The FIRST-NVSS-WENSS-GB6-SDSS Radio Object Catalog

This work was supported by NSF grant AST-0507259


Summary

We present here a combined catalog of radio objects drawn from four radio catalogs (FIRST, NVSS, WENSS, and GB6) and the optical SDSS. The two primary surveys used in creating the matched catalog are the two 20 cm surveys, FIRST and NVSS: a row in the matched catalog is either a FIRST object with an NVSS match, an NVSS object with a FIRST match, or an isolated object from either survey. All FIRST/NVSS catalog sources were than matched to the other two radio surveys, GB6 at 6cm and WENSS at 92cm.

The complete catalog contains 2,724,343 entries in the region of the sky north of -40° declination covered by the NVSS (which completely contains the part of sky observed by FIRST). A 2955 deg2 "overlap region" is defined by the area observed by all four radio surveys and photometrically with the SDSS. The overlap region contains 178,000 FIRST-NVSS matches (within 30"), 110,000 isolated FIRST sources, and 23,000 isolated NVSS sources. The catalog includes data parameters (from the original source surveys) for the closest positional matches. In addition, the catalog has "matching parameters" which give the distance to, and total number of, positional matches found within the chosen search radii.

The catalog is available for download at the bottom of the page. A detailed description of the calibration and a preliminary science analysis is presented in Kimball & Ivezić (2008) (hereafter KI08). Individual files with high resolution figures are available here. Please see below for updates.

The radio source surveys

The four radio surveys used to create this catalog are FIRST (Faint Images of the Radio Sky at Twenty Centimeters), NVSS (NRAO-VLA Sky Survey, 20cm), GB6 (Green Bank Telescope, 6cm) and WENSS (Westerbork Northern Sky Survey, 92cm).

FIRST and NVSS both observed the sky at 20cm (1.4 GHz), but with very different spatial resolution. FIRST has a 5.4" beam with astrometrical errors of 0.5-1", while NVSS has a 45" beam with 1-7" astrometric accuracy. FIRST therefore distinguishes small-scale structures with accurate positions, but underestimates flux for extended sources. NVSS has less accurate positional measurements and cannot distinguish small-scale structures, but measures more accurate fluxes for extended sources and can find low-surface brightness objects missed by FIRST. Combining observations from the two surveys where their coverage coincides provides more information about a single source than could be gained by observations done with a single 20cm radio survey. FIRST, designed to coincide with the region of sky covered by the Sloan Digital Sky Survey, covers 9,000 deg2 in the north Galactic cap and a smaller strip along the celestial equator. With a detection limit of 1 mJy, the FIRST survey contains over 800,000 sources. NVSS is complete to 2.5 mJy, covers the entire sky north of -40°, and contains more than 1.8 million sources.

The Green Bank survey at 6cm (4.85 GHz) covers the sky in the declination region 0°<δ<75°. It has a beam size of 3.6'×3.4'. This survey is complete to 25 mJy and contains 75,000 sources.

The WENSS survey observed at the longest wavelength of the included surveys, 92cm (325 MHz). WENSS covers the entire sky north of δ=29° down to a completeness limit of approximately 18 mJy. The beam size is 54"×54"cosec(δ), and the survey contains over 200,000 sources.

The matched catalog includes most parameters which are available in each of the individual surveys, with some additions and modifications. For example, the unified radio catalog includes an integrated 6cm source flux, which can be calculated from GB6 survey parameters. We have also modified the original survey parameters, where necessary, such that axis sizes and fluxes are in the same units (arcsec and mJy). All other parameters remain the same as in the original surveys.

Matching procedure

The first step in the catalog production was to match FIRST and NVSS. We undertook this step in two parts by first positionally matching FIRST objects to NVSS objects (using NVSS positions as search centers), then matching NVSS objects to FIRST objects (using FIRST positions as search center). For each FIRST and NVSS source, we store in the catalog the closest three neighbors from the other catalog, within 30". Most FIRST-NVSS matches show up twice in the catalog, once with the FIRST detection as the primary and once with the NVSS detection as the primary. Catalog users therefore have the option to examine matching results separately for either catalog. It is easy to distinguish between the two searches (NVSS to FIRST or FIRST to NVSS) using the catalog's matchflag parameters.

