots are shown in Figure and Appendix If the contour plot displays distinct hot spots at the edge of the lobes (as in Figure ), and the lobes are aligned, the source was classified as FRII. On the contrary, sources with collimated jets showing hot spots near the core and jets were classified as FRI (see... |
Contour plots were obtained for all but one (CoNFIG-225) of the CoNFIG sources. 53 sources were identified as extended but the resolution is too low to confirm their exact morphological types. Additional VLA observations were obtained for 31 of these sources (see § 3.2 ) while for the other 22 sources, data were found ... |
As specified in § 2.2 , 7 sources were deleted from the sample after examination of their contour plots. In one case, the source is a globular cluster and in another case, the source has no FIRST or NVSS contours. The other 5 NVSS ‘sources’, when observed with FIRST, actually consist of separate and independent sources... |
[MATH] [MATH] 1.3 Jy. The NVSS radio positions of these sources can be found in Table 3.2 VLA Observations Radio observations of 31 extended sources with uncertain morphological classification were taken at 8 GHz using the VLA in A configuration on July 29, 2007. The observations used the standard two IFs of 8435.1 and... |
The 8 GHz flux density of each source was derived from the 1.4 GHz NVSS flux density and spectral index (see details in § ), and the exposure time was computed for each source such as to provide a signal-to-noise ratio of at least 5. The sources were placed accordingly into 4 different exposure time groups (5, 10, 20 a... |
3.3 Final Classification Data were extracted from the VLA archives at other frequencies and configurations for some sources for which the 8 GHz contours were not satisfactory. All data were reduced using standard procedures incorporated within the AIPS software provided by NRAO, and the resulting images are grouped in ... |
Over 60% of sources in the CoNFIG sample are classified either as FRI or FRII. The rest of the sources are classified as compact or deleted from the sample (see Table ). Some sources in this group will be unresolved high redshift FRIIs, and some others are confirmed as truly compact from the VLBA calibrator list |
(Beasley et al., 2002 ; Fomalont et al., 2003 ; Petrov et al., 2005 2006 ; Kovalev et al., 2007 or from the Pearson-Readhead survey (Pearson & Readhead, 1988 . Some of these confirmed compact sources show a steep ( [MATH] ) spectral index and are possible Compact Steep Sources (CSS) (Fanti & Fanti, 1994 |
After looking at optical counterparts from the SuperCOSMOS Sky Survey (as discussed in § ), two particular subclasses of compact and FRII sources are identified. |
(1) Some compact sources show no optical counterpart. These sources can be classified either as CSS or as unresolved FRII. When confirmed as a compact source from the VLBA calibrator list |
(see Beasley et al., 2002 ; Fomalont et al., 2003 ; Petrov et al., 2005 2006 ; Kovalev et al., 2007 or the Pearson-Readhead survey (Pearson & Readhead, 1988 , the source is classified as CSS and is included in the compact sources statistics. Otherwise, the source is assumed to be an unresolved FRII. Their inclusion int... |
(2) Other sources are classified as FRII and present a stellar type optical identification. These sources are - on the basis of the unified model - steep-spectrum RQSOs, which occur when the line of sight of the observer is oriented at less than 45 with respect to the jets, enabling the observer to look inside the dust... |
The final classification for each source is shown in Table and the distribution of morphological types is presented in Table . Errors in flux density measurements are well defined for point sources in the catalogues, but corresponding error estimates for the extended sources were too uncertain to be worth including her... |
3.4 CoNFIG-2, 3 and 4 In order to improve the FRI/FRII statistics, three more samples of about 250 sources each, and complete to 0.8 Jy (CoNFIG-2), 0.2 Jy (CoNFIG-3) and 50 mJy (CoNFIG-4), were constructed from the NVSS catalogue over various areas, all included in the FIRST northern region. For each sample, the morpho... |
Because the morphology information is not complete for all sources in the CoNFIG-2, 3 and 4 samples, these objects were only used to compute the FRI/FRII source count (see § 6.2 ). |
Optical Identifications and Redshifts To determine core coordinates in order to retrieve redshifts, optical identifications (together with [MATH] , R1, R2 and I magnitudes) were obtained for 79% of the sources in the CoNFIG sample and 61% of the sources in the CoNFIG-2 sample from the SuperCOSMOS Sky Survey |
(SSS, Hambly et al., 2001 Redshifts were retrieved for 230 CoNFIG sources and 161 CoNFIG-2 sources, using the SIMBAD ( ) and NED ( |
) databases. Redshift and magnitude information are listed in Table and Appendix for CoNFIG, and in Table 10 for CoNFIG -2. When spectroscopic redshifts were not available, we estimated photometric redshifts via an empirical R - z relation derived from extended sources in the CoNFIG sample with galaxy identifications a... |
