Optical Corrections to the Véron-Cetty and Véron Quasar Catalogue

Abstract

Fixes are presented to be applied to the Véron-Cetty and Véron Quasar Catalogue, 13th edition. These are comprised of 39 de-duplications, 380 astrometric moves of 8 + arcsec of which 31 are over 10 arcmin, and 30 indicated de-listings.

1 INTRODUCTION

Quasars have been catalogued into complete collections from the earliest days of quasar surveys. The two most prominent such catalogues have been the Véron-Cetty and Véron (VCV) Catalogue of Quasars and Active Nuclei (Véron-Cetty & Véron 2010), published in 13 editions from 1984 to 2010, and the Hewitt and Burbidge catalogue, which published its last edition in 1993 (Hewitt & Burbidge 1993). These catalogues recorded quasar positions based on optical or radio or X-ray surveys in which optical positions were not always published or were often approximated into tiles of sky specified by the quasar names, e.g. ‘0450−132’ referred to a tile of sky bounded by the corners B1950 04h50m–13d12m and 04h51m–13d18m (the last digit of the name being tenths of a degree).

With the advent of large optical quasar surveys such as 2df QSO Redshift Survey (Croom et al. 2004) and Sloan Digital Sky Survey (Abazajian et al. 2009), we have moved to an optical standard of quasar cataloguing where optical photometry and arcsecond-accurate astrometry are the norm. To complete the transition to this new standard, there is accordingly a need to bring the older 1970s–1990s data into conformity. Most of the old data already bear arcsecond-accurate astrometry, by which red–blue photometry can be sourced from optical catalogues such as that of the Cambridge Automatic Plate Measuring (APM) machine (McMahon & Irwin 1992) and the United States Naval Observatory (USNO-B; Monet et al. 2003). In their recent releases, VCV have fixed considerable old data up to the optical standard, but a residue remains of quasars with only approximately known or mistaken positions. Of course, the latter of these are elusive to identify without some indicator to select them.

I have recently developed an unpublished but publicly available quasar catalogue, the ‘Million Quasars’ (Milliquas; Flesch 2012) catalogue, as a by-product of the recently published ‘Atlas of Radio/X-ray Associations’ (ARXA; Flesch 2010). It is serving as a platform in support of ongoing development, and as a resource for bulk querying of quasars and quasar candidates. It is built to be an optical catalogue of arcsecond accuracy, and I have relied on the VCV 13th edition as the authority on the early quasars. Nevertheless, the process of merging the VCV data with optical APM/USNO-B data has highlighted some hundreds of quasars which either match to no optical signature or to an untypical signature; this thus serves as the aforementioned indicator of quasar data which are poorly or mistakenly sited. Investigation of these has led me to identify in the VCV data:

• • Thirty-nine duplicate objects where a correctly sited quasar is matched by another, poorly sited, object which arose through either error or inexactitude by the original author (OA) in the discovery paper, or as a result of a cataloguing error.

• • Three hundred and eighty objects for which I find that the VCV position is offset at least 8 arcsec from the true optical position which I present. Thirty-one of these quasars require a move of at least 10 arcmin.

• • Thirty objects for which I recommend de-listing, as I have found, by a careful inspection of all evidence, either are not quasars or have information of such poor quality that the object is lost beyond any mechanism of recovery, i.e. even if the object were re-found, it could not be recognised as being the previous object.

These 449 objects represent all those where VCV objects were not well supported by optical data, and found to require a fix. For most of these objects, the diagnosis and correction have been clear. However, completeness requires me to include a few puzzling objects where the true identification is still unclear, for which I provide a ‘best’ available optical object only, rather than something certain. I flag such cases.

2 THE METHOD

The recent publications of the Nonparametric Bayes Classifier Kernel Density Estimate (NBCKDE: Richards et al. 2009) and SDSS-Extreme-Deconvolution QSOs (XDQSO: Bovy et al. 2011) photometric quasar candidate catalogues, plus the radio-/X-ray-associated candidates published in ARXA, now highlight good optical candidates for quasar searches which hitherto would have been too hard. These and other tools are as follows.

