The Gaia-WISE Extragalactic Astrometric Catalog
aa r X i v : . [ a s t r o - ph . GA ] A p r Draft version April 9, 2018
Typeset using L A TEX twocolumn style in AASTeX61
THE GAIA-WISE EXTRAGALACTIC ASTROMETRIC CATALOG
Jennie Paine, Jeremy Darling, and Alexandra Truebenbach Center for Astrophysics and Space AstronomyDepartment of Astrophysical and Planetary SciencesUniversity of Colorado, 389 UCBBoulder, CO 80309-0389, USA
Submitted to ApJABSTRACTThe
Gaia mission has detected a large number of active galactic nuclei (AGN) and galaxies, but these objects mustbe identified among the 1000-fold more numerous stars. Extant astrometric AGN catalogs do not have the uniform skycoverage required to detect and characterize the all-sky low-multipole proper motion signals produced by the barycentermotion, gravitational waves, and cosmological effects. To remedy this, we present an all-sky sample of 567,721 AGNin
Gaia
Data Release 1, selected using WISE two-color criteria. The catalog has fairly uniform sky coverage beyondthe Galactic plane, with a mean density of 12.8 AGN per square degree. The objects have magnitudes ranging from G = 8 . Gaia’s magnitude limit, G = 20 .
7. The catalog is approximately 50% complete but suffers fromlow stellar contamination, roughly 0.2%. We predict that the end-of-mission
Gaia proper motions for this catalog willenable detection of the secular aberration drift to high significance (23 σ ) and will limit the anisotropy of the Hubbleexpansion to about 2%. Keywords: astrometry — catalogs — galaxies: active — infrared: galaxies — proper motions —quasars: general
Corresponding author: Jennie [email protected]
Paine, Darling, & Truebenbach INTRODUCTIONThe
Gaia mission will provide astrometric and propermotion measurements for a large number of bright ac-tive galactic nuclei (AGN), but separating the ∼ extragalactic objects from the ∼ stars remainschallenging (Gaia Collaboration et al. 2016). Currentcatalogs include the Large Quasar Astrometric Cata-log (LQAC; Souchay et al. 2015), the V´eron Catalogof Quasars and AGN (V´eron-Cetty & V´eron 2010), theSecrest et al. (2015) catalog of mid infrared AGN, andthe Gaia
Universe Model Snapshot (GUMS), a simu-lated catalog (Robin et al. 2012). Many of these cat-alogs are dominated by the Sloan Digital Sky Survey(SDSS) footprint that covers 35% of the sky (Ahn et al.2012), which is problematic for all-sky proper motionstudies that attempt to detect low-multipole correlatedproper motion signals such as the secular aberrationdrift dipole (Titov & Lambert 2013; Xu et al. 2012;Truebenbach & Darling 2017), the stochastic gravita-tional wave background quadrupole (Gwinn et al. 1997;Titov et al. 2011; Book & Flanagan 2011; Darling et al.2017), or the isotropy of the Hubble expansion (Darling2014; Chang & Lin 2015; Bengaly 2016).Desirable features of extragalactic proper motion cat-alogs are all-sky, uniform selection, and low stellar con-tamination. Completeness is not very important: it im-pacts the signal-to-noise of correlated global proper mo-tions, which scales with the square root of the numberof objects. In this work, we consider only low-multipoleproper motion signals, but completeness will ultimatelydetermine the maximum multipole that can be studieddue to the limiting sky density of sources. Stellar con-tamination is the largest concern for detecting globalsignals of a few µ arcsec yr − because stellar proper mo-tions can be large and significant and therefore dominatethe individually non-significant extragalactic proper mo-tions. What stellar contamination remains in any givenextragalactic catalog may be addressed using a non-Gaussian permissive likelihood function as described inDarling et al. (2017).This paper presents the Gaia -WISE extragalactic as-trometric catalog, a catalog designed to have low stellarcontamination and fairly uniform sky coverage outside ofthe Galactic Plane. Section 2 presents the WISE color-color selection used to identify AGN and exclude stars,and Section 3 explores the sky distribution of the cata-log, its optical and mid-IR properties, its redshift distri-bution, and the expected end-of-mission proper motionuncertainties. Section 4 predicts the performance of thiscatalog in detecting the secular aberration drift causedby the barycenter acceleration about the Galactic Cen-ter. Section 4 also predicts the expected
