EDP Sciences
Free Access
Issue
A&A
Volume 533, September 2011
Article Number A79
Number of page(s) 34
Section Catalogs and data
DOI https://doi.org/10.1051/0004-6361/201117328
Published online 30 August 2011

© ESO, 2011

1. Introduction

Blazars are an extreme class of active galactic nuclei (AGN) characterized by high luminosity, rapid variability, and high polarisation. In the radio band, blazars are core-dominated objects with apparent superluminal speeds along relativistic jets pointed close to the observer’s line of sight. Blazars have flat spectral indices and include flat spectrum radio quasars (FSRQs) and BL Lacertae objects, the counterparts of high- and low-luminosity radio galaxies. Owing to their orientation with respect to the line of sight, blazars represent less than 5% of all AGN, a quite rare class of sources.

The first blazar samples were selected at relatively high limiting flux densities in the radio or X-ray band,  ~1 Jy or a few times 10-13 erg cm-2 s-1, respectively (Stickel et al. 1991; Wood et al. 1984; Kollgaard et al. 1996; Gioia et al. 1990; Stocke et al. 1991; Perlman et al. 1996).

In the last decade of the past century, many efforts have been made to select samples of blazars that could be representative of the whole population (Marchã 1994; Laurent-Muehleisen et al. 1999; Caccianiga et al. 1999; Bondi et al. 2001). A deeper, more comprehensive sample of blazars has been constructed by Perlman et al. (1998) and by Landt et al. (2001): the “Deep X-ray Radio Blazar Survey” (DXRBS). The DXRBS has been constructed by cross-correlating all ROSAT sources of the WGCAT catalogue (White et al. 1995) and radio sources with flat radio spectra. The radio flux densities have been taken from available catalogues, like the Green Bank GB6 (Gregory et al. 1996), NORTH20CM (White & Becker 1992), and Parkes-MIT-NRAO PMN (Griffith & Wright 1993).

The DXRBS sample is currently the faintest, down to  ~50 mJy at 5 GHz and power  ~1024 W Hz-1, and most extensive blazar sample with nearly complete optical spectroscopic identifications. The radio spectral information of the DXRBS sources is based on non-simultaneous measurements at just two frequencies. Simultaneous flux density measurements at several frequencies are needed to obtain more reliable spectral indices.

In this paper we present the results from the simultaneous multi-frequency campaign made with the Effelsberg 100-m telescope. In Sect. 2 we summarise the observations and data processing. In Sect. 3 we present results from the observations. Conclusions are presented in Sects. 4 and 5.

2. Observations and data reduction

The sample of faint blazars, subject of the present investigation, was constructed from the DXRBS catalogue. We selected all sources with Dec  ≥ −20° and obtained a sample of 103 blazars, 74 of which were originally classified as FSRQs, 17 as SSRQs, and 12 as BL Lac obiects. They represent a complete sample of objects with known optical spectroscopic identifications. Redshifts are available for 99 out of 103 sources.

We have made simultaneous flux density measurements with the Effelsberg 100-m telescope at 2.64 GHz, 4.85 GHz, 8.35 GHz, and/or 10.45 GHz of the full sample of 103 sources.

The observations were carried out in the period 1–6 July 2009. All sources in the sample are point-like to the Effelsberg telescope beams. They are also bright enough to be observed by cross-scanning along the azimuth and elevation axes, with four to eight subscans to determine the total intensity and polarisation characteristics.

Details about the observation mode, calibration and evaluation of flux density and polarised emission errors can be found in Mantovani et al. (2009).

