---

An optical spectroscopic survey of the 3CR sample of radio galaxies with z < 0.3

I. Presentation of the data

S. Buttiglione¹ - A. Capetti² - A. Celotti¹ - D. J. Axon³ - M. Chiaberge^4,5 - F. D. Macchetto⁴ - W. B. Sparks⁴

1 - SISSA-ISAS, via Beirut 2-4, 34014 Trieste, Italy
2 - INAF - Osservatorio Astronomico di Torino, Strada Osservatorio 20, 10025 Pino Torinese, Italy
3 - Department of Physics, Rochester Institute of Technology, 85 Lomb Memorial Drive, Rochester, NY 14623, USA
4 - Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA
5 - INAF-IRA, Via P. Gobetti 101, 40129 Bologna, Italy

Received 7 October 2008 / Accepted 18 December 2008

Abstract
We present a homogeneous and 92% complete dataset of optical nuclear spectra for the 113 3CR radio sources with redshifts <0.3, obtained with the Telescopio Nazionale Galileo. For these sources we could obtain uniform and uninterrupted coverage of the key spectroscopic optical diagnostics. The observed sample, including powerful classical FR II radio-galaxies and FR I, together spanning four orders of magnitude in radio-luminosity, provides a broad representation of the spectroscopic properties of radio galaxies. In this first paper we present an atlas of the spectra obtained, provide measurements of the diagnostic emission line ratios, and identify active nuclei with broad line emission. These data will be used in follow-up papers to address the connection between the optical spectral characteristics and the multiwavelength properties of the sample.

Key words: galaxies: active - galaxies: jets - galaxies: elliptical and lenticular, cD - galaxies: nuclei

1 Introduction

Radio galaxies (RG) are an important class of extragalactic objects for many reasons. They are extremely interesting in their own right, representing one of the most energetic astrophysical phenomena. The intense nuclear emission and collimated jets can influence the star formation history, excitation, and ionization of the ISM, representing a fundamental ingredient for the formation and evolution of their hosts but also of their larger scale environment. The host galaxies are giant early-type galaxies, often situated centrally in galaxy clusters; indeed, with their symbiotic relationship, central dominant ellipticals and the AGN phenomenon encompass many of the major cosmological and astrophysical issues we confront today. These include the ubiquity and growth of supermassive black holes, their relationship to the host, the onset of nuclear activity, the nature of their jets and the interaction with the interstellar medium, and the physics of the central engine itself. In particular, studies of radio-loud AGN are the key to understanding the processes leading to the ejection of material in relativistic jets, its connection with gas accretion onto the central black holes, the way that different levels of accretion are related to the process of jets launching, the origin of the AGN onset, and its lifetime.

The attractiveness of the 3CR catalog of radio sources as a basis for such a study is obvious, being one of the best studied sample of RG. Its selection criteria are unbiased with respect to optical properties and orientation, and it spans a relatively wide range in redshift and radio power. A vast suite of observations is available for this sample, from multi-band HST imaging, to observations with Chandra, Spitzer and the VLA, that can be used to address the issues listed above in a robust multiwavelength approach.

Quite surprisingly, despite the great interest of the community, the available optical spectroscopic data for the 3CR sample are sparse and incomplete.

To fill this gap, we carried out a homogeneous and complete survey of optical spectroscopy. We targeted the complete subsample of 113 3CR radio sources with z < 0.3, for which we can obtain uniform uninterrupted coverage of the key spectroscopic optical diagnostics. The observed sources include a significant number of powerful classical FR II RG, as well as the more common (at low redshift) FR Is (Fanaroff & Riley 1974), spanning four orders of magnitude in radio luminosity, thus providing a broad representation of the spectroscopic properties of radio-loud AGN.

The primary goals of this survey are (we defer the detailed list of bibliographical references to the follow-up papers dealing with the specific issues):

-

to measure luminosities and ratios of the key diagnostic emission line; these will be used to prove the possible existence (and to characterize the properties) of different spectroscopic subpopulations of radio-loud AGN;

-

to explore the presence of broad Balmer lines in the different subclasses of RGs, separated on the basis of e.g. radio morphology and diagnostic line ratios;

-

to perform a detailed test on the unified models for RL AGN, taking advantage of the uniform dataset, of the sample completeness, and of the accurate spectral characterization;

-

to compare the spectral properties of radio-loud and radio-quiet AGN;

-

to analyze the host galaxy properties from the point of view of its stellar component, looking for connections between the nuclear activity (and its onset) and the star formation history of the host;

-

to address the connection between the optical spectral characteristics and the multiwavelength properties of the sample, derived combining HST, Chandra, Spitzer and VLA data;

-

to study the environment of radio sources (using the target acquisition images);

-

to explore the properties of the off-nuclear gas (e.g. spatially resolved line ratios and velocity fields) and stellar component by using the full two dimensional dataset.

