The discovery of the first luminous z~6 quasar in the UKIDSS Large Area Survey

0.85 MB·English

by  B. P. Venemans

Mon.Not.R.Astron.Soc.000,1–5(2006) Printed5February2008 (MNLATEXstylefilev2.2)
z ∼ 6
The discovery of the first luminous quasar in the
UKIDSS Large Area Survey
B. P. Venemans1 ⋆, R. G. McMahon1, S. J. Warren2, E. A. Gonzalez-Solares1,
P. C. Hewett1, D. J. Mortlock2, S. Dye3, R. G. Sharp4
7
0 1 Institute of Astronomy, Universityof Cambridge, Madingley Road, Cambridge, CB3 0HA, United Kingdom
0 2 Astrophysics Group, Imperial College London, Blackett Laboratory, Prince Consort Road, London, SW7 2AZ, United Kingdom
2 3 Cardiff University,School of Physics & Astronomy, QueensBuildings, The Parade, Cardiff, CF24 3AA, United Kingdom
4 Anglo-Australian Observatory, P.O. Box 296, Epping, NSW1710, Australia
n
a
J

2
Accepted 2007January12.Received2007January10;inoriginalform2006December6
1

2
v ABSTRACT
2 We present the initial results from our search for high redshift, z & 6, quasars using
6 near infrared data from the UKIRT Infrared Deep Sky Survey (UKIDSS) Large Area
1 Survey(LAS).Ouranalysisof106deg2 ofskyfromDataRelease1(DR1)hasresulted
2 in the discovery of ULAS J020332.38+001229.2, a luminous (J = 20.0, J =
AB Vega
1 19.1, M = −26.2) quasar at z = 5.86. Following candidate selection from the
1450
6 combined IR and optical catalogue data and stacking of multiple epoch Sloan Digital
0
Sky Survey (SDSS) data, we have obtained optical spectroscopy for the only two
h/ z & 6 quasar candidates. The VLT FORS2 spectrum of ULAS J020332.38+001229.2
p shows broad Lyα + NV 1240 emission at ∼8350˚A and an abrupt continuum break
- due to absorption by the Lyα forest. The quasar is not present in the SDSS DR5
o
r catalogue and the continuum spectral index of α=-1.4 (Fν ∝ να) is redder than a
t composite of SDSS quasars at similar redshifts (α=-0.5). The discovery of one z ∼ 6
s
a quasar in ∼100 deg2 in a complete sample within our selection criteria down to a
: mediandepthofY =20.4(7σ)isconsistentwithexistingSDSSresults.Wedescribe
v AB
our survey methodology, including the use of optical data from the SDSS and the
i
X highlyeffective proceduresdevelopedtoisolatethe verysmallsurfacedensityofhigh-
r probability quasar candidates.
a
Key words: cosmology: observations, galaxies: quasars: general, galaxies: quasars:
individual: ULAS J020332.38+001229.2
1 INTRODUCTION interveningmaterialandcanbeusedtodeterminethebary-
onic content and physical conditions (metallicity, tempera-
Quasars are powerful probes of the high-redshift Universe
ture, ionization state) of the intergalactic medium (IGM)
since, by virtue of their high luminosities, they can be ob-
and gas within interveninggalaxies.
servedouttoimmensedistanceswithlargelookbacktimes.
Finding quasars above redshift z > 6.0 is of crucial
Thehost galaxies ofthehighest redshift quasarsare proba-
importance since it is in the epoch of reionization between
blystillformingandmostlikelyoccurinoverdenseregionsin
theredshiftwheretheStr¨omgrenspheresofthereionization
thematterdistributionoftheearlyUniverse.Studiesoftheir
sources overlap (zOVL = 6.1±0.15, Gnedin & Fan 2006),
spacedensities,hostgalaxiesandlocalenvironmentsprovide
as shown by the observations of the Gunn-Peterson optical
potentiallypowerfulconstraintsontheoriesoftheformation
depth at z ∼ 6, and the lower bound (1σ) to the epoch of
andgrowthofsupermassiveblackholes,thegrowthoflarge
reionization from the 3-year WMAP polarization results at
scale structure and the formation and evolution of the first
z=8.2(Pageetal.2006) whichcorrespondsroughlytothe
galaxies (e.g. Kauffmann & Haehnelt 2000).
epoch when theUniverse was half-ionized (byvolume).