2.1. Source coordinates and flux density measurements

The positions for the sources listed in the DXRBS catalogue are taken from the GB6 survey for 0° <  Dec  < 70° or PMN surveys for −20° <  Dec  < 0°. To improve the positional accuracy even more, we have derived the positions of the DXRBS objects from the FIRST survey (Becker et al. 1994) and from the NVSS (Condon et al. 1998) at 1.4 GHz. Images of a 15′ × 15′ field centred at the GB6 or PMN positions were extracted from the two surveys, and new coordinates were obtained using the AIPS task JMFIT. Positions for 63 DXRBS objects were obtained from the FIRST catalogue. For the remaining sources the coordinates were from the NVSS. The new coordinates are reported in Table 1 together with flux density measurements at 1.4 GHz from the NVSS and our own flux density measurements with the Effelsberg 100-m telescope. The source 3C 286 was observed as flux density and polarisation calibrator. The flux density measurements at all available frequencies are on the Baars et al. (1977) scale.

3. Results

Originally the spectral indices in DXRBS were defined from flux density measurements at 1.4 GHz and 5 GHz from the existing catalogues. This approach, however, causes a major problem. Blazars are variable sources at all bands of the electromagnetic spectrum. Measurements at different frequencies obtained at different epochs may cause a misleading classification of the spectral indices. Consequently, we performed accurate flux density measurements of the DXRBS sample at several of the frequencies available at the Effelsberg 100-m telescope.

Table 1

Flux densities at the available frequencies.

3.1. Spectral index behaviour

Single-epoch spectral indices are determined making use of flux density measurements taken at 2.64 GHz, 4.85 GHz, 8.35 GHz, and/or 10.45 GHz. Flux densities at 1.4 GHz were taken from the NVSS. We make the assumption that sources alter very little in flux density with time at frequencies lower than about 2 GHz (Padrielli et al. 1987). The spectral indices (α averaged over the whole spectra), have been classified as follows:

  • a)

    “steep” spectral index when α ≤ −0.5; “steep +” for α < −0.7;

  • “flat” spectral index when α > −0.5;

  • “steep-flat” spectral index when α is steep at lower frequencies and flattens at higher frequencies;

  • “flat-steep” spectral index when α is flat at lower frequencies and steep at higher frequencies;

  • a giga-Hertz peaked (“GPS”) spectral index when the shape is convex with a peak at about the intermediate observing frequencies;

  • “inverted” spectrum when the spectral index shape is mainly straight and the flux density clearly increases with frequency.

Examples of spectral index plots are shown in Fig. 1. The full set of spectral index plots is presented in Appendix A. A crude classification of the spectral index behaviour is summarized in Table 2.

The spectral index type is listed in Table 3. The α values between subsequent frequencies can be found in the spectral index plots in Appendix A. For the source J0513.8+0156 only one flux density measurement is available due to technical problems during the observations.

Table 2

Spectral index classification.

Table 3

Variability and spectral class.

3.2. Flux density variability

Observations made at 4.85 GHz have also allowed us to look for any flux density variability through a comparison with the measurements available in the GB6 and PMN catalogues.

The GB6 survey has been built from observations performed in 1986–87. The catalogue contains sources with S4.85   GHz > 18 mJy. A 20 mJy source has a typical error of 4–5 mJy. The PMN survey was conducted in 1990 using the Parkes 64-m radiotelescope and the NRAO seven-beam receiver at 4.85 GHz covering the whole sky in the declination range −87° < δ < 10°. The PMN survey in its equatorial region, which is the one relevant for the comparison with our southern sources, has a flux density limit of 40 mJy. The sources in our sample observed by Effelsberg at 4.85 GHz and listed in the GB6 and PMN catalogues are 67 and 35, respectively. Effelsberg measurements are reported in Table 1, while GB6 and PMN measurements are listed in Table 3.

Firstly, we checked for any possible systematic difference between our measurements and those from the GB6 and PMN catalogues. In Fig. 3 we plot the Effelsberg flux densities versus the GB6 (points) or PMN (triangle) flux densities at 5 GHz. The flux density measurements are well-placed around the equal flux density line, with a scatter given by the variability. More quantitatively, the mean value and dispersion for the ratio between the Effelsberg and GB6 flux densities are 1.00 and 0.07. From the comparison between the Effelsberg and the PMN flux density we obtain a slightly lower value with a much higher dispersion 0.97 and 0.54, respectively. This is consistent with the original finding from the comparison between the PMN and GB6 surveys.