The paper is organized as follows: in Sect. 2 we present the sample, the observational procedure and the data reduction, leading to an atlas of calibrated spectra. In Sect. 3 we provide measurements of the emission line fluxes, obtained after separating the starlight from the emission of the active nucleus. In Sect. 4 we describe the quality of the resulting measurements and describe in more details the RGs showing broad emission lines. A summary is given in Sect. 5.

Throughout, we have used H₀ = 71 km s^-1 Mpc^-1, and .

2 Observations and data reduction

2.1 Sample selection

The Third Cambridge (3C) Catalogue of radio sources was compiled by Edge et al. (1959): they used the Cambridge double Interferometer to detect all the radio emitters at a frequency of 159 MHz, resulting in 471 sources catalogued.

Bennett (1962a,b) revised the catalog creating the 3CR version. He obtained new observational data to provide a complete and reliable list for all sources with a flux density greater than  > 9 Jy at a frequency of 178 MHz and for which the declination is  degrees (apart from some limitations for sources of low surface brightness). The revision excluded some sources from the catalog (as being below the flux limit or as resolved blends of adjacent sources) and included other sources (as extended sources with diameters up to 1). 328 sources are catalogued in the 3CR. Spinrad et al. (1985) identified the optical counterparts and measured the redshift of the 298 extragalactic sources in the 3CR. Our sample comprises all of the 113 3CR sources from Spinrad et al. (1985) with z < 0.3. We only excluded 3C 231 (M 82, a starburst galaxy) and 3C 71 (NGC 1068, a Seyfert galaxy).

Here we present the optical spectra of 104 3CR objects.

2.2 Observations

All but 18 optical spectra were taken with the Telescopio Nazionale Galileo (TNG), a 3.58 m optical/infrared telescope located on the Roque de los Muchachos in La Palma Canary Island (Spain). The observations were made using the DOLORES (Device Optimized for the LOw RESolution) spectrograph installed at the Nasmyth B focus of the telescope. The observations were organized into four runs of 4-5 nights per run starting from November 2006 and ending on January 2008. The chosen long-slit width is 2 : it is centered on the nucleus of the sources and it is aligned along the parallactic angle in order to minimize light losses due to atmospheric dispersion.

Table 1:   Detectors used for the survey.

For the first run (22-25 November 2006) and the second run (12 April 2007), the detector was a Loral thinned and back-illuminated CCD. For the third run (6-18 August 2007) and the fourth run (4-8 January 2008) a new thinned, back-illuminated, E2V 4240 CCD had been installed. It has a smaller pixel size and the field of view is decreased by 1 arcmin. More details on the CCD detectors are listed in Table 1.

Table 2:   Grisms used for the observations.

Table 3:   Observational strategy.

For each target we took:

-

one acquisition image (240 s) in R band;

-

one (or two) low resolution spectrum with the LR-B grism (3500-8000 Å) with a resolution of 20 Å;

-

two high resolution spectra with the VHR-R (6100-7800 Å) or VHR-I (7200-8900 Å) grisms with a resolution of 5 Å.

The spectral coverages of the spectra are reported in Table 2. The grism wavelength ranges change slightly between the two CCDs because of the small difference in the pixel size. The observational strategy is summarized in Table 3. In order to compensate for the galaxies' dimming with (1+z)⁴, the exposure times depend on the source redshifts. We divided the targets into three groups:

-

for z<0.1 (LR-B) = 500 s;

-

for 0.1<z<0.2 (LR-B) = 750 s;

-

for 0.2<z<0.3 (LR-B) = 1000 s.

The 1000 s exposures were divided into two subexposures of 500 s. The high resolution spectra have an exposure time twice the low resolution ones. For each target, the VHR-R or VHR-I grisms are chosen in such a way that the [O I] 6300, 64 Å lines are observable, i.e.:

-

VHR-R for z<0.175;

-

VHR-I for z>0.175.