[MATH] , as shown in Figure . A simple fit to these data is [EQUATION] Photometric redshift estimate is given to 14 CoNFIG objects falling into this category, but with no spectroscopic redshift information, bringing the redshift coverage to 244 sources (89% of the sample). |
For the CoNFIG sample, redshifts range from z=0.0033 to z=3.522 with a mean redshift of z=0.715 and a median redshift of z=0.580. The morphology-dependent distribution is shown in Figure |
and details of the redshift distributions are given in Tables and . The FRI distribution is concentrated at low redshifts ( [MATH] ) while the FRII distribution covers a wider range, up to z=2.5. |
Note that for 38 CoNFIG and 34 CoNFIG-2 sources, redshift information was retrieved from the literature using the radio position of the object, but no counterpart was found in SSS. |
Luminosity Distributions In order to compute the radio luminosity, the spectral index [MATH] (defined as [MATH] ) of each source needs to be determined. To achieve this, flux densities at different frequencies for each source were compiled and the spectral index computed following the relation: |
[EQUATION] A summary of the different frequencies and corresponding surveys used to retrieve the flux density information is given in Table |
We made independent estimates for the low- and high-frequency spectral indices (with [MATH] and [MATH] respectively). The low frequency spectral indices are used to compute the luminosities; since |
[MATH] , the luminosity emitted at [MATH] will correspond to an observed flux at frequency [MATH] Values for spectral indices are listed in Table . The median value is [MATH] . However, the extended sources are grouped around a median value of [MATH] as seen in Figure , while the compact sources, as is well known, show... |
Figure presents the luminosity distributions for FRI and FRII objects. The distribution of FRII sources only (excluding unresolved sources) is shown as a solid-line distribution on the plot. As stated in § 3.3 , the inclusion of the unresolved sources with the FRII sources has very little effect on the structure of the... |
FRI/FRII Source Counts One of the challenges in modelling the space densities of FRI and FRII sources is to compute an accurate source count for each morphological type. This implies determining the morphological type of all sources used. |
To increase the number of sources with morphology information, five samples at different flux density limits were combined with the CoNFIG sample. These samples include the CENSORS (Best et al., 2003 and BDFL |
(Bridle, Davis, Fomalont & Lequeux, 1972 samples, as well as the 3 extra CoNFIG samples (see details in § 3.4 and Table ). The morphologies of the sources in each sample were determined by looking at the FIRST and NVSS contour plots as described in § (with the exception of the CENSORS sources and some of the BDFL sourc... |
6.1 The CENSORS and BDFL samples 6.1.1 The CENSORS sample The CENSORS (Combined EIS-NVSS Survey Of Radio Sources) sample (Best et al., 2003 ; Brookes et al., 2005 2007 contains 150 sources complete to |
[MATH] =7.2 mJy, selected from NVSS over the ESO Imaging Survey (EIS) Patch D. The sample has a median flux density of [MATH] [MATH] 15 mJy and optical identifications for 68% of the sources. |
The wide field EIS comprises a relatively wide-angle survey of four distinct patches of sky up to 6 [MATH] each (Nonino et al., 1999 . Patch D is the most northerly, with a limiting magnitude of I [MATH] 23 |
(Benoist et al., 1999 and an area of 2 [MATH] [MATH] centred on [MATH] [MATH] (J2000). The goals of the CENSORS sample (Best et al., 2003 are to constrain evolution of the top end of the black hole mass function, study the environments around radio sources at different luminosities, test the K-z relation for radio sour... |
Little classification of the CENSORS sources has been done, although Brookes et al. ( 2007 presented a list of the possible RQSOs. To attempt a preliminary morphological classification, we used the NVSS images to determine the morphology of the sources; in most cases though, the resolution is far too low to determine t... |
[EQUATION] This relation was then applied to the 136 sources from CENSORS using the flux density and magnitude information from Best et al. ( 2003 using a K-correction value of [MATH] . As a result, 49 sources were classified as FRI and 87 as FRII. Note that, for 63 CENSORS sources, B-magnitude information were not ava... |
6.1.2 The BDFL sample The BDFL sample (Bridle, Davis, Fomalont & Lequeux, 1972 contains 424 sources and is complete to [MATH] [MATH] 1.7 Jy in the area of sky [MATH] |
[MATH] To improve definition of the high flux-density ends of the morphological source counts, sources from this sample with [MATH] [MATH] 4.0 Jy were selected, yielding a total number of 90 sources. The morphology of each source was determined from either |
Laing, Riley & Longair ( 1983 or Kharb & Shastri ( 2004 , or by looking at the NVSS contours. As a result, of the 90 sources, 21 sources are classified as FRI, 38 as FRII and 31 as compact. |
6.2 Compiling the Source Count In total, there are [MATH] 500 sources from the combined BDFL (down to 4.0 Jy), CENSORS (at 7.2 mJy) and CoNFIG (at 1.3 Jy) samples. Still, as illustrated in Figure 10 , large areas of the P-z plane are not covered. |
