Astron. Astrophys. 363, 887-900 (2000)

2. Candidate high frequency peakers

CSS and GPS sources have convex radio spectra peaking at frequencies ranging from about hundred MHz to a few GHz; we assume/define that HFPs have similar spectral properties, with the peak occurring at a few GHz or higher frequencies.

The availability of large areas covered by radio surveys, and the need of a spectral peak at high frequencies made the choice of the NVSS (Condon et al. 1998) and of the 87GB (Gregory et al. 1996) catalogues quite natural.

We cross correlated the 87GB catalogue at 4.9 GHz with the NVSS catalogue at 1.4 GHz and selected the sources with inverted spectra, and in particular those with a slope steeper than -0.5 (). We defined two samples of candidates: the "bright " sample, with sources brighter than 300 mJy at 5 GHz and covering nearly the whole area of the 87GB (declination between 0 and +75^o), excluding objects with to avoid the galactic plane and ease the optical identification work; the "faint " sample is restricted to the area covered by the FIRST survey (Becker et al. 1995) as well, and limited to sources brighter than 50 mJy at 4.9 GHz. This work presents the "bright" sample, while a forthcoming paper (Stanghellini et al. 2001) will describe the "faint" sample.

The search for candidates for the bright sample started with the 1795 sources from the 87GB stronger than 300 mJy and in the region of the sky described above. We used a simple fortran program to make a cross correlation of the positions of these sources with the catalogue derived from the NVSS. The error in the position for the sources in the 87GB is much larger than that associated with the NVSS, and is generally of the order of 10-15 arcseconds in both RA and DEC. We considered positionally coincident the sources with difference in either coordinate ( or ) smaller than the largest between 45 arcsec and 3 times the error reported in the 87GB catalogue. Only 55 sources could not be identified (3.1%) since they fall in areas not yet covered by our release of the NVSS catalogue (as of July, 1999). Among the remaining 1740 objects, 164 (9.4%) had an inverted spectra with slope steeper than -0.5 and they were included in our starting list. We then inspected the NVSS images to make sure that the component in the catalogue accounted for the whole flux density. The extended objects (typically FRII and a few FRI or complex radio sources), resolved by the NVSS but a single source in the 87GB, were removed. Our list of "bright" HFP candidates consisted of 103 sources. After a search for optical identification one source have been dropped, since it is associated with a planetary nebula (J1812+0651), hence to a completely different physical phenomenon. Therefore, the final list is made up with 102 (5.9% of the starting dataset) candidates and is presented in Table 1: Column 1 gives the J2000 name; Columns 2 and 3 provide the J2000 coordinates from the NVSS catalogue (very accurate positions can be found in the Jodrell Bank VLA Astrometric Survey, JVAS catalogue; Patnaik et al. 1992; Browne et al. 1998; Wilkinson et al. 1998); Columns 4, 5 and 6 report the flux densities in the NVSS, 87GB and JVAS catalogues, respectively; Column 7 shows the spectral index between the NVSS and the 87GB; Column 8 reports whether the source belongs to other relevant samples (see below); Columns 9, 10 and 11 give the optical ID, magnitude and redshift; finally Column 12 provides the B1950 source name.

Table 1. Candidates observed with the VLA. All columns are self-explicative, except Column 8, where a reference to other samples is reported; a full description is given in the text.

Table 1. (continued)

We found 14 sources in common with the new GPS candidate starting list in Marecki et al. (1999) (`m' in Column 8), three objects are in the `bright' GPS sample from Stanghellini et al. (1998) (`st'), two sources in the `faint' GPS sample from Snellen et al. (1998) (`sn'). It has been useful to compare our list to the 550 compact extragalactic objects in Kovalev et al. (1999) (`K'), where nearly simultaneous radio spectra are available between 1 and 22 GHz. We searched our HFP candidates in the Caltech-Jodrell Bank flat-spectrum sample (Taylor et al. 1996and references therein) (`pr', `cj1' and `cj2') and with the Kellermann et al. (1998) (`k') VLBA survey at 15 GHz, in order to have images of the pc-scale radio morphology. Finally most of our HFP candidates have very short snapshot images in the VLBA Calibrator Survey (VCS, Peck & Beasley 1998).

The optical ID (capitals) and redshift are from the NED database, when available. We also report our optical ID on the digitised red plates of the Palomar Observatory Sky Survey (POSS) (small `g' and `s' for extended or stellar) when no other optical information is available; a `?' following the optical identification means that the classification is uncertain. We remark that the optical magnitudes, mostly from the NED database, reported in Table 1 are not homogeneous (i.e. in the same band). Also variability plays an important role, given that a significant fraction of the sources are associated with blazars. In fact among the candidates there are also 3 BL Lac objects from the 1 Jy sample (Stickel et al. 1991) and a few other sources known to be variable.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk@link.springer.de