• • Radio/X-ray associations to optical objects as presented in ARXA, and similarly in Milliquas which has updated versions of these. This is especially useful for originally X-ray-selected quasars, such as from the Einstein satellite, which are thus also expected to be X-ray associated in ARXA.

• • The photometric quasars published by NBCKDE and XDQSO. NBCKDE provides reliable photometric redshifts, and some needed XDQSO redshifts were kindly provided by Adam Myers.

• • Optical magnitudes and colours to match to the original paper, although the systematic offsets of that paper need to be ascertained; see the ‘personal equation’ paragraph of Section A5 in Flesch & Hardcastle (2004).

• • Recently surveyed quasars/galaxies of matching magnitude and redshift, especially those which are offset by right ascension only or declination only, as such errors do happen but rarely in both simultaneously.

• • Finding charts and positional information given by the OAs, which later can have been overlooked or wrongly superseded.

• • Online optical look-ups, being the SDSS-DR7 ¹ and DR8 ² finding charts and the Digitised Sky Survey. ³

Sometimes these tools immediately yield a strong candidate. An example is my search for the optical object for Q 1233 + 4749, an approximately sited quasar found serendipitously in the search for ‘primeval galaxies’ by Thompson & Djorgovski (1995), who presented no astrometry or photometry for this object, but their search was for objects of 19.5≤R<24 and B−R≈1.5. This object should be within the tile of sky bounded by the corners B1233 + 4749 and B1234 + 4750; VCV placed it at the B1233 + 4749 corner. Inspection of all available quasars and radio-/x-ray-associated objects and photometric quasar candidates within this tile of sky yielded only the two objects displayed in Table 1. The first of these is the VCV-catalogued object which has no photometry and, being approximately located, matches to no optical object. The second object, offset 521 arcsec from the first, has R = 20.1, B = 21.2 at J123614.2 + 473259, i.e. B123351.6 + 474930, with both radio and X-ray associations—the X-ray association is a ROSAT All-Sky Survey, from Voges, W., et al. 2000, on http://www.xray.mpe.mpg.de/rosat/survey/rass-bsc/ and ../rass-fsc/ source with a 30-arcsec error circle. ARXA assigns this object an 80% likelihood of being a quasar based on the radio association and stellar point spread function. This object has the right photometry for the discovery paper, and is clearly the true Q 1233 + 4749.

Table 1. Objects Found for Q 1233 + 4749.

Of course, many searches for approximately sited quasars are not so straightforward and yield multiple candidates including some slightly outside the B1950 sky tile, but usually one candidate is found to be the clearly best match. A typical example is the search for Q 1510 + 155, from Sargent, Boksenberg, & Steidel (1988), which gives no photometry or astrometry, but states this is one of Cyril Hazard’s quasars, which is typically a bluish v = 17.5–18.5. Hazard had a large collection of unpublished quasars which he lent out to researchers, and this paper states that ‘the accurate positions of these objects will be reported (in) Sargent, Hazard, and McMahon (1988)’, which, however, never appeared. The list of all quasars and candidates in the tile of sky bounded by B1510 + 155 and B1511 + 156 is displayed in Table 2. There, we first see the VCV object sited at the B1510 + 155 corner (although VCV usually positioned their approximate objects at the tile centres) with redshift of 2.106, then the third object is an SDSS quasar of redshift 2.110 which is a very strong candidate for duplication. Next is an NBCKDE candidate with photometric redshift of 2.225, also a good match, and the remaining objects have non-matching redshift. To select between the SDSS quasar and the NBCKDE candidate, we look at the object magnitudes: the SDSS quasar fits the Hazard profile of v≈18 and the NBCKDE candidate is two magnitudes too faint, so the SDSS quasar is selected as the true Q 1510 + 115. This quasar is catalogued in VCV as ‘SDSS J15128 + 1119’, so we have now found a duplication in VCV. The Q 1510 + 115 row needs to be deleted, but I recommend renaming ‘SDSS J15128 + 1119’ as ‘Q 1510 + 115’ in order to retain the original name.