The combination of the LR-B and VHR ranges of wavelengths enables us to cover the most relevant emission lines of the optical spectrum and in particular the key diagnostic lines H, [O III] 4959, 5007, [O I] 6300, 64, H, [N II] 6548, 84, [S II] 6716, 31. The high resolution spectra are a sort of zoom on the H region with the aim of resolving the H from the [N II] doublet, as well as the two lines of the [S II] doublet.

Table 4 provides the journal of observations for the four runs and the main information on the sources: (1) name of the source; (2) and (3) J2000 coordinates; (4) redshift; (5) date of observation; (6) CCD used in reference to Table 1; (7) number of low resolution spectra; (8) exposure time for each low resolution spectrum; (9) high resolution grism used; (10) number of high resolution spectra; (11) exposure time for each high resolution spectrum. In some cases the exposure times were increased because of bad sky conditions (i.e. seeing >2.5 ): these sources are identified with an ``a'' in the note column of Table 4.

Nine sources of our sample (namely 3C 020, 3C 063, 3C 132, 3C 288, 3C 346, 3C 349, 3C 403.1, 3C 410, 3C 458) could not be observed due to scheduling problems and time constraints. Since these sources were excluded for random reasons from the observing list we believe that they do not induce any bias in the sample. We verified a posteriori that they are apparently randomly distributed in the sky, as well as within the distributions of radio power and redshift of the whole sample.

2.3 Data reduction

The LONGSLIT package of NOAO's IRAF reduction software was used in order to:

-

subtract the bias from each science image (we used the over-scan region of each image);

-

divide the science images for the flat field (created from the halogen lamp emission);

-

subtract the background; most observations were split into two subexposures (with only the exception of the LR-B spectra of low redshift galaxies, see Table 4) obtained moving the target by 25 or 50 along the slit. Subtraction of the two subexposures removes most of the sky background. The residual background (due to changes in sky brightness or for the single LR-B spectra) was subtracted measuring the average on each pixel along the dispersion direction in spatial regions immediately surrounding the source spectrum;

-

wavelength calibrate the images using the calibration lamps (Ar, Ne and He);

-

correct the optical distortions by fitting polynomial functions along both the horizontal and vertical directions of the CCD to the lamp spectra;

-

flux calibrate the spectra using spectrophotometric standard stars. We observed two or three standard stars per night using the same telescope configuration used for the targets.

We then extracted and summed a region of 2 along the spatial direction, resulting in a region covered by our spectra of 2   2 . For each galaxy, multiple exposures taken with the same setting were averaged into a single spectrum.

The accuracy of the relative flux calibration was estimated from the residuals of the calibration of the standard stars; comparison of the calibrated spectra obtained using the different standard stars observed during the same night implies that our flux calibration agrees to within a <5% level.

More than 75% of the nights were photometric and we therefore expect a rather stable absolute flux calibration. This result was checked for 20 objects of the sample randomly selected using their HST R-band images (de Koff et al. 1996; Martel et al. 1999). The procedure used for the comparison is the following: we smoothed the HST image to the seeing measured in the acquisition TNG image, extracted a region corresponding to 2   2 , and measured its flux based on the HST absolute flux calibration. We convolved the LR-B TNG spectra with the filter transmission used to obtain the HST images (the broad band F702W filter of the Wide Field Planetary Camera 2) and measured the flux within this spectral band. We compared these two measurements and found an agreement within 20%. The absolute flux calibration of high resolution spectra was obtained scaling them to the low resolution calibrated spectra.

Table 4:   Log of the observations.

The telluric absorption bands were usually left uncorrected except in the few cases in which an emission line of interest fell into these bands. In these cases we corrected the atmospheric absorptions using the associated standard stars as templates.

Spectra for 18 3CR objects are available from the Sloan Digital Sky Survey (SDSS) database (York et al. 2000; Stoughton et al. 2002; Yip et al. 2004), Data Release 4-6. Since they have comparable signal to noise, resolution and wavelength coverage (3800-9200 Å) of the TNG spectra, we decided not to observe them at the TNG and to use the SDSS spectra for our analysis.

Figure 1 shows, for all the observed targets, the low resolution spectrum (upper image) and the high resolution one (bottom image). The calibrated spectra are in units of 10^-18 erg cm^-2 s^-1 Å^-1. The wavelengths (in Å units) are in the observer frame in the axes below the images while they are in the source frame in the axes above them. The three main telluric absorption bands are indicated as circles with a cross inside.