To improve coverage, the three other CoNFIG samples described in 3.4 were used. In all three of these samples, sources with [MATH] [MATH] 1.3 Jy were removed as they are already present in the CoNFIG sample. |
The relative differential source counts [MATH] for FRI and FRII sources were then computed from the combined sample of 244 FRI and 736 FRII sources. Data for the source count presented in Figure 11 can be found in Table 12 |
It is seen that FRII sources dominate the total count, except at low flux densities (log [MATH] [MATH] ), where the FRI sources suddenly take over. Since most of the FRI count at low flux densities is composed of low-luminosity sources at low redshift, our results show that FRI objects must undergo some mild evolution.... |
[MATH] [MATH] 1Jy. This is in agreement with previous investigations stretching back to Longair ( 1966 An illustrative exponential evolution model of space density, as described by Wall, Pearson & Longair ( 1980 , was used to fit the data, as shown in Figure 11 . In the case of both FRI and FRII sources, the luminosity... |
[MATH] , where [MATH] is the local luminosity function and [MATH] an evolution function transforming [MATH] into the redshift and luminosity-dependent [MATH] . The details of the evolution function [MATH] successfully fitting the two morphology counts (with the same set of parameters) are as follows: |
[EQUATION] where [MATH] is the redshift cutoff (maximum redshift at which a population exists) and [MATH] is the look-back time [EQUATION] |
with M(P) defined as [EQUATION] The best fit values for this model are [MATH] [MATH] [MATH] [MATH] and [MATH] This is only an illustrative model. Accurate modelling of the FRI/FRII source counts and luminosity functions will be the subject of a future study. |
The morphological counts and the model fit raise two points: (1) If a single model, albeit one with differential evolution, can describe both populations, are the populations separate, or necessarily one |
(Snellen & Best, 2001 ; Rigby, Best & Snellen, 2007 ? At present the question is a semantic one: we do not have an immediate hypothesis to test requiring FRIs and FRIIs to be one population or two. What is of interest in this is whether FRIs and FRIIs at the same radio luminosity show exactly the same evolution, and ou... |
(2) Our model is a hands-off best-fit to the composite and morphologically-divided source counts at 1.4 GHz – down to 10 mJy. It is not a valid model, because it predicts far too many sources in total at lower flux densities. A valid model needs a rapid downturn to be achieved at about 0.01 Jy, and it must be that furt... |
Summary The CoNFIG sample is constructed as a sample of 274 radio sources from NVSS with [MATH] Jy. Redshift information is available for |
[MATH] 80% of the sample, and morphological classifications were obtained for all of the sources, either from NVSS and FIRST contour plots, or, for 46 sources, from 8 GHz VLA observations. These data allow us to compute morphology-dependent luminosity distributions and source counts. |
To increase the number of sources with morphology information, three more samples were constructed in sub-areas of the main region with flux density limits of 0.8 Jy, 0.2 Jy and 50 mJy. Morphological identifications were obtained only from NVSS and FIRST contour plots for those sources. Morphological information for th... |
Our data show mild evolution of the FRI sources at low redshift; however, they do not participate in the “evolution bump” around |
[MATH] [MATH] 1 Jy. The results also support the observation that a large number of mJy sources are FRIs galaxies and not starburst galaxies as previously assumed. |
Acknowledgements We are very grateful to Jim Dunlop for many helpful suggestions, and to Rick Perley and Eric Greison for their precious help with the VLA observation and data reduction. We also thank the referee for very helpful comments. |
This work was supported by the National Sciences and Engineering Research Council of Canada. The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc. This research has made use of the SIMBAD database, operated at CDS,... |
Appendix A CoNFIG Sample Data for CoNFIG sample. The 3CRR sources from Laing, Riley & Longair ( 1983 are indicated by stars. The RA and DEC gives the NVSS position of the source. Note that 3 sources (CoNFIG-015, 076 and 225) were deleted later in the sample construction process and therefore do not appear here. |
The sources of C* type are confirmed compact sources from the VLBA calibrator list (see Beasley et al., 2002 ; Fomalont et al., 2003 ; Petrov et al., 2005 2006 ; Kovalev et al., 2007 or the Pearson-Readhead survey (Pearson & Readhead, 1988 . Sources of S* type are confirmed compact sources which show a steep ( [MATH] )... |
The values for [MATH] [MATH] [MATH] [MATH] and [MATH] were retrieved from the samples listed in Table . References for the redshift information and comments are given in the last two column. |
The spectral index [MATH] (where [MATH] ) corresponds to [MATH] as defined in . The [MATH] values with sign correspond to sources for which the value is derived from a non-satisfactory single-power-low fit of the [MATH] relation. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.