In addition to storing data for the three closest matches within 30", we also provide the total number of neighbors found with 5", 10", 30", and 120" (matchtot parameters).

We then positionally matched to neighboring sources in GB6 and WENSS using a 120" search radius. When available, we used the more accurate position of the FIRST source in each catalog row for this step. For NVSS objects with no nearby FIRST neighbor, the NVSS position was used for this step. (See note below on positions.) Each catalog entry contains data describing the closest WENSS and GB6 neighbor; the total number of GB6 and WENSS matches within 120" is stored in the respective matchtot parameter. GB6 and WENSS each cover a different portion of the sky than the 20cm surveys; for primary sources outside GB6 or WENSS survey coverage, the respective matchtot parameter is set to -1.

We also matched the radio catalog with the SDSS photometric survey: we retain photometric metadata, and also spectroscopic metadata for objects observed by the SDSS spectroscopically. Due to the very different astrometric accuracy of NVSS and FIRST, we chose two different search radii for positional matching to FIRST (where available) and NVSS (when no FIRST match available). We used a search radius of 60".

Due to the high density of photometric SDSS sources (about 7.5 arcmin-2), it is easy to find serendipitous line-of-sight matches between intrinsically unrelated objects. We therefore provide data separately for both the nearest and the brightest SDSS photometric matches. If the nearest source is within 3" of the primary object (see Matchflags below), we also find the brightest object within 3". If the nearest object is further than 3" but within 10", we find the brightest object within 10". If no photometric neighbors were found within 10", we used a 30" search radius. If that failed, we used a 60" search radius. In 98.8% of cases where an SDSS match was found within 3", the nearest object and the brightest object are the same source.