3 Data analysis

3.1 Starlight subtraction

The extraction apertures contain, in addition to the emission produced by the active nucleus, a substantial contribution from the host galaxy starlight. In order to proceed in our analysis it is necessary to separate these two components by modeling the stellar emission.

In order to prepare the spectra for the stellar removal, we corrected for reddening due to the Galaxy (Burstein & Heiles 1984,1982) using the extinction law of Cardelli et al. (1989). The galactic extinction E(B-V) used for each object was taken from the NASA Extragalactic Database (NED) database and is listed in Table 6. We also deredshifted the spectra using the value of redshift from NED.

The spectra modeling was done considering two wavelength ranges of 1000 Å centered on the H line from the HR spectrum and on the H line from the LR-B spectrum. For each of the two ranges we subtracted the best fit single stellar population (SSP) model taken from the Bruzual & Charlot (2003) library. We selected a grid of 33 single stellar population models varying over 11 values of stellar ages (0.1, 0.3, 0.5, 1, 2, 3, 5, 7, 9, 11, 13 Gyr) and three values of absolute metalicity (0.008, 0.02, 0.05), see Table 2. The templates assume a Salpeter Initial Mass Function (IMF) formed in an instantaneous burst, with stars in the mass range . The parameters free to vary independently of each other in order to obtain the best fit are the stellar age, the metalicity, the normalization of the model, the velocity dispersion (line-broadening function), and a small adjustment to the published redshift. We allowed for different models for the population in the H and H regions. In fact, as discussed in Sect. 3.3, the stellar content of radio galaxies is often composed by populations of various ages and thence the dominant population can vary in different bands. We excluded from the fit the spectral regions corresponding to emission lines, as well as other regions affected by telluric absorption, cosmic rays or other impurities. The SSP subtracted from the spectra are listed in Table 5: for each source the best fit SSP and the are reported for the two wavelength regions around H and H.

In Fig. 2 we show a few examples of the process of removal of the galaxy starlight. The six spectra presented are representative of the different quality in the sample (in order of decreasing signal-to-noise ratio from top to bottom). We show separately galaxies with different levels of line equivalent widths (high EW in the right two columns, low EW in the left 2 columns).

In a few galaxies the continuum is essentially featureless (see for example, 3C 084 in Fig. 1) and it is likely to be dominated by non-stellar emission. No starlight subtraction was performed for these objects. In several spectra a first attempt to model the starlight left substantial positive residuals in the region surrounding the H and/or H lines, indicative of the presence of a broad line component. In these cases we flagged a spectral region extending over between 100 and 200 Å around the broad line, extended the spectral range considered, and repeated the fitting procedure (see Fig. 3 for an example).

Table 5:   Summary of the single stellar population fitting.

For a few objects, we used a different technique to remove the galactic emission because we did not find a satisfactory fit using the template models. We then used as template the spectra extracted from two off-nuclear regions, flanking the nuclear aperture, summed over a spatial aperture of 2 and appropriately scaled to match the nuclear spectrum. An example of this method is given in Fig. 4. The objects in which one of these different approaches was adopted are marked appropriately in Table 4.

The case of 3C 386 deserves a special note. Lynds (1971) reports that the peculiar observational characteristics of 3C 386 result from the chance superposition of a F7 star on the galaxy's nucleus. This is confirmed by our spectra that show strong Balmer absorption lines at zero redshift accompanied by absorption and emission features at the redshift of the radio galaxy, z=0.0177 (see Fig. 5). Before modeling the stellar population of this galaxy, we removed the contribution of the foreground star. We obtain the best match of the absorption lines of the star using a F5V stellar template (from Pickles 1998), with the same reddening of the galaxy, E(B-V)= 0.335. The low resolution spectrum resulting from subtracting the foreground star is shown in Fig. 5 where we also present the H spectral region from the high resolution spectrum, with a well visible H+[N II] triplet and [S II] doublet. No broad H line is visible in this spectrum in contrast to the result of Simpson et al. (1996).

3.2 Measurement of the emission lines fluxes

The next step of our analysis consists of the measurement of the emission line intensities for which we used the specfit package in IRAF. We measured line intensities fitting Gaussian profiles to H, [O III] 4959, 5007, [O I] 6300, 64, H, [N II] 6548, 84, and [S II] 6716, 31. Some constraints were adopted to reduce the number of free parameters: we required the FWHM and the velocity to be the same for all the lines. The integrated fluxes of each line were free to vary except for those with known ratios from atomic physics: i.e. the [O I] 6300, 64, [O III] 4959, 5007 and [N II] 6548, 84 doublets. Where required, we insert a linear continuum.