For the sources in the GB6 catalogue we can compare observations made at three epochs: our data taken in 2009 and the two measurements made in 1986 and 1987, which were used for the GB6 catalogue. We used the JAVA applet available at the GB6 catalogue web page (http://pulsar.phas.ubc.ca/) to retrieve the flux densities at epochs 1986 and 1987 for all our sources with declination  >0°.

The long-term variability in the GB6 catalogue is parametrized using the following quantity: where σ is the combined error of the two flux density measurements. Out of the 67 sources in common, four do not possess a calculated R value because they lack either the 1986 or 1987 flux density measurement. We calculated the same quantity R using our 2009 Effelsberg observations and the 1987 measurements from the GB6 survey and we plot the outcome in Fig. 4. We consider variables the sources with R > 2. We find that 55% of the sources (35 objects) do not show significant variations both in the time intervals 1986 to 1987 and 1987 to 2009, 11% of the sources (seven objects) were found variable between 1986–1987 compared to 40% (twentyfive objects) of variable sources between 1987 and 2009; finally 5% of the sources (three objects) showed significant variability between 1986 and 1987 but were found non-variable on the longer time 1987–2009. Considering also the four objects for which we do not have the R86−87 values, we have a total of 28 radio sources (42%) with R87−09 > 2 and therefore, according to the threshold adopted, variable in a time interval of 22 years.

We calculated R also for the 35 sources with δ ≤ 0 that have the PMN measurements and we find that 45% of the sources (16 objects) have R90−09 > 2. This fraction is consistent with that of the northern sources and we can conclude that 43% of the DXBRS sample presented in this paper show significant variability on a time interval of about 20 years. The quantity R is reported in Table 3.

3.3. Polarised emission

The Effelsberg observations were carried out in full polarisation. We consider a source to be polarised when the intensity of the polarised emission is three times the rms error estimated for the polarised emission source by source. Results are presented in Table 4.

Table 4

Percentage of polarised emission, electric vector position angles, and rotation measures.

At 1.4 GHz, 27 sources show a polarised emission below the detection limits according to the data extracted from the NVSS. One source in our list was not found by the extraction process. The remaining sources have a median value of the fractional polarisation m of %.

At 2.64 GHz we found 25 sources polarised with a median value of m of 4.9 %. At 4.85 GHz we found 36 sources polarised with a median m of (5.8 ± 0.9)%. These integrated values are considered typical of AGN. They can be compared with those achieved by Mantovani et al. (2009) on a sample of compact steep-spectrum sources and by Klein et al. (2003) on a sample of steep spectrum extended radio sources selected from the B3-VLA sample, as shown in Table 5. The trend is a quick decrease in the fractional polarisation with increasing wavelength. However, cases of repolarisation, i.e. an increase of fractional polarisation towards longer wavelengths, were reported, like that for the source 3C 455 (see Mantovani et al. 2009), worth of further investigation. Repolarisation is seen for thirteen faint blazars in our sample, which show a peak of m at about 2.6 GHz. Systematic instrumental effects to possibly explain this behaviour can be excluded. More accurate observations are required to confirm this trend of the fractional polarisation.

Table 5

Comparison between percentage of polarised emission for different samples of objects.

Rotation measure could be calculated for 27 sources, of which seventeen are classified as bona fide blazars (FSRQs), nine as SSRQs, and one as a GPS object, according to our flux density measurements. The RM values are in the range from 0 to 1950 rad m-2 in the source rest frame (i.e. RM redshift corrected). The majority of the sources show a quite low RM. No clear correlation of m with the spectral index type (SSRQs or FSRQs) could be found. An example of RM and m plot is presented in Fig. 2. The full set of plots can be found in Appendix B.

4. Comments on individual sources

There are a few sources worth of brief comments.