In addition to being of intrinsic interest, bright high-
Over the last few years there has been considerable
redshift quasars are particularly valuable as probes of the
progress in wide area searches for high-redshift quasars.
Universeviaabsorptionstudiesofcosmologicallydistributed
Storrie-Lombardi et al. (2001) reported the final results of
a large survey for z ∼ 4 quasars over ∼8000deg2 which re-
⋆ E-mail:[email protected] sulted in ∼50 quasars in the redshift range 4.0 < z < 4.8.
2 B. P. Venemans, et al.
This survey used photographic data and was based on the
dropacrossLyαcausedbytheLyαforest,depressingtheob-
servedfluxintheB-bandrelativetotheR-band.Theadvent
oftheCCD-basedSloanDigitalSkySurvey(SDSS,Yorket
al. 2000) extended wide-area surveys to longer wavelengths
with the inclusion of the i- and z-bands, enabling searches
forquasarsbasedoni−z colourandredshiftsofz ∼6were
reached (Fan et al. 2000). There are now around 20 z ∼ 6
quasarsknown,withthehighestredshiftquasaratz=6.42
(Fan et al. 2003, 2006).
However, finding quasars beyond z = 6.4 using the
SDSS is practically almost impossible because the sources
become too faint in the reddest passband (the z-band) due
toabsorptionbytheinterveningLyαforest.Toreachhigher
redshift requires wide-field near-infrared imaging to sub-
stantially fainter fluxes then the 2 Micron All Sky Survey
(2MASS,Skrutskieetal.2006).Thisistheparameterspace
explored by the UK Infrared Telescope (UKIRT) Infrared
DeepSkySurvey(UKIDSS)LargeAreaSurvey(LAS).One
of the main science drivers of the UKIDSS LAS is the dis-
covery of quasars with z>6.
In this Letter, we report the first high-redshift quasar Figure 1. i−Y vs Y −J diagram in the AB magnitude sys-
discovered with UKIDSS, ULAS J020332.38+001229.2 at
temillustratingsimulatedcoloursofstars,ellipticalgalaxiesand
z =5.86. All magnitudes are given in the AB system using quasars (from Hewett et al. 2006) and observed colours from
theVegatoABconversionsprovidedinHewettetal.(2006) UKIDSSandSDSS.Alsoplottedaredatafrom20deg2 fromthe
andaΛ-dominatedcosmology withH0 =70kms−1Mpc−1, UKIDSSEarlyDataRelease.Thecloudofbluepointsareobjects
ΩM =0.3, and ΩΛ =0.7 is adopted. classifiedasstars,redpointsareobjectsclassifiedasgalaxies.The
regionincolour-colourspaceoccupiedbyhigh-redshiftquasarsis
outlinedbythedashedlines.QuasarULASJ0203+0012isplotted
asopencirclewitherrorbar(SDSSDR5i-bandphotometry)and
2 SURVEY IMAGING DATA assolidcircle(co-added i-bandphotometry). Thesolidsquareis
theMdwarfULASJ022905.12+005506.5.
ThenearinfrareddataarefromtheUKIDSSLAS(Lawrence
et al. 2006) which is being carried out using the UKIRT
Wide Field Camera (WFCAM, Casali et al. 2006) on the
(September,October and November) in each of three years
3.8-m UKIRT.WFCAMhasasparsepackedmosaicoffour
(2005-2007) as part of the SDSS-II project. In addition
2048×2048pixelRockwellHawaii-IIarrayswithapixelscale
Stripe82wasimaged10-20timesduringtheSDSS-Iproject
of 0′.′4 pixel−1.Theinstantaneous fieldof viewis 0.21 deg2.