Table 6 summarizes the intensities of the main emission lines (dereddened for Galactic absorption) relative to the intensity of the narrow component of H, for which we give flux and luminosity. To each line we associated its relative error, as a percentage. We placed upper limits at a 3 level to the undetected, but diagnostically important, emission lines by measuring the noise level in the regions surrounding the expected positions of the lines, and adopting as line width the instrumental resolution. When no emission lines are visible, we only report the upper limit for H. The missing values in Table 6 correspond to lines outside the coverage of the spectra or severely affected by telluric bands.

For the galaxies with a broad-line component we first attempted to reproduce the BLR emission with a gaussian profile, allowing for velocity shifts with respect to the narrow lines and also for line asymmetry. Most line profiles were well reproduced, but in some cases (e.g. 3C 332) two gaussians had to be included. Nonetheless, in 3C 111 and in 3C 445 the broad line profile is so complex, and the prominence of the broad component with respect to narrow lines is so large, it precludes any attempt to measure the [N II] doublet and the narrow component of H (or H in the blue region of the spectrum). For these objects we use the [O III] line as reference instead of the narrow H component. No narrow lines are visible in 3C 273 and we only report its broad H flux. For all 18 galaxies with a broad H component we also give its flux in Table 6. These objects will be discussed in more detail in Sect. 4.3.

3.3 On the effects of template mis-match on the emission lines fluxes

An important issue related to the subtraction of the stellar component is the effect of the template mis-match for the measurements of the emission lines. This is particularly important for the estimate of H in the galaxies showing emission lines of relatively low equivalent width. In fact they can be strongly affected by the level of the absorption features associated to the stellar population.

We consider as an example the case of 3C 310. Its spectrum is of average quality in our dataset and it shows only weak emission lines, while the H line is, at most, marginally detected in its spectrum. The best fitting stellar population has an age of 5 Gyr and a metalicity of Z=0.05 and it is shown in red in the middle panel of Fig. 6 superposed on the original spectrum. In the residual spectrum a well defined H line emerges in emission, caused by the removal of a substantial absorption associated with the stellar emission. This indicates that its intensity is strongly influenced by the choice of the stellar template. To associate a proper error to this procedure it is necessary to establish the range of acceptable stellar models (in terms of age and metalicity) and the resulting uncertainty in the line measurement.

As a formal error propagation across all steps of the data reduction is clearly unfeasible, we estimated the typical error of our spectra by measuring the rms flux in continuum dominated spectral regions. Even with this approach, the value of minimum reduced  obtained from the best fitting stellar population is often larger than the value indicative of a good fit ( ) and this is in contrast with the fact that the stellar models seem to trace in general the spectra rather well. There are several reasons for this discrepancy: i) the is not properly normalized (e.g. because not all data points are independent); ii) our estimate of the signal-to-noise ratio does not include the uncertainties in e.g. the wavelength and flux calibration; iii) there are real mismatches between data and models, in part caused by the use of a discrete grid of stellar models. We then decided, when necessary, to rescale our error bars such that the overall best fitting model provides /d.o.f. = 1, following the approach proposed by Barth et al. (2001) in a different context. This is a conservative approach since it has the effect of increasing the range of the acceptable templates with respect to what would have been obtained only using the measured rms of each spectrum.

We consider as acceptable templates all stellar populations for which the resulting is increased by less than  = 9.21, corresponding to a confidence level of 99% (2.6 level) for 2 parameters. In the case of 3C 310, the error derived setting is 9.5  10^-18 erg cm^-2 s^-1 Å^-1. This value is similar to the rms of the stellar subtracted spectrum, which is 8.2  10^-18 erg cm^-2 s^-1 Å^-1. Adopting the procedure outlined above, the best fit is the only statistically acceptable model. Nonetheless, in Fig. 6 we show the two models with the lower values of  , differing by one step in metalicity or age. In both cases significant residuals already emerge in the continuum subtracted spectrum, confirming visually the poorer quality of the fit. We re-estimate the H flux resulting from these two templates. They differ by 4% with respect to our original measurement using the best fit model. This is lower than the statistical error associated by specfit to this measurement, which is 7%. We conclude that, in the case of 3C 310, a possible template mis-match does not have a significant effect on the H flux measurement.

Table 6:   Emission line measurements.