Gaia sensitivityto anisotropy in the Hubble expansion. We discuss theramifications of this work and the future prospects forextragalactic proper motion studies in Sections 5 and 6.We assume a Hubble constant of H = 72 km s − Mpc − and a flat cosmology (other cosmological assumptionsare not required). CATALOG SELECTION METHODThe WISE survey is an all-sky mid-infrared (MIR)survey in the 3.4, 4.6, 12 and 22 µ m bandpasses(W1, W2, W3, and W4, respectively; Wright et al.2010). The AllWISE data release, used in this work,combines data from the cryogenic and post-cryogenic(Mainzer et al. 2011) survey phases, and provides bet-ter sensitivity and accuracy over previous WISE datareleases. WISE colors have been shown to cleanly sep-arate AGN from stars and normal galaxies, and sev-eral methods exist in the literature for selecting AGNwith WISE (e.g. Assef et al. 2013; Mateos et al. 2012;Stern et al. 2005, 2012; Truebenbach & Darling 2017).To create our catalog of
Gaia
AGN, we did not considerselection methods using only a W1-W2 color cut in or-der to avoid contamination from brown dwarfs at lowGalactic latitudes, which can reside in the color spaceselected by single color cuts (Kirkpatrick et al. 2011).We employed the ALLWISE catalog of MIR AGN de-scribed in Secrest et al. (2015). The catalog is based onthe WISE two color selection technique of Mateos et al.(2012) which has cuts in the W1 − W2 and W2 − W3color space, referred to as the color wedge. This AGNcolor wedge was defined based on the Bright Ultrahard
XMM-Newton survey (BUXS), one of the largest flux-limited samples of ‘ultrahard’ X-ray-selected AGN, butthe method does not employ X-ray selection directly.BUXS is comprised of 258 objects, of which 56.2% aretype 1 AGN and nearly the rest are type 2. BUXStype 2 AGN are intrinsically less luminous than type1 AGN. Since the completeness of the MIR wedge hasa strong dependence on luminosity, the wedge prefer-entially selects type 1 AGN. Secrest et al. (2015) se-lected 1.4 million MIR AGN using ALLWISE profilefitting magnitudes with S/N ≥ cc flags = “0000” to avoid sources contaminated byimage artifacts.We cross-matched the Secrest et al. (2015) cata-log of MIR AGN with Gaia
Data Release 1 using allwise best neighbour , the precomputed WISE cross-match table provided in the
Gaia archive (Marrese et al.2017). The table includes only the most likely matchesbetween the WISE and
Gaia catalogs, called “bestneighbours.” Since
Gaia is used as the leading catalog incross-matching, a
Gaia source may be matched to mul-tiple sources from an external catalog. Marrese et al.(2017) then determine the best match to the
Gaia sourceusing the angular distance, position errors, epoch dif-ference, and density of sources in the external cata-log. A small number of
Gaia sources have
G >
Gaia’s nominal magnitude limit of 20.7,which are likely incorrectly determined magnitudes aia-WISE Extragalactic Astrometric Catalog
Gaia
MIR AGNcontains 567,721 objects. The first ten objects are givenin Appendix A, and the full catalog is available online.2.1.
Completeness
The completeness of the WISE color wedge selec-tion is dependent on the ratio of the AGN luminos-ity to the host luminosity because host galaxy lightcan contaminate the MIR emission (Mateos et al. 2012;Padovani et al. 2017). Thus, lower luminosity AGN willhave colors of normal galaxies and will be excluded bythe color wedge. To assess the completeness of our cata-log, we compared the catalog to the sample of SDSS DR9QSOs (Ahn et al. 2012) in
Gaia . SDSS QSOs were iden-tified in the
Gaia source catalog via the cross-matchingalgorithm provided in the
Gaia archive with a matchingradius of 1 arcsecond. 44.6% of all
Gaia -SDSS QSOswere also identified by the WISE color wedge, suggest-ing that our sample is missing more than half of all AGNin the
Gaia catalog. Only 49.3% of
Gaia -SDSS QSOshave S/N > Gaia -WISE catalog is thereforedue to non-detections in the least-sensitive WISE W3band. Among the WISE-detected Gaia-SDSS QSOs,90.2% lie in the WISE MIR color wedge. The remainingquasars generally have bluer W1 − W2 colors than thecolor wedge, likely due to contamination by host galaxystarlight. 2.2.