(see Adelman-McCarthy et al. 2006). The SDSS data is re-
The arrays are spaced by 0.94 times the detector width.
leased as DRSN1 (Sako et al. 2005) and DRsup (Adelman-
A mosaic of four pointings gives a tile with a continuous
McCarthyetal.2006).TheareacoveredbytheLASinboth
coveredareaof0.77deg2.SeeDyeetal.(2006)fortechnical
Y- and J-bands in the SDSS Southern Survey area is 106
details of UKIDSS. deg2. In this area ∼950000 objects are detected in both Y
The LAS consists of nominal 40-s exposures in the Y-,
and J with a signal-to-noise ratio in Y >7.
J-,H-andK-bands.TheY-band(0.97–1.07µm)isspecially
WematchedtheUKIDSSDR1datawiththecatalogues
designed to fall between the z-band and the J-band, occu-
from the SDSS DR5, using a search radius of 2arcsec. The
pying the clean wavelength range between the atmospheric
colours of objects are computed using the SDSS PSF mag-
absorption bands at 0.95µm and 1.14µm. The Y-band is
nitudes and the UKIDSS apermag3 magnitudes which are
needed to distinguish high-redshift quasars from the more
computed using a diameter of 2arcsec, corrected for aper-
numerouscoolstars,suchasL-andT-dwarfs(Hewettetal.
tureslosses on a pointing by pointing basis.
2006, see also Fig. 1).
In this Letter, we use data from the Data Release 1
(DR1,Warren et al. 2006). The UKIDSSobservations were
obtained during the period 2005 August to 2006 January.
3 QUASAR SELECTION
The data was released to theESO community in 2006 July
(see Warren et al. 2006 for further details). The median Candidatehigh-redshift(z>5.8) quasarsdowntoasignal-
5σ point source depths in AB magnitudes are YAB=20.8, to-noise ratio of ∼7 in Y, corresponding to a median depth
JAB=20.5, HAB=20.2 and KAB=20.1 (Warren et al. 2006). ofYAB=20.4,wereselectedonthebasisoftheirblueY −J
We have restricted our analysis to the UKIDSS LAS area UKIDSS colour and, either, (i) red i−Y colours from our
that is imaged at multiple epochs by the SDSS Southern combined UKIDSS-SDSS catalogue, or, (ii) absence in the
Survey (−25◦ < RA < 60◦;−1.25◦ < Dec < 1.27◦, called SDSS i- and/or z-band catalogue. The colour-colour dia-
Stripe 82 in the SDSS footprint, York et al. 2000). This re- gram shown in Fig. 1 demonstrates the principles behind
gion of sky is being imaged by the SDSS collaboration in ourselection criteria.Thecolourcutsusedwerei−Y >2.5
5 passbands (u, g, r, i, z) repeatedly during three months and Y −J <0.6. TheY −J cut is used toavoid theL and
Discovery of a luminous z = 5.9 quasar in the UKIDSS LAS 3
T dwarfs. The i−Y cut selects quasars above z =5.8 and Table 1. Photometry of the z = 5.86 quasar, ULAS
discriminates against foreground Galactic stars. J020332.38+001229.2. Magnitudes areintheABsystem.
