The Million Quasars (MILLIQUAS) catalogue
version 8 (Final Edition)
Eric Wim Flesch, 2 August 2023
================================================================================
This is a compendium of 907,144 type-I QSOs and AGN, largely complete from the
literature to 30 June 2023. 66,026 QSO candidates are also included, calculated
via radio/X-ray association (including double radio lobes) as being 99% likely
to be quasars. Blazars and type-II objects are also included, bringing the
total count to 1,021,800. 60.7% of all objects show Gaia-EDR3 astrometry.
Citations are provided in a single file, milliquas-references.txt. This is the
final edition of Milliquas.
Low-confidence/quality or questionable objects (so deemed by their researchers)
are not included in Milliquas. Additional quality cuts can apply as detailed
in the HMQ paper (Flesch 2015,PASA,32,10). Full QSO/AGN classification is via
spectral lines yielding a reliable spectroscopic redshift; two spectral lines
are required, or one spectral line refining a compatible photometric redshift.
Obscured AGN with redshift from the host only, are taken to be type-II objects.
Some legacy quasars with no good spectra nor radio/X-ray association were
flagged by Gaia-EDR3 as 5-sigma moving (i.e., stars), and so removed from
Milliquas. All objects are de-duplicated across source catalogues. The aim
here is to present one unique reliable object per each data row.
Two NIQs offset <2 arcsec can be reported as a single object if within the same
host. Lenses are reported as single objects onto the brightest quasar image;
Milliquas is not a catalogue of lenses.
The catalog format is simple, each object is shown as one line bearing the
J2000 coordinates (of whatever epoch), its original name, object class, red and
blue optical magnitudes, PSF class, redshift, the citations for the name and
redshift, plus up to four radio/X-ray identifiers where applicable.
Please cite as Milliquas v8, Flesch, E.W. 2023,OJAp,6,49. (arXiv:2308.01505)
Questions/comments/praise/complaints may be directed to me at eric@flesch.org.
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
milliquas.txt 188 1021800 The catalogue
milliquas-references.txt 150 2399 index to citations
Milliquas-ReadMe.txt 80 . This file
--------------------------------------------------------------------------------
Byte-by-byte Description of the Milliquas (Million Quasars) file:
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 11 F11.7 deg RA right ascension J2000 (degrees)(0)
13- 23 F11.7 deg DEC declination J2000 (degrees)(0)
26- 50 A25 --- Name ID from the literature, or J2000 (1)
52- 55 A4 --- Type classification of object, and associations (2)
57- 61 F5.2 mag Rmag red optical magnitude (3)
63- 67 F5.2 mag Bmag blue optical magnitude (3)
69- 71 A3 --- Comment comment on optical object (4)
73 A1 --- R red optical PSF class (5)
75 A1 --- B blue optical PSF class (5)
77- 82 F6.3 z Z redshift from the literature or estimated (6)
84- 89 A6 --- Cite citation for name (7)
91- 96 A6 --- Zcite citation for redshift (7)
98-119 A22 --- Xname X-ray ID, if any (8)
121-142 A22 --- Rname radio ID, if any (8)
144-165 A22 --- Lobe1 radio lobe ID or extra R/X ID, if any (8)
167-188 A22 --- Lobe2 radio lobe ID or extra R/X ID, if any (8)
--------------------------------------------------------------------------------
Note (0): These are to 7 decimals with a miniscule round-up to avoid truncation.
60.7% of rows show Gaia astrometry, 27.8% show SDSS astrometry, and 8.3%
show Pan-STARRS astrometry. See Note (4) for the individual indices.
Note (1): Nameless radio/X-ray associated objects here show the J2000 position
in HHMMSS.SS+DDMMSS.S for the convenience of the user. If needing a name
for it, just preface the J2000 with "MQ", e.g., MQ J000001.89+443053.8 .
Note (2): Legend of type/class:
Q = QSO, type-I broad-line core-dominated, 860100 of these.
A = AGN, type-I Seyferts/host-dominated, 47044 of these.