The influence of the stellar templates on the other emission lines (and in particular on the H line that is 3 brighter that H while they have the same EW in absorption) is negligible because the stellar population has not significant absorption features at the line wavelengths.

For the majority of the sources, only models differing by at most one step in age or in metalicity are consistent with the data, without a significant impact on the H flux. Only for the lower quality spectra can this range be broader. For low SNR spectra, coupled with emission lines of low equivalent widths, the resulting measurement error is dominated by the choice of the stellar template. In these cases we adopted the following strategy: we subtracted the acceptable template with the lowest age (i.e. with the strongest H absorption) and considered the H measurement as an upper limit. One of such galaxy is 3C 173.1, already shown in the bottom right panel of Fig. 2.

Finally we note that our approach to the starlight subtraction is limited to the choice of the best fitting single population model, similarly to the method adopted by Kauffmann et al. (2003) for the SDSS sample of emission line galaxies and by Tadhunter et al. (1993) for the 2 Jy sample of radio galaxies. Nonetheless, this procedure does not grasp the complexity of the emission processes in radio galaxies. As clearly shown by several studies in the literature the stellar content of radio galaxies is often composed by populations of different ages, and there are also evidence for internal absorption, particularly associated to the young stars component (e.g. Wills et al. 2004; Raimann et al. 2005; Tadhunter et al. 2005; Holt et al. 2007). The presence of stars of various ages is indirectly confirmed by the fact that the best fit stellar population found in the H regions is on average  one step in age younger than for the H band. This can be explained by the higher relative contribution of young stars at shorter wavelengths that drives the age of the SSP to lower values with respect to the redder part of of the spectrum. Furthermore, other mechanisms contribute to the continuum emission, e.g. nebular continuum and light from the accretion disk (direct or via scattering). However, a full separation of these contributions requires a detailed analysis on a object-by-object basis in order to properly measure the rather large set of free parameters describing the various components. Moreover this fitting requires spectra with a high signal to noise, and this is not always achieved in our data. We defer such a study to a forthcoming paper.

There is however a general consideration that can be drawn already at this stage. Nuclear emission and nebular continuum are expected to be in general bluer that the starlight. This biases the derived age of the single stellar population that appears younger that it actually is. It also dilutes the stellar absorption features, thus biasing the metalicity to lower values. The main resulting effect is an overestimate of the correction of the H line and, consequently, an over-rating of this line, particularly in the objects with brightest optical nuclei.

4 Results

4.1 Emission line data quality

We summarize here the data quality of our spectra from the point of view of the measurements of the emission line intensities. Only in three radio galaxies (namely 3C 052, 3C 129.1 and 3C 130) did we fail to detect any emission line; in all cases an accurate redshift can be measured by fitting the stellar population models based on stellar absorption features.

Leaving aside these objects, the H line can be measured with a statistical accuracy of better than 10% with only three exceptions. In the red part of the spectrum, the [O I]6300 line is detected in 95% of the 3C galaxies. The completeness of the [S II] doublet measurements is slightly lower (87%) due to our choice to favor the inclusion of the [O I] in the high resolution spectra but, when covered by the data, the [S II] line is also measurable in 95% of the objects. In the blue part of the spectrum there are 15% of the sources in which H and/or [O III] are not detected.

All together, these measurements will enable us to locate the vast majority of the sources in the diagnostic planes that compare pairs of emission line ratios and also to perform a detailed analysis of the relationship between the multiwavelength characteristics of the 3C sources with their emission line properties.

In addition, a broad H line is seen in 18 galaxies and these are discussed in more detail in the Sect. 4.3.

4.2 Comparison with previous results

There is a extremely vast amount of spectroscopic data accumulated for the 3CR sample in the literature over the last 40 years and it is therefore, on the one hand, very interesting to contrast our measurements with those available, but, on the other hand, difficult to perform a full comparison on an object-by-object basis. We therefore decided to focus on three papers, those having the largest overlap with our sample.

Willott et al. (1999) collected [O III] measurements from optical spectra and images of 3CR objects from the literature (these data are available at the web-page: http://www.science.uottawa.ca/~cwillott/3crr/3crr.html). There are 28 sources in common with our sample. Figure 7 (left panel) shows the comparison between their data collection and our measurements, performed after scaling their reported luminosities to our cosmological parameters. These data show large discrepancies, even in excess of a factor of 100, particularly for the sources with the lowest [O III] fluxes. Sources with stronger fluxes are in better agreement with our results, but there are still large differences.