Table 2. Objects Found for Q1510 + 115.

Sometimes there is no help from the radio/X-ray data, and no photometric-redshift. For only nine such approximately sited cases only optical matching is available, and for them I designate the best bluish object with magnitudes consistent with the discovery paper. I flag these objects (in column 10 of Tables 5–7) according to how confident the selection is, and it is likely that I have missed the true object for some of these nine cases.

Some objects are corrected by identification of simple offsets, such as IXO 15 which is at the true location J033311.3− 361137 (Woo 2008), but recorded in VCV at J033311.9− 371135, i.e. offset 1° to the south. Such simple transcription errors account for ≈50 of the fixes presented here.

A few quasar discovery papers could be termed ‘treasure hunt’ papers because the presented astrometry is not of quasars, but of radio/X-ray sources or galaxies with offsets supplied, so that the reader must follow the trail to find the promised quasars. Notable among these is Appenzeller et al. (1998), which lists 674 X-ray sources with offsets to objects which VCV diligently followed but ultimately missed 76 quasar positions which I present here. Also notable is Arp (1981), which presents central galaxies with offsets provided to ‘companion’ galaxies, and from there, further offsets to the described quasars. Arp’s treasure hunt was made doable by his arcsecond-accurate offsets, and I present the search for his first object, ‘NGC 157#1’, as an example. Arp’s Table 1 lists the secondary galaxy as offset 30 arcmin north of NGC 157—this will be MCG -1-02-034, which is 29.7 arcmin due north of NGC 157. Arp now specifies a quasar of V = 19.0 at a distance of 119 arcsec from the secondary galaxy, and I show, in Table 3, the list of photometric candidates within 250 arcsec of that galaxy. There are three objects, the bottom of which is the galaxy, and the middle line shows a BOSS candidate at exactly 119 arcsec offset from the galaxy, R = 18.8 and B = 19.1, which is undoubtedly Arp’s quasar. In this way, I have identified seven object positions which VCV recorded only as ‘approximate’. SDSS has since resurveyed these objects, finding one confirmed quasar, five photometric quasars, and one star.

Table 3. NGC 157#1: Objects Within 250 arcsec from MCG -1-02-034.

Many discovery papers present finding charts for their objects, which I have used in tandem with STScI Digitised Sky Survey to secure the exact astrometry. The cut-off for inclusion in this paper is a move of 8 arcsec. Smaller moves are not always trivial and are provided in the online Milliquas catalogue.

3 DUPLICATES

Thirty-nine duplicates in the VCV 13th edition are presented in Table 4. Each duplication consists of a master object which has accurate optical astrometry, and the duplicate to be removed, which is typically positioned onto blank sky. The offsets range from 48 arcsec to across the sky. I find that 13 of these duplications arose because the original published astrometry was approximate only, 6 other cases were due to astrometric error by the OAs, and 20 cases were cataloguing errors, of which 11 were simple transcription errors causing an N/S or E/W offset. Table 4 lists the duplicate first and then the master. Columns are (1) line number; (2) VCV name of the duplicate object; (3) duplicate object J2000; (4) VCV redshift; (5) VCV V magnitude; (6) VCV table number (1 = QSOs, 2 = Bl Lacs, 3 = active galactic nebulae, AGNe); (7) VCV reference number for the original paper; (8) astrometric offset from the duplicate to the master; (9) VCV name of the master object; (10) master object J2000; (11) master object redshift; (12) optical red magnitude; (13) optical blue magnitude; and (14) a comment explaining the original error and/or fix. For 12 of these objects, I recommend reassigning the duplicate name to the master because the duplicate name is the original historic name.

Table 4. Thirty-Nine Duplications in VCV.