  • J0204.8+1514 –

    Adding 180° to the EVPA at 1.4 GHz produces just a slightly worse fit. In this case the RM is −22.5 rad m-2.

  • J1025.9+1253 –

    A fit of equal goodness and a RM of −85.5 rad m-2 results from rotation of the EVPA at 2.64 GHz by 180° and the EVPA at 1.4 GHz by 360°.

  • J0937.1+5008 –

    The source J0937.1+5008 exhibits a high degree of variability at 5 GHz, of the order of 60% to 70% on a time scale of about 22 years comparing the Effelsberg flux density and the GB6 flux density. Moreover, it shows an increase in flux density by a factor of 2.3 comparing the Effelsberg flux density with the flux density measured few months later (22 Oct. 2009) with the European VLBI Network observations at the same frequency (Mantovani et al., in prep.). Short-term variability was not found comparing the two measurements made in 1986 and 1987 by the GB6. A preliminary milli-arcsecond resolution image of J0937.1+5008 (resolution better than 5 mas; rms noise 1.5 mJy/beam) shows that the source is point-like (Mantovani et al. 2010). J0937.1+5008 was detected by the Fermi Gamma-Ray Space Telescope (Abdo et al. 2010). This source presents a peculiar “concave” spectral index as shown in Fig. 5. A similar spectral index shape is also shown by the source J1028.6–0336.

    thumbnail Fig. 1

    Examples of spectral index behaviour, namely: “steep” (J0447.9−0322), “flat” (J0110.5−1647), “steep-flat” (J0847.2+1133), “flat-steep” (J0210.0−1004), “GPS” (J0227.5−0847), “inverted” (J1028.5−0236).

    Open with DEXTER

  • J1626.6+5809 –

    Removing the EVPA at 1.4 GHz, a good fit is achieved using the three Effelsberg measurements and a rotation by 180° at 2.64 GHz. The RM in this case is −16.4 rad m-2.

  • J1648.4+4104 –

    A better fit to the data can be achieved by rotating the EVPA at 1.4 GHz by 360°. The RM is now 170 rad m-2.

thumbnail Fig. 2

Examples of position angles of the electric vector χ (dots) and fractional polarisation m (triangles) versus λ2 plot. σ values assess the quality of the best fit.

Open with DEXTER

thumbnail Fig. 3

Effelsberg flux densities versus GB6 (points) or PMN (triangles) flux densities at 5 GHz. The straight line means a ratio of 1.

Open with DEXTER

thumbnail Fig. 4

The R values for both short and long-term variability.

Open with DEXTER

thumbnail Fig. 5

The “concave” spectral index of the source J0937.1+5008.

Open with DEXTER

5. Conclusions

Our investigation aims at veryfing the spectral index classification of the DXRBS sample. We define a spectral index α ≤ −0.5 as “steep”. If α > −0.5 it is considered “flat”. However, the spectra often show a complex shape, i.e. not a steep or flat straight line. This makes the classification more difficult. We have attempted a “crude” classification source by source in Table 3, while spectral indices are available in the plots we produced for each source (see Appendix A).

The majority of sources, 66 out of 102 (for one source we do not have enough measurements) can be considered bona fide blazars, namely those with “flat”, “steep-flat”, “inverted”, and possibly also “flat-steep” spectra in the present classification.

Several sources changed their spectral index classification with respect to that reported in the original list by Perlman et al. (1998) and by Landt et al. (2001). Among them we have nine sources with a “convex” spectral index peaking in the GHz regime, which should be more properly defined as giga-Hertz peaked sources (GPS). These sources cannot be considered blazars. We found that 17 sources previously classified as FSRQs are actually SSRQs. Moreover, we found six sources with spectral indices even steeper than −0.7, failing one of the original selection criteria. In total, we found that 27 sources can be classified as SSRQs showing a spectral index  <−0.5. This result suggests that a statistically meaningful comparison between FSRQs and SSRQs in the same flux density limited complete sample is possible.