In the central and right panels we compare our [O III] fluxes with the values found by Simpson et al. (1996) and Tadhunter et al. (1993), derived from long-slit spectra of two complete samples, with respectively 8 and 16 sources in common with our 3CR subsample. Their data have been corrected for Galactic absorption before comparison. Quite reassuringly, essentially all sources have [O III] fluxes differing by less than a factor of 2.

4.3 Radio galaxies with broad emission lines

In 18 sources, all of them FR II galaxies, we detected a broad H component. Broad lines can be very strong, as in 3C 273, or, as in 3C 184.1, they become clearly visible only after the subtraction of the stellar light. Figure 8 shows all spectra showing the H broad components: the dashed line, representing the continuum emission, corresponds to the zero flux level when the SSP has been subtracted, otherwise it represents the AGN power law emission.

In some cases they are symmetric and centered on the narrow component (3C 287.1), but more often they are highly red or blue shifted (3C 184.1, 3C 227). In most objects the H line presents a highly irregular profile, with ``bumps'' in the line wings. For example, in 3C 197.1 and 3C 445 the broad component has a step-shaped emission in the red wing, while in 3C 227 the step-shaped emission is on the blue side of the line. We also confirm the ``double-humped'' shape of the BLR of 3C 332 (Eracleous & Halpern 1994).

The technique used to measure the narrow line fluxes (described in Sect. 3.2) provides us also with flux estimates for the broad components. In the cases where no satisfactory fit was obtained with multiple gaussians, we measured the flux by directly integrating the area defined by the broad component. The H broad line fluxes are reported in the last column of Table 6.

We will explore in more detail the issue of the presence and properties of broad lines in a forthcoming paper.

5 Summary

We presented a homogeneous and 92% complete dataset of optical nuclear spectra of the 113 3CR radio sources with redshift <0.3. The data were obtained mostly from observations performed at the Telescopio Nazionale Galileo and are complemented with SDSS spectra, available for 18 galaxies. For each source we obtained a low (20 Å) resolution spectrum covering the 3500-8000 Å spectral range and a higher (5 Å) resolution spectrum centered on the H line and 1700 Å wide. From the two dimensional calibrated data, we extracted spectra over a 2   2  nuclear region. A complete atlas of the optical nuclear spectra is presented.

We separated the contribution of the emission produced by the active nucleus from the galaxy starlight. This has been achieved by subtracting the best fit model of stellar emission from a grid of single stellar population templates. We then measured and tabulated the fluxes of the main emission lines by fitting gaussians to the residual spectra. We discussed the effects of the possible mis-match between the galaxies stellar population and the adopted templates; the conclusion of this analysis is that they do not affect significantly the line measurements, with only very few exceptions where the estimate of the H flux is compromised.

Only in three radio galaxies did we fail to detect any emission line. In most cases (90% of the sample) we were able to measure all the brightest lines, e.g. H, [O III] 4959, 5007, [O I] 6300, 64, H, [N II] 6548, 84, and [S II] 6716, 31.

In addition, a broad H line is seen in 18 galaxies. They show a wide variety in terms of line profiles, from well behaved symmetric lines centered onto the narrow H, to ``double humped'' lines, but also include highly irregular shapes, with secondary peaks and steps.

The data presented here will be used in follow up papers to address the connection between the optical spectral characteristics and the multiwavelength properties of the sample.

Acknowledgements

We thank the referee, Katherine Inskip, for her useful comments and suggestions. S.B. and A.Ce. acknowledge the Italian MIUR for financial support. A.Ca. acknowledges COFIN-INAF-2006 grant financial support. 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. This research has made use of NASA's Astrophysics Data System (ADS). This research has made use of the SDSS Archive, funding for the creation and distribution of which was provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Aeronautics and Space Administration, the National Science Foundation, the US Department of Energy, the Japanese Monbukagakusho, the Max Planck Society, and The Higher Education Funding Council for England. The official SDSS Web site is www.sdss.org. The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, University of Cambridge, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Institute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, University of Portsmouth, Princeton University, the United States Naval Observatory, and the University of Washington.