The Catalog Format

Data Field Units Format Description
-- Row Identifiers --
1 uniq_id long integer Unique catalog row identifier
2 ra degree double Designated right ascension of entry (see note on positions)
3 dec degree double Designated declination of entry
-- FIRST data --
4 first_id long integer unique FIRST identifier
5 first_ra degree double right ascension (J2000) values from FIRST survey; for details see the FIRST website
6 first_dec degree double declination (J2000)
7 first_wflag char warning: possible sidelobe of nearby bright source
8 first_peak_flux mJy/beam float peak flux at 20cm
9 first_flux mJy float integrated flux density at 20cm
10 first_sky mJy/beam float local noise estimate at source position
11 first_major arcsec float FWHM of major axis (after PSF deconvolution)
12 first_minor arcsec float FWHM of minor axis (after PSF deconvolution)
13 first_posangle degree float position angle east of north (after PSF deconvolution)
14 first_fmajor arsec float fitted major axis (before PSF deconvolution)
15 first_fminor arsec float fitted minor axis (before PSF deconvolution)
16 first_fposangle degree float fitted position angle east of north (before PSF deconvolution)
17 first_field char name of source field
-- NVSS data --
18 nvss_id long integer unique NVSS identifier
19 nvss_ra degree double right ascension (J2000) values from NVSS survey; for details see the NVSS website. Also, please see note on NVSS corrections.
20 nvss_dec degree double declination (J2000)
21 nvss_peak_int mJy/beam float peak intensity at 20cm (Stokes I)
22 nvss_flux mJy float integrated flux at 20cm
23 nvss_major arcsec float FWHM major axis after deconvolution (please see note on NVSS corrections)
24 nvss_minor arcsec float FWHM minor axis after deconvolution (please see note on NVSS corrections)
25 nvss_posangle degree float position angle east of north
26 nvss_q mJy/beam float interpolated Stokes Q value at position of I peak
27 nvss_u mJy/beam float interpolated Stokes U value at position of I peak
28 nvss_p_flux mJy float integrated linear polarized flux
29 nvss_i_rms mJy/beam float RMS noise in Stokes I image
30 nvss_pol_rms mJy/beam float RMS noise in Stokes Q,U images
31 nvss_res_rms mJy/beam float RMS Stokes I residual
32 nvss_res_peak mJy/beam float peak Stokes I residual
33 nvss_res_flux mJy/beam float integrated Stokes I residual
34 nvss_xpixel pixel float "X" pixel of source center
35 nvss_ypixel pixel float "Y" pixel of source center
36 nvss_field char name of (4x4 deg) field
37 nvss_jd_processed day long integer Julian date of processing
-- WENSS data --
38 wenss_id long integer unique WENSS identifier
39 wenss_ra degree double right ascension values from the WENSS survey; for details see the WENSS paper
40 wenss_dec degree double declination
41 wenss_type char label: Single, Component, Multiple, or Extended
42 wenss_flag char set to "*" when source-finding algorithm failed
43 wenss_peak_flux mJy/beam float peak flux density at 92cm
44 wenss_flux mJy float integrated source flux at 92cm
45 wenss_major arcsec float source major axis size
46 wenss_minor arcsec float source minor axis size
47 wenss_posangle degree float source angle on sky east of north
48 wenss_sky mJy/beam float local rms noise level
49 wenss_frame char name of WENSS frame
-- GB6 Data --
50 gb6_id long integer unique GB6 identifier
51 gb6_ra degree double right ascension (J2000) values from GB6 survey; for details see the GB6 catalog paper
52 gb6_dec degree double declination (J2000)
53 gb6_ra_err degree float error in right ascension
54 gb6_dec_err degree float error in declination
55 gb6_peak_flux mJy/beam float peak flux at 6cm
56 gb6_peak_flux_err mJy/beam float error in peak flux
57 gb6_major arcsec float FWHM major axis
58 gb6_minor arcsec float FWHM minor axis
59 gb6_flux mJy float integrated source flux
60 gb6_posangle degree float fitted major axis position east of north
61 gb6_sky mJy float local sky level
62 gb6_xpixel pixel float "X" pixel of object center
63 gb6_ypixel pixel float "Y" pixel of object center
64 gb6_eflag char flag indicating significantly extended source
65 gb6_wflag char warning flag
66 gb6_cflag char flag indicating confusion
-- SDSS photometric data (nearest) --
67 near_ra degree double right ascension values from SDSS photometric survey; for details see www.sdss.org
68 near_dec degree double declination
69 near_type integer indicates source photometric type
70 near_flags char SDSS quality flags
71 near_modelmag_u magnitude float u,g,r,i,z magnitudes
72 near_modelmag_g magnitude float
73 near_modelmag_r magnitude float
74 near_modelmag_i magnitude float
75 near_modelmag_z magnitude float
76 near_modelmagerr_u magnitude float u,g,r,i,z magnitude errors
77 near_modelmagerr_g magnitude float
78 near_modelmagerr_r magnitude float
79 near_modelmagerr_i magnitude float
80 near_modelmagerr_z magnitude float
-- SDSS photometric data (brightest) --
81 bright_ra degree double right ascension values from SDSS photometric survey; for details see www.sdss.org
82 bright_dec degree double declination
83 bright_type integer indicates source photometric type
84 bright_flags char SDSS quality flags
85 bright_modelmag_u magnitude float u,g,r,i,z magnitudes
86 bright_modelmag_g magnitude float
87 bright_modelmag_r magnitude float
88 bright_modelmag_i magnitude float
89 bright_modelmag_z magnitude float
90 bright_modelmagerr_u magnitude float u,g,r,i,z magnitude errors
91 bright_modelmagerr_g magnitude float
92 bright_modelmagerr_r magnitude float
93 bright_modelmagerr_i magnitude float
94 bright_modelmagerr_z magnitude float
-- SDSS data (spectroscopic) --
95 spec_type integer indicates source spectroscopic type values from SDSS spectroscopic survey; for details see www.sdss.org
96 spec_mjd day long integer Julian date of observation
97 spec_plate integer SDSS plate #
98 spec_fiber integer SDSS fiber #
99 spec_redshift double spectroscopic redshift
100 spec_redshifterr double error in spectroscopic redshift
101 spec_redshift_warning long integer warning flag for redshift calculation
-- Matching parameters --
102 matchflag_first integer set to -1 if FIRST is primary; otherwise equals rank of FIRST match (see matchflag note below)
103 matchflag_nvss integer set to -1 if NVSS is primary; otherwise equals rank of NVSS match (see matchflag note below)
104 distance arcsec double distance between NVSS and FIRST source in this row.
105 overlap integer set to 1 if row entry falls within overlap region; 0 otherwise (see note on positions)
106 spec_overlap integer set to 1 if row entry falls within SDSS spectroscopic overlap region; 0 otherwise (see note on positions)
107 matchtot5 integer # of NVSS or FIRST neighbors within 5" of primary source
108 matchtot10 integer # of NVSS or FIRST neighbors within 10" of primary source
109 matchtot30 integer # of NVSS or FIRST neighbors within 30" of primary source
110 matchtot120 integer # of NVSS or FIRST neighbors within 120" of primary source
111 wenss_distance arcsec double distance between primary and nearest WENSS source within 120" (see note on positions)
112 wenss_matchtot120 integer # of WENSS neighbors within 120" of primary source (see note on positions)
113 gb6_distance arcsec double distance between primary and nearest GB6 source within 120" (see note on positions)
114 gb6_matchtot120 integer # of GB6 neighbors within 120" of primary source (see note on positions)
115 near_distance arcsec double distance between primary and nearest SDSS photometric source within 60" (see note on positions)
116 bright_distance arcsec double distance between primary and nearest bright SDSS photometric source within 3", 10", 30", or 60" (see notes on positions and SDSS photometry)
117 sdss_matchtot integer # of photometric neighbors within 60" of primary (see notes on positions and updates)