We can anticipate that all 42 sources belonging to our sample observed so far with the EVN at 5 GHz have been detected (Mantovani et al., in prep.). Six sources, previously classified SSRQs (and found with α < −0.7), host AGN. Two of them show a core-jet structure, while the remaining four are point-like at a resolution better than 5 milliarcsec. The EVN to Effelsberg flux density ratio ranges from 0.05 to 0.8. Most of their radio emission is therefore not coming from the cores of these sources.

Fourty-three percent of the DXRBS sample show significant variability on a time scale of 19–22 years; 11% show short-term variability comparing Green Bank flux density measurements taken one year apart in 1986 and 1987.

About 25% of the sources show polarised emission above the detection limits of our observations. We also extracted polarimetric information from the NVSS to collect at least the three measurements needed to compute the RM diminishing the  ambiguity. The majority of the sources show |RM|  < 200 rad m-2 in the source rest frame, possibly produced by the Faraday depth of the Galactic foreground. In nine cases, namely J0029.0+0509, J0322.6−1335, J0434.3−1443, J0510.0+1800, J0518.2+0624, J0744.8+2920, J1400.7+0425, J1419.1+0603, and J1648.4+4104 the rest frame RM are in the range of 200  <    |RM|    <  1940 rad m-2. The highest value found is 1938.8 rad m-2 for the source J0434.3−1443.

Acknowledgments

We thank an anonymous referee for his/her very helpful comments and suggestions, and for a careful reading of the manuscript of this paper. This work is based on observations with the 100-m telescope of the MPIfR (Max-Planck-Institut für Radioastronomie) at Effelsberg. It has benefited from research funding from the European Community’s Framework Programme under RadioNet R113CT 2003 5058187. F.M. likes to thank Anton Zensus for the kind hospitality at the Max-Planck-Institut für Radioastronomie, Bonn, for a period during which part of this work has been done. The authors like to thank Heinz Andernach for suggesting to check for short-term flux density variability making use of Green Bank measurements. 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. Part of this work was supported by the: COST Action MP0905 “Black Holes in a Violent Universe”.

References

Appendix A: Spectral index plots

thumbnail Fig. A.1

Spectral index plots of sources in Table 3.

Open with DEXTER

Appendix B: RM and m plots

thumbnail Fig. B.1

Position angles of the electric vector χ (dots) and fractional polarisation m (triangles) versus λ2 plots of sources in Table 4. σ values assess the quality of the best fit.

Open with DEXTER

All Tables

Table 1

Flux densities at the available frequencies.

Table 2

Spectral index classification.

Table 3

Variability and spectral class.

Table 4

Percentage of polarised emission, electric vector position angles, and rotation measures.

Table 5

Comparison between percentage of polarised emission for different samples of objects.

All Figures

thumbnail Fig. 1

Examples of spectral index behaviour, namely: “steep” (J0447.9−0322), “flat” (J0110.5−1647), “steep-flat” (J0847.2+1133), “flat-steep” (J0210.0−1004), “GPS” (J0227.5−0847), “inverted” (J1028.5−0236).

Open with DEXTER
In the text
thumbnail Fig. 2

Examples of position angles of the electric vector χ (dots) and fractional polarisation m (triangles) versus λ2 plot. σ values assess the quality of the best fit.

Open with DEXTER
In the text
thumbnail Fig. 3

Effelsberg flux densities versus GB6 (points) or PMN (triangles) flux densities at 5 GHz. The straight line means a ratio of 1.

Open with DEXTER
In the text
thumbnail Fig. 4

The R values for both short and long-term variability.

Open with DEXTER
In the text
thumbnail Fig. 5

The “concave” spectral index of the source J0937.1+5008.

Open with DEXTER
In the text
thumbnail Fig. A.1

Spectral index plots of sources in Table 3.

Open with DEXTER
In the text
thumbnail Fig. B.1

Position angles of the electric vector χ (dots) and fractional polarisation m (triangles) versus λ2 plots of sources in Table 4. σ values assess the quality of the best fit.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.