Stellar Contamination
Mateos et al. (2012) find that contamination by nor-mal galaxies in the MIR wedge is minimal. For astro-metric purposes, however, objects need only be extra-galactic, so unresolved galaxies are acceptable. Con-tamination by Galactic stars is of much greater concerndue to their large proper motions.To assess any remaining stellar contamination afteromitting the Tycho stars, we cross matched our sam-ple with the SDSS DR12 catalog (Alam et al. 2015).229,073 AGN in our sample reside within the SDSS foot-print, and 65,575 have a spectroscopic classification fromSDSS. Of those, only 104 objects (0.16%) are identifiedby their spectroscopic classification as stars. Extrapo-lating to the whole sky gives approximately 910 totalstars in our sample, suggesting negligible contamina-tion from stars. We also consider contamination fromdusty stars that would not be found in our SDSS cross-match. Nikutta et al. (2014) find that a majority of ob-jects brighter than W1= 11 are Galactic stars. Oursample contains 1,836 objects with W1 <
11, which in- dicates a maximum of 0.32% contamination from dustystars. RESULTS3.1.
Sky Distribution
Figure 1 illustrates the distribution of
Gaia -WISEAGN on the sky. The lower density of AGN at lowGalactic latitudes is due to a combination of dust alongthe Galactic plane and the effectiveness of the MIR colorwedge at excluding stars. Additionally, WISE photom-etry is limited by confusion near the Galactic plane dueto high source density (Wright et al. 2010). The higherdensities near the ecliptic poles are due to increased cov-erage by both WISE and
Gaia . The mean and mediandensities above the Galactic plane ( b > ◦ ) are 12.8and 12.0 objects per deg , respectively, and the maxi-mum density is 55 objects per deg .3.2. Optical PropertiesGaia surveys the sky down to G = 20 .
7, with a smallfraction of objects at
G >
21 (Gaia Collaboration et al.2016). As illustrated in Figure 2, the majority of WISEAGN lie at the fainter end of
Gaia’s magnitude distri-bution. Statistics for the distribution of G magnitudesare listed in Table 1.3.3. Mid-IR Properties
The WISE two color distribution for our catalog isshown in Figure 3, along with the Mateos et al. (2012)wedge. The majority of objects reside in a locus nearthe bluer end of the color wedge, with a small number ofoutliers with redder colors. The distribution around thelocus tapers before the color cuts, suggesting that thecolor wedge captures most of the AGN population, ex-cept for the bottom right cut where AGN colors begin tooverlap with the color space occupied by normal galax-ies. The distributions of WISE W1, W2, and W3 mag-nitudes, and W1 − W2 and W2 − W3 colors are shown inFigure 4; statistics for these distributions are given inTable 1. 3.4.
Redshifts
Redshifts were obtained for objects with spectroscopicredshifts from SDSS. Redshifts with nonzero warningflags or negative errors were discarded, since a negativeredshift error indicates a poor fit even if the warningflag is zero. This yielded redshifts for 90,365 objects( ∼ z = 4, which isunexpectedly high considering Gaia’s magnitude limit.However, a majority of these are confirmed quasarsin the SDSS Baryon Oscillation Spectroscopic Survey(BOSS) quasar catalog, of which many were selected forthe survey using WISE colors (Pˆaris et al. 2017).
Paine, Darling, & Truebenbach
Table 1.