Overlap regions

We have defined a region, about 2955 deg2, where the sky coverage of the four radio surveys and the SDSS photometric survey overlap. The SDSS spectroscopic survey is not yet complete; the spec_overlap parameter refers to a slightly smaller 2894 deg2 region defined by the overlap of the four radio surveys and the SDSS spectroscopic survey as of SDSS public Data Release 5. The data parameters overlap and spec_overlap are set to 1 if the row entry lies within the relevant sky region and is set to 0 otherwise.

NVSS corrections

The integrated NVSS fluxes and the NVSS major and minor axis sizes were calculated according to Condon et al. 1998 (available from the
the NVSS website), where the FWHM of the beamsize is θ = 45" everywhere. However, the code used by the NVSS catalog browser to calculate angular sizes has since been updated to account for a non-circular beam. The values in the original version of the Unified Catalog were calculated using the earlier circular beam approximation. Now available for download is an updated version, with the following changes:

Positions

Because FIRST has more accurate astrometry than NVSS, we use the FIRST position to designate the right ascension and declination for each row, where available. For an NVSS source with no FIRST match, we use the NVSS position to designate ra and dec parameters for the corresponding row in the catalog. This decision results in the following: each FIRST source will have the same near and bright optical matches every time it appears in the catalog; an NVSS source can have different designated optical neighbors if it appears in several catalog rows with different FIRST matches.

SDSS spectroscopic type

The type or class of an SDSS spectroscopic source is determined from the spectrum by the survey's spectroscopic pipeline.

SDSS photometry: nearest vs. brightest

Parameters labeled near correspond to the closest SDSS photometric match within 60". If that object is within 3" the parameters labeled bright correspond to the brightest neighbor within 3", and sdss_matchtot records the total number of neighbors within 3". If the nearest neighbor is further than 3" away but within 10" (or 30" or 60"), the bright and sdss_matchtot parameters correspond to neighbors within 10" (or 30" or 60", respectively).