B = BL Lac type object, 2814 of these. (FSRQs are typed as QSOs here)
K = NLQSO, type-II narrow-line core-dominated, 6048 of these.
N = NLAGN, type-II Seyferts/host-dominated, 39768 of these. Incomplete,
and includes an unquantified residue of legacy NELGs/ELGs/LINERs,
plus some unclear AGN. This is the catch-all category.
S = star classified but showing quasar-like photometry and radio/X-ray
association, thus included as a quasar candidate; 124 of these.
R = radio association displayed, see Note (9) for index.
X = X-ray association displayed, see Note (9) for index.
2 = double radio lobes displayed (declared by data-driven algorithm).
Note (3): Optical photometry is 97% from ASP (2017,PASA,34,25) which presents
data from the APM (http://www.ast.cam.ac.uk/~mike/apmcat), USNO-A & USNO-B
(http://tdc-www.harvard.edu/catalogs/ub1.html), and SDSS (http://sdss.org).
Optically faint photometry is supplemented from Pan-STARRS and DESI data.
APM/USNO-B galaxies < mag 17 are usually shown too bright due to PSF
modelling. Integer magnitudes (e.g., 22.00) are estimates if both bands are
integer or one band empty. Note: many SDSS magnitudes are
"extinction-corrected" ~0.3 mag brighter than observed.
Note (4): Legend:
p = optical magnitudes are POSS-I O (violet 4050A) and E (red 6400A).
These are preferred because O is well-offset from E, and those plates
were always taken on the same night, thus the red-blue color is
correct even for variable objects. Epoch is 1950's.
j = optical magnitudes are SERC J (Bj 4850A) and R (red 6400A) from the
POSS-II or UKST surveys. Red-blue color is less reliable because the
red & blue plates were taken in different epochs, i.e., years apart.
b = blue magnitude is Vega 4400A (Johnson), red is 6400A (Cousins).
g = blue magnitude is SDSS-type green 4900A, red is r 6200A. Also denotes
SDSS astrometry if not otherwise flagged.
n = blue magnitude is Pan-STARRS green 4900A, red is r 6200A. Also denotes
Pan-STARRS astrometry if not otherwise flagged.
d = magnitudes are DES DR2 AB r & g. Also denotes DESI astrometry if not
otherwise flagged.
u = blue magnitude is SDSS ultraviolet 3850A.
v = red magnitude is visual 5500A, or estimated from a sky chart/viewer.
i = red magnitude is infrared 7500A.
z = red magnitude is infrared z 8600A.
r = red magnitude is r 6200A.
(blank) = red alone is 6400A (Cousins); if both mags present = estimates.
G = Gaia-EDR3 astrometry shown, precessed to J2000 by CDS. If 'G' is alone
then the magnitudes are Gaia RP & BP, or Gaia G if red band only.
N = Pan-STARRS astrometry shown. If 'N' is alone, PS photometry also.
D = DES DR2 astrometry shown. If 'D' is alone, DES photometry also.
+ = variability nominally detected in both red/blue over multi-epoch data.
m = proper motion detected. If from Gaia-DR2/3 ('G' also present in this
field), then this is proper motion or parallax which usually signifies
a star, but optical centroids can deflect within the optical gradient
of a near moving neighbor or if either object is variable; some will
be quasars. If not Gaia, it is from USNO-B which is nominal only.
e = USNO-B1.0 "epoch 2000" projected location based on nominal proper
motion, can miss true location by many arcsec.
% = swap of two "unplugged" SDSS spectra which crossed wires (7 of these).
a = object is host-dominated with faint nuclear activity, such as an SDSS
pipeline galaxy with an AGN subclass or AGN-classed elsewhere, see its
citation. Milliquas class is 'A' if BROADLINE, else 'N'. (see note 2)
Note (5): The APM, USNO-B, SDSS, and Pan-STARRS provide PSF class, albeit using
different criteria. These are shown here as:
- = point source / stellar PSF (APM notation: -1, here truncated)
1 = fuzzy / galaxy shape (APM notation: 1 and some 2)
n = no PSF available, whether borderline or too faint to tell, etc.
x = unseen / unclear in this band.