Catalog Statistics G W1 W2 W3 W1 − W2 W2 − W3 Redshift σ µ,RA a σ µ,Dec a (mag) (mag) (mag) (mag) (mag) (mag) ( µ as yr − ) ( µ as yr − )Mean 19.3 15.2 14.0 10.9 1.2 3.0 1.3 236 218Median 19.4 15.3 14.1 11.1 1.2 3.0 1.2 205 191Minimum 8.8 4.8 3.7 0.2 0.5 2.0 0.0 2 3Maximum 21.0 18.8 17.1 12.9 2.2 5.8 7.0 1062 797 a Gaia expected end-of-mission proper motion uncertainty (see Section 3.5). -180° -120° -60° -0° 60° 120°-90°-60°-30°-0°30° 60° 01020304050 C a t a l o g S o u r c e D e n s i t y ( d e g − ) Figure 1.
Gaia -WISE extragalactic astrometric catalog density plot in Galactic coordinates. The colorbar indicates the numberof objects per deg . Proper Motion UncertaintiesGaia
DR2 will include positions, proper motions, andparallaxes — or limits on these quantities — for all ob-jects. Predicted proper motion standard errors can becalculated ahead of the release using Gaia performancecharacteristics. The PyGaia Python toolkit is an imple-mentation of Gaia performance models that can be usedfor basic simulation and analysis of Gaia data, includingcalculation of proper motion uncertainties. We utilizedthe PyGaia Python toolkit to calculate predicted propermotion uncertainties for each AGN, shown in Figure 6.This calculation relies on each object’s G magnitude, V − I C color, and ecliptic latitude. For objects wherethe V − I C color was not available, this value was setto zero, which has a negligible impact on the predictedproper motion uncertainty. The reported uncertaintiesinclude known instrumental effects. Statistics for thedistributions of predicted uncertainties are given in Ta- ble 1. The uncertainties in right ascension proper mo-tion are generally larger than in declination, which is aconsequence of the Gaia’s scanning law. APPLICATIONSAlthough proper motions for
Gaia
AGN will not beavailable until DR2, we can use the predicted uncer-tainties to test
Gaia’s potential capability to detect orconstrain select proper motion signals. For this purpose,we generate a null proper motion catalog by randomlyselecting proper motions consistent with zero based oneach object’s expected errors and assuming Gaussian-distributed errors. One can then add proper motionsignals to the noisy null catalog to study the expectedsensitivity of the
Gaia -WISE catalog to various corre-lated proper motions. These include the secular aberra-tion drift (Section 4.1), an anisotropic Hubble expansion(Section 4.2), and a stochastic long-period gravitationalwave background (Darling et al. 2017).4.1.
Secular Aberration Drift aia-WISE Extragalactic Astrometric Catalog G (mag) N Figure 2.
Distribution of
Gaia
G-band magnitudes in the
Gaia -WISE extragalactic astrometric catalog. The greendotted line indicates
Gaia’s nominal magnitude limit, G =20 . : − : P D J : − : P D J Figure 3.
WISE colors for
Gaia
MIR AGN. The dashedlines indicate the color wedge of Mateos et al. (2012). Thecolor bar indicates the logarithm of the number of objectsper hexagonal bin.
The aberration of light is an apparent angular deflec-tion of light rays caused by an observer’s velocity acrossthe rays and the finite speed of light. Aberration canbe caused by the Earth’s annual motion or the secularSolar motion in the Galaxy or with respect to the cos-mic microwave background rest frame. If the observerexperiences a constant acceleration then the aberrationwill exhibit a secular drift that manifests as an apparentproper motion of objects in a dipole pattern convergingtoward the acceleration vector direction.The secular aberration drift caused by the solar sys-tem’s acceleration toward the Galactic Center (a con-sequence of its orbit) is detectable in extragalacticproper motions as a dipole vector field that resem-bles an electric field and converges on the GalacticCenter (e.g. Xu et al. 2012; Titov & Lambert 2013;Truebenbach & Darling 2017). The expected solar ac-celeration and corresponding secular aberration driftdipole amplitude can be predicted using the distance tothe Galactic center ( R ) and the orbital speed of theSun (Θ + V ⊙ ), which includes solar motion V ⊙ in the di-rection of Galactic rotation Θ : a = (Θ + V ⊙ ) /R and | ~µ | = a/c . Reid et al. (2014) measured R = 8 . ± . + V ⊙ = 255 . ± . − from the trigono-metric parallaxes and proper motions of masers associ-ated with young massive stars. These yield an accel-eration of a = 0 . ± .