SDSS photometric type

SDSS photometric type is determined morphologically in a simple way: objects whose light distribution matches the point-spread function are classified as unresolved (type=1 in this catalog); extended objects are classified as resolved (type=2 in this catalog).

SDSS magnitudes

The SDSS uses inverse hyperbolic sine ("asinh") magnitudes, often referred to as "luptitudes". For >10σ detections, "luptitudes" are practically the same as standard magnitudes. For more details, please see the SDSS website and the Lupton et al. 1999 AJ paper.

Matchflags

The FIRST-NVSS matching was done in two ways: we both positionally matched NVSS objects to FIRST sources, and positionally matched FIRST objects to NVSS sources. If a FIRST detection was the primary object, matchflag_first = -1. If an NVSS object was the primary, then matchflag_first = 1 for the row containing the nearest FIRST match, matchflag_first = 2 for the row containing the second closest FIRST match, and matchflag_first = 3 for the row containing the third closest FIRST match (within 30"). Any remaining matches within 30" are not stored, although the number of such matches can be recovered from the matchtot30 parameter. Equivalent rules govern matchflag_nvss.



Download the catalog in fits or csv (text) format

Version 1.1 of the catalog is available below. See the note above on NVSS corrections for a discussion of the changes.

For details regarding the selection of sources for each file, see KI08. The file extension .csv refers to a comma-separated-variable text file. To unzip a .gz file, type "gunzip filename" at the command line (exclude quotation marks).

  1. The complete catalog (version 1.1) is available as a compressed, tarred archive of files each with a 5° width in right ascension. There are 2,724,343 rows. File names have the format Complete_0RA5 (containing sources between 0° and 5° in right ascension). The data have the format described in the above table. To unzip the compressed archives, type "bunzip2 file.tar.bz2" at the command line (exclude quotation marks). To extract files from the unzipped archive, type "tar -xvf file.tar" at the command line (exclude quotation marks).

    Download the fits archive (271M compressed; 1.4G uncompressed) or the csv archive (227M compressed; 1.3G uncompressed).

  2. A small subset of the catalog could be useful for familiarizing yourself with the catalog format and for testing code. It covers a region of the sky about 106 deg2 in the range 150° < R.A. < 165°, 40° < dec. < 50°. It contains 16,453 rows, with the format given above.

    Download the fits file (2.1M compressed; 8.7M uncompressed) or the csv file (9.5M).

  3. The set of sources detected by FIRST and NVSS in the 2955 deg2 overlap region containing 141,881 sources. Matching radius is 25". Parameters include: uniq_id, ra, dec, first_peak_flux, first_flux, nvss_flux, gb6_flux, wenss_flux, distance, wenss_distance, gb6_distance.

    Download the fits file (5.5M compressed; 8.7M uncompressed) or the csv file (9.2M).

  4. The set of sources detected by all four radio catalogs and the SDSS in the 2955 deg2 overlap region containing 4732 sources. Matching radii are: (FIRST-NVSS) 25", (FIRST-WENSS) 30", (FIRST-GB6) 70", (FIRST-SDSS) 2". In addition to those listed in item 3 above, parameters include: near_type, near_flags, the near model magnitudes and their errors.

    Download the fits file (470K compressed; 625K uncompressed) or the csv file (837 K).

  5. The set of galaxies with optical spectra detected by NVSS, FIRST, WENSS, and SDSS in the 2955 deg2 overlap region containing 2885 sources. Matching radii are: (FIRST-NVSS) 25", (FIRST-WENSS) 30", (FIRST-SDSS) 2". Parameters include: uniq_id, ra, dec, first_peak_flux, first_flux, nvss_flux, wenss_flux, near_type, near_flags, distance, wenss_distance, near_distance, the near_modelmags and their errors, and the SDSS spectroscopic parameters.

    Download the fits file (298K compressed; 439K uncompressed) or the csv file (570K).