^{a a}This Einstein HRI-detected object was evidently misplaced and lost in preparation by the original authors, Reichert et al. (1982). It was presented out of sequence in their RA-ordered Table 4, as ‘2352 + 073’ between ‘2353 + 283’ and ‘2353 + 072’, by which it is seen that the RA was originally inscribed as ‘2353’. The Einstein position presented was (B1950) 23 52 47 + 07 16 15, which is, however, only a 3σ detection without any eligible optical counterpart—so the original optical object was not recovered. The authors knew that they had lost this object, for on their page 440, line 12, they wrote: ‘Four … (HRI) counterparts’ followed by only three names sans this object. An X-ray-associated doppelganger of the same redshift and similar magnitude, RX J00013 + 0728, is at (B1950) 23 58 44 + 07 11 45, which would have been written as ‘2358 + 072’. If that was the original object, then a double transcription error is indicated, causing the loss.

4 ASTROMETRIC MOVES

A total of 380 VCV objects are found to be positioned 8 + arcsec from their true optical positions. I present these in four tables: Table 5 shows offsets of 30 + arcsec, Table 6 shows offsets of 15–29 arcsec, Table 7 shows offsets of 8–14 arcsec, and Table 8 shows 82 objects which came late to hand, mostly from VCV Table 3 of AGNe. Columns are as follows: (1) line number; (2) VCV object name; (3) VCV J2000; (4) VCV redshift; (5) VCV V magnitude; (6) VCV table number (1 = QSOs, 2 = Bl Lacs, and 3 = AGNe); (7) VCV reference number for the original paper; (8) astrometric move required, in arcsec; (9) optical object J2000; (10) flag on optical object: p = a photometric quasar from NBCKDE/XDQSO, r = radio-associated, x = X-ray-associated, ! = a standout optical-only fit with no other good candidates; ? = best optical-only fit but other candidates present; ?? = good optical-only fit but other good candidates present; (11) optical red magnitude; (12) optical blue magnitude; and (13) a comment explaining the original error and/or fix.

Table 5. One Hundred and Five Moves of 30 + arcsec.

Table 6. Eighty-Nine Moves of 15–29 arcsec.

Table 7. One Hundred and Four Moves of 8–14 arcsec.

Table 8. Eighty-Two Additional Objects

5 DELETIONS

I recommend de-listing 30 VCV-catalogued objects because they are either not quasars or are so poorly described that recognition is excluded. It is not suitable to make a table out of these objects because there are so many different circumstances to describe, so a simple list follows. One paper accounts for 11 of these objects, which I will discuss at the end.

1. NGC 3726 B1 is a star. Approximately located in VCV at J113448.1 + 470025, z = 1.13, it is an Arp (1981) treasure hunt object, as discussed above. Arp placed it at 100 arcsec from MCG 8-21-061, and it is seen exactly there, but SDSS-DR8 finds it to be a star, SDSS J113456.62 + 470014.8, r = 17.9, b = 18.5.

2. WEE 140 from Weedman (1985), putatively z = 2.27, is found to be a star by SDSS-DR5, SDSS J160250.34 + 280541.4. Weedman did say in his Table 2 notes that the lines for this object were weak.

3. Q 1052 + 04 from Lanzetta et al. (1991), approximately located in VCV at J105505.2 + 041400, v = 18.1, z = 3.391, does not exist. NED ⁴ identifies it with SDSS J105433.04 + 040027.4, z = 3.301, but that object is 34 arcsec outside of the B1052 + 04 sky tile, and the redshift offset of 0.09 is large. Another candidate is NBCK J105510.14 + 034730.0, v = 18.5, which, however, has a photometric redshift of 0.535. VCV’s catalogue paper explains (illustrated by its Figure 2) that old z = 3.3 redshifts were often wrong because low-z Mg ii −2800 Å lines were mistaken for Lyman alpha. In either scenario, this object was wrongly presented. No other good candidates are seen, and the original paper gives no help, so I recommend de-listing.

4. Q 0124−365 from Savage et al. (1984) is lost. It was presented on an object-poor finding chart which I could not place anywhere in or near B0124-365, nor could VCV. Perhaps there was a typo in the name, or maybe it was the wrong finding chart. No astrometry was stated in the paper, so there is no means of recovering this object.

5. MC 1227 + 120 never existed. It was catalogued in Burbidge, Crowne, & Smith (1977) with reference to a paper ‘in press’, that paper being Smith et al. (1977), in which however it did not appear. This object is only an artefact of the literature.