04 cm s − yr − and a dipoleamplitude of | ~µ | = 5 . ± . µ as yr − .An E-mode vector field dipole painted on the sky, ˜V E ( α, δ ), can be expressed as a ℓ = 1 vector sphericalharmonic following the notation of Mignard & Klioner(2012): ˜V E ( α, δ ) = s Re r π sin α + s Im r π cos α ! ˆe α + s r π cos δ + s Re r π cos α sin δ − s Im r π sin α sin δ ! ˆe δ where the coefficients s Re,Imℓm determine the directionand amplitude of the dipole, α and δ are the right ascen-sion and declination coordinates, and ˆe α and ˆe δ are theunit vectors in those directions. In this formalism, theexpected E-mode dipole caused by the solar orbit aboutthe Galactic Center (266 . ◦ , − . ◦ ) is ( s , s Re , s Im ) =( − . ± . , . ± . , − . ± . µ as yr − .In order to predict the Gaia sensitivity to the secularaberration drift signal, we assigned a proper motion toeach object that is consistent with no proper motion byrandomly sampling its predicted Gaussian proper mo-tion error distribution (Section 3.5). Over 1000 randomtrials, we added the expected secular aberration driftsignal to the noisy null proper motions, omitting the un-certainties in the input dipole, and used a least squares
Paine, Darling, & Truebenbach
W1 (mag) N W2 (mag) W3 (mag) W1−W2 (mag) N W2−W3 (mag) Figure 4.
Distribution of W1, W2, and W3 band magnitudes, and W1 − W2 and W2 − W3 colors in the
Gaia -WISE extragalacticastrometric catalog. Green dotted lines show the nominal S/N = 5 magnitudes for each band (16.9, 16.0, and 11.5 for W1, W2,and W3, respectively). 5 H G V K L I W 1 Figure 5.
Distribution of redshifts in the
Gaia -WISE extra-galactic astrometric catalog, where available (Section 3.4). minimization to fit a dipole to the data. The result-ing mean of the best fit parameters is ( s , s Re , s Im ) =( − . ± . , . ± . , − . ± . µ as yr − , con- Figure 6.
Predicted proper motion uncertainties in bothright ascension (blue) and declination (pink), with overlap-ping values shown in magenta. sistent with the original input dipole, with mean Z-scoreof 23. We therefore predict that
Gaia will produce aia-WISE Extragalactic Astrometric Catalog
Anisotropic Cosmic Expansion
Extragalactic proper motions can test the isotropy ofthe Hubble expansion in the current epoch. If we ne-glect the peculiar motions of galaxies caused by den-sity inhomogeneities, an isotropic Hubble expansion pro-duces no extragalactic proper motions. In contrast,anisotropic expansion will cause extragalactic objects tostream toward directions of faster expansion and awayfrom directions with slower expansion. All-sky propermotion observations can therefore measure the expan-sion isotropy and constrain cosmological models that at-tempt to explain accelerating expansion without invok-ing dark energy, such as Lemaitre-Tolman-Bondi modelsand Bianchi universes (e.g. Amendola et al. 2013).Quercellini et al. (2009) and Fontanini et al. (2009)showed that a triaxial expansion can be described us-ing a Bianchi I model, which has the metric ds = − dt + a ( t ) dx + b ( t ) dy + c ( t ) dz . (1)This metric permits different expansion rates along thethree axes: H x = ˙ a/a , H y = ˙ b/b , and H z = ˙ c/c .The observed Hubble parameter would be H = ddt ( abc ) / / ( abc ) / , and the Friedmann-Robertson-Walker metric is recovered for a ( t ) = b ( t ) = c ( t ). Theexpansion can therefore be characterized by the frac-tional departure from the isotropic Hubble expansionalong the coordinate i using a unitless shear parameter:Σ i = H i, H − . (2)The principal shearing axes can be arbitrarily orientedon the sky, and Darling (2014) showed that the propermotion induced by this anisotropy model can be com-pletely described by a quadrupolar E-mode vector field.To test the catalog’s potential to constrain anisotropy,we performed 1,000 trials of adding a randomly gener-ated anisotropy signal to the noisy null proper motionsand fitting the anisotropy model to attempt to recreatethe original input signal. We used the shear equation(Equation A1) of Darling (2014) to form these artificialanisotropy signals. For each trial, shear terms Σ x , Σ y ,and Σ z were drawn from Gaussian distributions withmean of zero and random standard deviation sampledfrom a uniform distribution between 0 and 0.1. Therotation angles were randomly selected from a uniformdistribution between 0 and 2 π , assuming that there isno preferred direction for anisotropy. After the signal isadded to the null proper motions, we use a least squaresminimization to fit the shear equation to the data in anattempt to recover the original signal.The shear equation parameters are degenerate due tothe rotation degeneracy of the principal axes (no partic-ular axis is required to be the direction of maximum or í í í 0 D [ L P X P , Q S X W 6 K H D U í í 0 D [ L P X P ) L W 6 K H D U Figure 7.