  6. The set of quasars with optical spectra detected by NVSS, FIRST, WENSS, and SDSS in the 2955 deg2 overlap region containing 1288 sources. Matching radii are: (FIRST-NVSS) 25", (FIRST-WENSS) 30", (FIRST-SDSS) 2". For each source, the file includes all 74 parameters originally provided in the Schneider et al. (2007, AJ, 134, 102) SDSS DR5 known quasar catalog. Additionally, parameters from the unified catalog include: uniq_id, ra, dec, first_peak_flux, first_flux, nvss_flux, wenss_flux, the near_modelmags and their errors. The file includes a header describing each of the columns.

    Download the text file (755K).

  7. The set of isolated FIRST-NVSS sources. It is comprised of objects observed in both FIRST and NVSS (matched within 15") with no other FIRST detection within 30" of the NVSS source. The file contains all 109,825 such sources within the 2955 deg2 overlap region. Parameters include: uniq_id, ra, dec, first_peak_flux, first_flux, nvss_flux, wenss_flux, gb6_flux, distance, wenss_distance, gb6_distance described above. The two FIRST fluxes and the NVSS flux are necessary and sufficient to divide the data into the three morphological classes described in KI08.

    Download the fits file (4.1M compressed; 6.8M uncompressed) or the csv file (7.0M).

  8. The set of isolated FIRST-NVSS sources with a photometric SDSS match containing 42,144 sources. This file is a subset of sources from item 7 above. FIRST-SDSS matching radius is 2". It includes all data parameters described in items 7 above, as well as SDSS data (near_ra, near_dec, near_distance, near_type, near_flags, model magnitudes, and model magnitude errors).

    Download the fits file (3.7M compressed; 5.7M uncompressed) or the csv file (6.9M).

  9. The set of 9337 high-redshift (z > 1) galaxy candidates detected by FIRST, NVSS, WENSS (and not SDSS). These objects were selected based on their lack of an optical counterpart, and their steep radio spectra (α 9220 < -0.5). (See notes on updates.) The selection of these sources is described in Section 5.2 of KI08. These objects: are detected by FIRST, NVSS, and WENSS, have no SDSS match within 3", have compact radio morphology, and have steep radio spectra (α 9220 < -0.5). (Please note: these objects are a subset of sample D in table 8, not of sample E as is stated in that table. I.e.: GB6 criteria were not used in their selection.)

    Download the fits file (2.5M compressed; 5.0M uncompressed) or the csv file (5.8M).

  10. An auxiliary text file containing NVSS angular sizes in agreement with the NVSS catalog browser. See the note above on NVSS corrections for details. The file contains one row for each NVSS source in the Unified Catalog. Parameters include: NVSS id, major axis [arcsec], major axis flag, minor axis [arcsec], minor axis flag. Flags are 0 for unresolved (upper limit only), 1 for resolved (actual angular size).

    Download the csv file (12M compressed; 38M uncompressed).


Updates

(Posted 10 Oct 2008) Correction: The parameter sdss_matchtot records the total number of SDSS neighbors within 60", not within the pre-defined search radius as was previously stated (see note on SDSS photometry).

(Posted 11 Oct 2008) Correction: The original set of high-redshift (z > 1) galaxy candidates for download mistakenly included sources with flat spectra and uncorrected NVSS fluxes. The version currently posted here is correct. Note that the corrected version is smaller than the dataset referred to in KI08.

(Posted 9 March 2009) Now available: version 1.1 of the catalog contains corrected NVSS fluxes and major/minor axis sizes. See the note above on NVSS corrections for details. We thank Dr Haida Liang, Dr Hui Shi, and Dr Jess Broderick for bringing this issue to our attention.



If you use this catalog in your research, we would greatly appreciate it if you add a link to this site and a reference to Kimball & Ivezić 2008, AJ, 136, 684 in your publications.
   
   Amy E Kimball
   National Radio Astronomy Observatory
   PO Box O  
   Socorro, NM 87801
   USA

   Željko Ivezić
   Department of Astronomy
   Box 351580, U.W.
   Seattle, WA   98195-1580
   USA

Please send your questions, comments, or suggestions to: akimball@nrao.edu