6. 1ES 1249 + 174W is lost, if it ever existed. It was presented as one of two optical candidates for a single Einstein source by Perlman et al. (1996). The E object is an SDSS-DR7 quasar, this W object, a ‘BL candidate’, had no redshift presented—the VCV-listed redshift belonged to the E object. This W object was not measured and is unseen.

7. 2E 1510 + 3902, an Einstein X-ray source at J151230.7 + 385051 with optical attributes of v = 19.0 and z = 0.228, appears to be SDSS J151224.30 + 385112.7 of v = 18.2 and z = 0.202; there are no other eligible objects. The redshift is a poor match, and the discovery paper Reichert et al. (1982), which presented spectra for its new AGNe, stated problems with this object (Table 5, footnote b) and presented no spectrum. Recommend de-listing due to poor fit and no other candidate.

8. X404-23, from Zamorani et al. (1999), is an X-ray source without an optical object. The finding chart presents a marked object which is, however, 30 arcsec outside the X-ray error circle. As Zamorani et al. state, ‘None of them is a convincing identification’.

9. Q 0411−789, approximately located in VCV Table 3 with V = 16.0 and z = 0.019, is the only catalogued object from Campusano & Pedreros (1978). At such high latitude, the B1950 tile of sky is quite small, 3 × 1 arcmin, and no v = 16 galaxy is seen there. It is meant to be a radio-detected object, but the nearest SUMSS Sydney University Molonglo Sky Survey, from Murphy, T., Mauch, T., Green, A., Hunstead, R. W., Piestrzynska, B., Kels, A. P., Sztajer, P. 2007, MNRAS, 382, 382 radio detection is 2 arcmin beyond the box.

10 and 11. Two nameless Table 3 entries, row 6 502 (J081121.6 + 631943) and row 12 235 (J101805.6 + 004318), are from Appenzeller et al. (1998), but they were identified as stars in that paper.

12 and 13. Two objects from Brissenden et al. (1987), approximately sited in VCV, are not seen at all. They are 1H 0217 − 639 (J021906.4 − 634428, z = 0.073) and 1H 2044−032 (J204711.6−030400, z = 0.015). These objects should be bright, V<18, but no photometry is presented in a paper which does give photometry for its other objects. 1H 2044−032 was also presented with a 69-mJy radio detection, but the nearest NVSS NRAO VLA Sky Survey catalog, from Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F., Perley, R. A., Taylor, G. B., Broderick, J. J. 1998, AJ, 115, 1693 source (5 + mJy) is 2 arcmin outside the box.

14. IRAS 22040 + 0332 (J220634.2 + 034655, z = 0.064, VCV Table 3) from Gu et al. (1995) is not seen in or near the prescribed B1950 tile of sky. No finding chart or optical magnitude was presented. This IRAS Infrared Astronomical Satellite source is not found in any IRAS source catalogue. No NVSS radio sources are within 5 arcmin.

15 and 16. SDSS J09557 + 2525, z = 2.262, and SDSS J10162 + 2649, z = 0.383, were removed by SDSS-DR8 as they were satellite streaks, SDSS J095546.30 + 252534 and SDSS J101615.16 + 264902.4.

17. ROTSE J11568 + 5427, v = 18.1, z = 1.02, was presented on Astronomer’s Telegram #1515 (http://www.astronomerstelegram.org) as an optical transient with suspected AGN origin. However, subsequently, Telegram #1644 announced it to be ‘an extremely luminous type II supernova at z = 0.21’.

18 and 19. Q 0000−029 (J000301.4−024141, v = 18, z = 2.31) and Q 2355 + 003 (J235803.4 + 004000, v = 19, z = 2.84) were so-called ‘optically violent variables’ from Zhan & Chen (1986), which appeared on a single UKST UK Schmidt Telescope Sky survey plate, but were not detected by deeper plates or CCD scans, and are today not seen on the DR8 finding charts. The authors theorised these were AGNe but concluded ‘We cannot reject other explanations including galactic nova. . .’. Also to be considered are plate artefacts, since it is most unlikely to find two such optical transients on a single plate.