Maximum absolute value of the fit shear vs. theinput shear for Hubble expansion anisotropies added to thesynthetic
Gaia -WISE AGN catalog proper motions. Non-significant fits are displayed as upper limits. minimum expansion), and therefore individual fit pa-rameters do not necessarily match the original inputparameters. Instead, we compare the maximum inputshear to the maximum fit shear, as shown in Figure 7.There is a roughly one-to-one correlation for large in-put values; however, for maximum input shear below ∼ × − , noise dominates and the fit parameters tendtoward a noise floor of 0.018 (a 1.8% departure fromanisotropy). The fit, however, is not significant for suchlow input anisotropy. DISCUSSIONPrior to the first
Gaia data release, the
Gaia
Universemodel snapshot (GUMS) simulated a synthetic catalogof objects that
Gaia could have potentially observed(Robin et al. 2012). GUMS simulated that nearly onemillion quasars would be observed by
Gaia . Our sampleroughly agrees with that number, given that it is about50% incomplete. However, unlike GUMS, our sampleconsists of real objects actually detected by
Gaia .The Large Quasar Astrometric Catalog (LQAC3;Souchay et al. 2015), is a collection of 321,957 objectsand represents the complete set of already identifiedquasars as of 2015. While the LQAC3 reliably containsextragalactic objects, the LQAC3-
Gaia cross-match isdominated by the SDSS footprint. Our catalog has amore uniform sky distribution, and is therefore prefer-able for the study of low-multipole proper motion sig-nals.We expect
Gaia -WISE AGN to be able to measure thesecular aberration drift with 23 σ significance. Mignard Paine, Darling, & Truebenbach (2012) predicted that
Gaia would detect the secularaberration drift with about 10 σ accuracy, assuming 10 – 10 quasars observed by Gaia with proper motionerrors lower than predicted here. Titov et al. (2011)predicted
Gaia to measure the dipole parameters withabout 10% relative precision. We find that the catalogshould be able to measure the dipole parameters withhigher precision, with the exception of the s Re compo-nent.While isotropy is a fundamental pillar of cosmologyand is well constrained by the cosmic microwave back-ground (Planck Collaboration et al. 2016), Gaia -WISEAGN will be able to probe the isotropy of expansionfor the relatively local universe since the majority areat redshift below 2.5 (95th percentile value). We pre-dict that
Gaia -WISE AGN will constrain the anisotropyof the Hubble expansion to about 2%. Darling (2014)showed that the expansion is isotropic to within 7% inthe most constrained direction using a catalog of 429 ra-dio sources. Local anisotropy has been previously mea-sured using the Hubble parameters derived from Type1a supernovae. Chang & Lin (2015) find that the max-imum anisotropy of the Hubble parameter is 3% ± z < . . ± .
86) km s − Mpc − for z < .