References

• Barth, A. J., Sarzi, M., Rix, H., et al. 2001, ApJ, 555, 685 [NASA ADS] [CrossRef] (In the text)
• Bennett, A. S. 1962a, MNRAS, 125, 75 [NASA ADS]
• Bennett, A. S. 1962b, MmRAS, 68, 163 [NASA ADS]
• Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 [NASA ADS] [CrossRef] (In the text)
• Burstein, D., & Heiles, C. 1982, AJ, 87, 1165 [NASA ADS] [CrossRef]
• Burstein, D., & Heiles, C. 1984, ApJS, 54, 33 [NASA ADS] [CrossRef]
• Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245 [NASA ADS] [CrossRef] (In the text)
• de Koff, S., Baum, S. A., Sparks, W. B., et al. 1996, ApJS, 107, 621 [NASA ADS] [CrossRef]
• Edge, D. O., Shakeshaft, J. R., McAdam, W. B., Baldwin, J. E., & Archer, S. 1959, MmRAS, 68, 37 [NASA ADS] (In the text)
• Eracleous, M., & Halpern, J. P. 1994, ApJS, 90, 1 [NASA ADS] [CrossRef] (In the text)
• Fanaroff, B. L., & Riley, J. M. 1974, MNRAS, 167, 31P [NASA ADS] (In the text)
• Holt, J., Tadhunter, C. N., González Delgado, R. M., et al. 2007, MNRAS, 381, 611 [NASA ADS] [CrossRef] (In the text)
• Kauffmann, G., Heckman, T. M., White, S. D. M., et al. 2003, MNRAS, 341, 33 [NASA ADS] [CrossRef] (In the text)
• Lynds, R. 1971, ApJ, 168, L87 [NASA ADS] [CrossRef] (In the text)
• Martel, A. R., Baum, S. A., Sparks, W. B., et al. 1999, ApJS, 122, 81 [CrossRef]
• Pickles, A. J. 1998, PASP, 110, 863 [CrossRef] (In the text)
• Raimann, D., Storchi-Bergmann, T., Quintana, H., Hunstead, R., & Wisotzki, L. 2005, MNRAS, 364, 1239 [NASA ADS] [CrossRef] (In the text)
• Simpson, C., Ward, M., Clements, D. L., & Rawlings, S. 1996, MNRAS, 281, 509 [NASA ADS] (In the text)
• Spinrad, H., Marr, J., Aguilar, L., & Djorgovski, S. 1985, PASP, 97, 932 [NASA ADS] [CrossRef] (In the text)
• Stoughton, C., Lupton, R. H., Bernardi, M., et al. 2002, AJ, 123, 485 [NASA ADS] [CrossRef]
• Tadhunter, C. N., Morganti, R., di Serego-Alighieri, S., Fosbury, R. A. E., & Danziger, I. J. 1993, MNRAS, 263, 999 [NASA ADS] (In the text)
• Tadhunter, C., Robinson, T. G., González Delgado, R. M., Wills, K., & Morganti, R. 2005, MNRAS, 356, 480 [NASA ADS] [CrossRef] (In the text)
• Willott, C. J., Rawlings, S., Blundell, K. M., & Lacy, M. 1999, MNRAS, 309, 1017 [NASA ADS] [CrossRef] (In the text)
• Wills, K. A., Morganti, R., Tadhunter, C. N., Robinson, T. G., & Villar-Martin, M. 2004, MNRAS, 347, 771 [NASA ADS] [CrossRef] (In the text)
• Yip, C. W., Connolly, A. J., Vanden Berk, D. E., et al. 2004, AJ, 128, 2603 [NASA ADS] [CrossRef]
• York, D. G., Adelman, J., Anderson, Jr., J. E., et al. 2000, AJ, 120, 1579 [NASA ADS] [CrossRef]

Footnotes

... data

Based on observations made with the Italian Telescopio Nazionale Galileo operated on the island of La Palma by the Centro Galileo Galilei of INAF (Istituto Nazionale di Astrofisica) at the Spanish Observatorio del Roque del los Muchachos of the Instituto de Astrofisica de Canarias.

... IRAF

IRAF is the Image Reduction and Analysis Facility of the National Optical Astronomy Observatories, which are operated by AURA, Inc., under contract with the US National Science Foundation. It is also available at http://iraf.noao.edu/

... redshift

The redshifts resulting from the spectral fitting always agree within the uncertainties with the published values with the only exception of 3C 130, for which we estimate z=0.032 instead of z=0.109, based on the position of the Na D absorption feature.

All Tables

Table 1:   Detectors used for the survey.

Table 2:   Grisms used for the observations.

Table 3:   Observational strategy.

Table 4:   Log of the observations.

Table 5:   Summary of the single stellar population fitting.

Table 6:   Emission line measurements.

All Figures

---

Copyright ESO 2009