20–30. All quasars from Afanas’ev et al. (1990b), namely SA68 #110 (v = 19.4, z = 0.78), SA68 #105 (v = 20.6, z = 0.71), SA68 #090 (v = 21.3, z = 1.00), SA68 #094 (v = 19.3, z = 1.08), SA68 #143 (v = 20.7, z = 2.11), SA68 #095 (v = 21.3, z = 1.24), M82 #95 (v = 19.4, z = 1.01), M82 #69 (v = 19.4, z = 0.93), M82 #22 (v = 19.6, z = 0.96), SA57 #216 (v = 22.2, z = 0.77), and SA57 #431 (v = 21.7, z = 0.94). Afanas’ev et al. presented these quasars from three fields, SA68 (i.e. ‘selected area 68’) to a depth of B<22, M82 to B<22.5, and SA57 to B<23.5. The stated astrometric accuracy is <2 arcsec. All three fields are now covered by the SDSS-DR8 finding charts; thus, we are able to optically investigate Afanas’ev et al.’s three fields to their plate limits. Unfortunately, few, if any, of their quasars are seen.

Field SA68: Afanas’ev et al. present six new quasars in this field, but five of them have no credible optical counterparts in the DR8 finding charts, while one of them, SA68 #95, b = 21.7, is offset 6 arcsec from SDSS J001735.30 + 155207.1, g = 22.7, and redshift unknown. This near-miss looks random.

Field M82: Seven quasars were presented, consisting of four already-known ones, HOAG 1, HOAG 2, and HOAG 3 from Burbidge et al. (1980) and NGC 3031 U4 (née M82 #4) from Arp (1983), and three new ones, M82 #22, M82 #69, and M82 #95. The four known quasars are all present in the DR8 finding charts, but the three Afanas’ev et al. objects are not seen at all.

Field SA57: Seven quasars were presented, consisting of five already-known ones, KKC 30, KKC 36, KKC 37, KKC 41, and KKC 43 from Koo, Kron, & Cudworth (1986), and two new ones, SA57 #216 and SA57 #431. The five KKC quasars are all present in the DR8 finding charts, and here we finally see evidence of Afanas’ev et al. quasars: SA57 #216, b = 22.6, matches exactly with SDSS J130841.02 + 291857.4, g = 22.5, and redshift unknown. In the case of SA57 #431, Afanas’ev et al. mention that this quasar is near another object, and indeed at that location there is a red object close to a blue object. However, the blue object is bright, v = 17.3, while Afanas’ev et al.’s quasar has v = 21.7. DSS POSS National Geographic - Palomar Observatory Sky Survey-I confirms that the blue object was also v = 17 in the 1950’s epoch, so it is unreconcilably bright.

The final outcome is a head-scratcher, but the SDSS-DR8 finding charts are clear: there is only one credible match to Afanas’ev et al.’s 11 objects. Searches at increasing radii turn up nothing useful. With this performance, the one apparent match should be discounted as a possible random artefact; therefore, all of these should be de-listed. Afanas’ev et al. published many other quasars in other papers, and those quasars are seen, confirmed and catalogued.

6 SUMMARY

This paper presents 39 duplication removals, 380 astrometric moves of 8 + arcsec, and 30 de-listings, to be applied to the Véron-Cetty and Véron Quasar Catalogue, 13th edition. This is to bring the VCV data up to the astrometric standard of today’s large optical quasar surveys, and so enable accurate inclusion into dynamic databases such as NED and SIMBAD (http://simbad.u-strasbg.fr/simbad).

ACKNOWLEDGMENTS

Great thanks to Mira and Philippe Véron for their definitive catalogue, for critiques of my early work for this paper, and encouragement to publish. Thanks to Adam Myers for XDQSO redshifts for some objects, and to Cedric Ledoux and Mike Irwin for help with single objects. Thanks also to Steve Willner for supplying a copy of the Afanas’ev et al. paper.