1, which corresponds to a maximum departurefrom isotropy of 3 . ± . Gaia isotropy mea-surement will therefore be competitive with and orthog-onal to other more traditional methods.Our analysis of the astrometric signals that may bedetected using
Gaia -WISE AGN has assumed that theproper motions of all objects will be determined withthe same precision as point sources. In reality, somegalaxies may appear extended to
Gaia , in which casethe precision of the image centroid position will be di-minished. The intrinsic variability of AGN will be an ad-ditional proper motion noise source, since variable AGNflux can cause the image centroid to move by up to a fewmas for nearby AGN (Popovi´c et al. 2012). Microlens-ing of quasars may also cause the image centroid to shiftdue to the appearance or disappearance of microimages(Williams & Saha 1995; Lewis & Ibata 1998). The ef-fect on the centroid position may be as large as tensof µ as due to stellar mass objects in the lensing galaxy(Treyer & Wambsganss 2004) or a few mas due to stel-lar clusters (Popovi´c & Simi´c 2013). The effects of bothAGN variability and microlensing will add uncorrelatednoise to the proper motions. They will therefore be aver-aged out in the determination of correlated signals suchas the secular aberration drift and anisotropic expan-sion, despite adding to the overall noise in the signals. CONCLUSIONSWe presented a catalog of
Gaia
AGN selected usingthe WISE two color method of Mateos et al. (2012).The catalog contains 567,721 objects, and we estimate that this sample is roughly 50% complete. We find thatthe WISE wedge reliably selects extragalactic objects,with only a negligible portion (0.2%) of our sample likelycontaminated by stars. We demonstrated two potentialapplications of the catalog, a precise measurement ofthe secular aberration drift and strong constraints onthe isotropy of the Hubble expansion. Based on theexpected end-of-mission proper motion uncertainty foreach object in the
Gaia -WISE catalog, we predict a mea-surement of the secular aberration drift with ∼ σ sig-nificance and a limit on the anisotropy of the Hubbleflow of ∼ Gaia ( ),processed by the Gaia
Data Processing and AnalysisConsortium (DPAC, ).Funding for the DPAC has been provided by nationalinstitutions, in particular the institutions participatingin the
Gaia
Multilateral Agreement.This publication makes use of data products from theWide-field Infrared Survey Explorer, which is a jointproject of the University of California, Los Angeles, andthe Jet Propulsion Laboratory/California Institute ofTechnology, funded by the National Aeronautics andSpace Administration.Funding for SDSS-III has been provided by the AlfredP. Sloan Foundation, the Participating Institutions, theNational Science Foundation, and the U.S. Departmentof Energy Office of Science. The SDSS-III web site is . SDSS-III is managed by theAstrophysical Research Consortium for the Participat-ing Institutions of the SDSS-III Collaboration includ-ing the University of Arizona, the Brazilian Participa-tion Group, Brookhaven National Laboratory, CarnegieMellon University, University of Florida, the FrenchParticipation Group, the German Participation Group,Harvard University, the Instituto de Astrofisica de Ca-narias, the Michigan State/Notre Dame/JINA Partic-ipation Group, Johns Hopkins University, LawrenceBerkeley National Laboratory, Max Planck Institute forAstrophysics, Max Planck Institute for ExtraterrestrialPhysics, New Mexico State University, New York Uni-versity, Ohio State University, Pennsylvania State Uni-versity, University of Portsmouth, Princeton University,the Spanish Participation Group, University of Tokyo,University of Utah, Vanderbilt University, University ofVirginia, University of Washington, and Yale University.This research has made use of the NASA/IPAC Extra-galactic Database (NED) which is operated by the JetPropulsion Laboratory, California Institute of Technol- aia-WISE Extragalactic Astrometric Catalog