Note (6): Spectroscopic/grism redshifts are required for objects classified as
Q/A/K/N and is optional for B (BL Lac type). Photometric redshifts are
rounded to 0.1z (and can be identified by that rounding) and are either
taken from the cited catalogue or calculated here using the four-colour
method of Flesch 2015,PASA,32,10, Appendix 2, using 4 colours from SDSS
ugriz, Pan-STARRS grizy (or ogriz with calibrated POSS-I O), or WISEA
colours B-R, R-W1, W1-W2, and W2-W3. Photometric redshifts apply onto a
quasar template mostly and not so well to galaxies. GAIA3 candidates show
the "QSOC" estimated redshift provided by Gaia-DR3, rounded here to 0.01z.
Note (7): Legend (with counts of name and redshift) and references are indexed
in the accompanying file "milliquas-references.txt". The citation for the
classification (e.g., that the object is a quasar) is from either the name
or redshift citation.
Note (8): Four columns of Radio/X-ray detections are presented:
1st column: best X-ray detection (i.e. highest probability association).
2nd column: best core Radio detection.
3rd column: a radio lobe if the type (see note 2) shows a "2",
otherwise this is an additional radio or X-ray detection.
4th column: a radio lobe if the type (see note 2) shows a "2",
otherwise this is an additional radio or X-ray detection.
Legend of Radio/X-ray detection prefixes and catalog home pages:
FIRST: VLA FIRST survey, 13Jun05 version, https://sundog.stsci.edu
VLA (abbrev of VLASS1QLCIR): VLASS Quick Look, https://cirada.ca/catalogues
RACS: Rapid ASKAP Continuum, https://research.csiro.au/racs/ , source file.
RACD: RACS as above, but from their main detection ("Gaussian") file.
ILT: LoTSS-DR2, https://lofar-surveys.org/ , main source catalogue.
ILD: LoTSS-DR2 as above, but from their main detection ("Gaussian") file.
NVSS: NRAO VLA sky survey, https://www.cv.nrao.edu/nvss
SUMSS: Sydney U. Molonglo, http://www.astrop.physics.usyd.edu.au/sumsscat/
MGPS: Molonglo galactic plane, www.astrop.physics.usyd.edu.au/mgpscat/
1RXH: ROSAT HRI (high resolution), http://cdsarc.cds.unistra.fr/cat/?IX/28A
2RXP: ROSAT PSPC (proportional), https://cdsarc.cds.unistra.fr/cat/?IX/30
2RXF: https://heasarc.gsfc.nasa.gov/W3Browse/rosat/rospspcftot.html
1WGA: White, Giommi & Angelini, https://heasarc.gsfc.nasa.gov/wgacat/
CXOG: Chandra ACIS source catalog, Wang S. et al, 2016,ApJS,224,40
CXO: Chandra Source Catalog v1.1, https://asc.harvard.edu/csc1/
2CXO: Chandra Source Catalog v2.0, https://asc.harvard.edu/csc2/
CXOX: XAssist Chandra, https://asd.gsfc.nasa.gov/xassist/pipeline4/chandra/
2XMM/2XMMi: XMM-Newton DR3, https://cdsarc.cds.unistra.fr/cat/?IX/41
4XMM: XMM-Newton DR13, https://www.cosmos.esa.int/web/xmm-newton/xsa
XMMSL: XMM-Newton Slew Survey Release 2.0, same attribution as 4XMM
XMMX: XAssist XMM-Newton, https://asd.gsfc.nasa.gov/xassist/pipeline5/xmm/
LSXPS: Swift X-ray Point Sources, https://www.swift.ac.uk/LSXPS (01July23)
RASS (ROSAT All-Sky Survey) is not included as its low resolution is not
usable in isolation. Optical field solutions are calculated from the raw
source positions of all these catalogs (except 2CXO) as described in my
MORX v1 paper 2016,PASA,33,52.
This research has made use of the SIMBAD database and CDS cross-match service to
obtain Gaia-EDR3 and Pan-STARRS photometry provided by CDS, Strasbourg, France.