Software: astropy (Astropy Collaboration et al.2013), pyGaia, STILTS (Taylor 2006), TOPCAT(Taylor 2005)REFERENCES
Ahn, C. P., Alexandroff, R., Allende Prieto, C., et al. 2012,ApJS, 203, 21Alam, S., Albareti, F. D., Allende Prieto, C., et al. 2015,ApJS, 219, 12Amendola, L., Eggers Bjæ lde, O., Valkenburg, W., &Wong, Y. Y. Y. 2013, JCAP, 12, 042Assef, R. J., Stern, D., Kochanek, C. S., et al. 2013, ApJ,772, 26Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., etal. 2013, A&A, 558, A33Bengaly, C. A. P., Jr. 2016, JCAP, 4, 036Book, L. G. & Flanagan, ´E. ´E. 2011, Phys. Rev. D, 83,024024Chang, Z., & Lin, H.-N. 2015, MNRAS, 446, 2952Darling, J. 2014, MNRAS, 442, L66Darling, J., Truebenbach, A. E., & Paine, J. 2017, ApJ,submittedFontanini, M., West, E. J., & Trodden, M. 2009, Phys. Rev.D, 80, 123515Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al.2016, A&A, 595, A2Gwinn, C. R., Eubanks, T. M., Pyne, T., Birkinshaw, M.,& Matsakis, D. N. 1997, ApJ, 485, 87Kirkpatrick, J. D., Cushing, M. C., Gelino, C. R., et al.2011, ApJS, 197, 19Lewis, G. F., & Ibata, R. A. 1998, ApJ, 501, 478Mainzer, A., Bauer, J., Grav, T., et al. 2011, ApJ, 731, 53Marrese, P.M., Marinoni, S., Giuffrida, G., and Fabrizio, M.2017,A&A, submittedMateos, S., Alonso-Herrero, A., Carrera, F. J., et al. 2012,MNRAS, 426, 3271Mignard, F. 2012, Mem. Soc. Astron. Italiana, 83, 918Mignard, F., & Klioner, S. 2012, A&A, 547, A59Nikutta, R., Hunt-Walker, N., Nenkova, M., Ivezi´c, ˇZ., &Elitzur, M. 2014, MNRAS, 442, 3361 Padovani, P., Alexander, D. M., Assef, R. J., et al. 2017,A&A Rv, 25, 2Pˆaris, I., Petitjean, P., Ross, N. P., et al. 2017, A&A, 597,A79Planck Collaboration, Ade, P. A. R., Aghanim, N., et al.2016, A&A, 594, A16Popovi´c, L. ˇC., Jovanovi´c, P., Stalevski, M., et al. 2012,A&A, 538, A107Popovi´c, L. ˇC., & Simi´c, S. 2013, MNRAS, 432, 848Quercellini, C., Cabella, P., Amendola, L., Quartin, M., &Balbi, A. 2009, Phys. Rev. D, 80, 063527Reid, M. J., Menten, K. M., Brunthaler, A., et al. 2014,ApJ, 783, 130Robin, A. C., Luri, X., Reyl´e, C., et al. 2012, A&A, 543,A100Secrest, N. J., Dudik, R. P., Dorland, B. N., et al. 2015,ApJS, 221, 12Souchay, J., Andrei, A. H., Barache, C., et al. 2015, A&A,583, A75Stern, D., Eisenhardt, P., Gorjian, V., et al. 2005, ApJ, 631,163Stern, D., Assef, R. J., Benford, D. J., et al. 2012, ApJ,753, 30Taylor, M. B. 2005, Astronomical Data Analysis Softwareand Systems XIV, 347, 29Taylor, M. B. 2006, Astronomical Data Analysis Softwareand Systems XV, 351, 666Titov, O., Lambert, S. B., & Gontier, A.-M. 2011, A&A,529, A91Titov, O., & Lambert, S. 2013, A&A, 559, A95Treyer, M., & Wambsganss, J. 2004, A&A, 416, 19Truebenbach, A. E., & Darling, J. 2017, MNRAS, 468, 196Truebenbach, A. E., & Darling, J. 2017, ApJ, submittedV´eron-Cetty, M.-P., & V´eron, P. 2010, A&A, 518, A10Williams, L. L. R., & Saha, P. 1995, AJ, 110, 1471Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al.2010, AJ, 140, 1868-1881Xu, M. H., Wang, G. L., & Zhao, M. 2012, A&A, 544, A135 Paine, Darling, & Truebenbach
APPENDIX A. CATALOGTable 2 lists the first ten rows of the
Gaia -WISE extragalactic catalog. The full catalog containing 567,721 objectswill be available as a machine-readable table provided by the publisher. a i a - W I S EE x t r a g a l a ct i c A s t r o m e t r i c C a t a l o g Table 2.
Gaia -WISE Extragalactic Catalog
Gaia
ID RA σ RA Dec σ Dec G ALLWISE