An additional requirement for candidate quasars was
that they should be undetected in the SDSS u-, g- and r- Band Source mag[AB] Date
bands.WeregardedanobjectasundetectedinSDSS,ifthe
i SDSSDR5 >23.1 2002-09-05
SDSSPSF-magnitudewasfainterthanthe3σ limitingmag-
nitude of the SDSS 2048×1489 pixel (13.′5×9.′8) field in Co-add(DRSN1,DRsup) 23.7±0.2 2002–2005
z SDSSDR5 21.2±0.3 2002-09-05
whichtheobjectwasfound.Wemadeanempiricaldetermi- Co-add(DRSN1,DRsup) 20.9±0.1 2002–2005
nation of the 3σ magnitude limit for each SDSS field from zNTT NTT/EMMI 20.69±0.06 2006-07-18
theSDSScatalogue.Ifanobjectwasdetectedinthei-band, Y UKIDSS 20.40±0.12 2005-11-26
wealsorequiredthati−z >2.2followingthecriterionused J UKIDSS 19.98±0.10 2005-11-26
by Fan et al. (2003). K UKIDSS 19.21±0.08 2005-12-09
We performed various steps to remove false positives
from our candidate quasar sample. These steps included:
4 CANDIDATE FOLLOW-UP OBSERVATIONS
4.1 ESO NTT imaging observations
(i) Removal of intra-quadrant cross-talk images. Cross-
Candidateswereobservedwithfilter♯611intheESOMulti-
talk images are artifacts caused by bright stars (see Dye et
Mode Instrument (EMMI, Dekker, Delabre & Dodorico
al. 2006 for a detailed description). The cross-talk images
1986), at the 3.6-m New Technology Telescope (NTT)1,
appear offset from the bright star by multiples of 51.2 arc-
sec (or 128 pixels with scale 0′.′4) corresponding to a single for 900s on the night beginning 2006 August 18. The fil-
read out channel on the same quadrant. More than 99per ter/detector combination is referred to hereas zNTT and is
similar to the SDSS z-band. Using the measured CCD sen-
cent of the z-dropouts are cross-talk images. We used the
sitivitycurve,andthefiltertransmission curve,wefollowed
2MASScatalogtolocateallstarsinthesurveyedareadown
theprocedure used by Hewett et al. (2006) to establish the
ttoo rJeAmBov=e a1l5l.5c.anWdeidathteesntuhsaetdwtehreeploocsaittieodnswiotfhitnheaserasdtaiurss colourrelation zNTT =z−0.05(i−z)(in ABmagnitudes),
fordwarfstars,OtoM.PhotometryiniandzofSDSSDR5
of 5arcsec of multiples of 51.2 arcsec in the North/South
orEast/Westdirectionasappropriateforthequadrant(see sources in the field was converted to zNTT, and the bright-
Dye et al. 2006) out to ≃6 arcmin on the Y-band images ness of the source was measured via relative photometry
using a fixed aperture size. For a typical quasar at z = 5.9
from the 2MASS stars.
thedifference between zNTT and z is at most 0.05mag.
(ii) Independentrepeat photometry on theSDSS images
With the zNTT-band photometry eight of the 10 ob-
using the pixel data of the SDSS to measure the photom-
servedobjectsweretooblueini−z andtooredinz−J to
etry of candidate quasars in the g-, i- and z-bands. This
beconsideredaslikelyquasars(seeTableA.1forthenames
removedobjectsthatdonotappearintheSDSScatalogues
oftheserejectedcandidates).Suchobjectsappeartobethe
but are present on the images, such as in the case where
maincontaminantinoursearch.Duetophotometricerrors,
objectslienearperturbingbrighterobjectsandtheedgesof
the z-band magnitudes are scattered to a brighter value,
theSDSS data. This photometric analysis also provided 3σ
makingtheobjectsappearredderthani−z>2.2andblue
upperlimits.
in z−J.
(iii) Co-adding of up to seven epochs of SDSS data from
Thetworemainingcandidates,wherethezNTT magni-
DRsup and DRSN1 taken in photometric conditions to de-
tudeconfirmedtheco-addedSDSSz-bandmagnitude,were
tect objects that lie below theSDSSsingle-epoch catalogue
selected for spectroscopic follow-up.
limits in the optical i- and z-bands. The co-added images
wereobtained byaveraging multiplesingle-epoch SDSSim-
ages centered on thelocation of theUKIDSSobject, scaled
4.2 ESO VLT spectroscopic observations
to a common zeropoint using theDR5 catalogue.
(iv) Screening by eye to identify spurious candidates Spectra of the two candidates were obtained on 2006
which hadnot been removedat anyearlier stage (e.g. mov- September1and2usingtheFOcalReducer/lowdispersion
ingobjectssuchasasteroidsandcorruptedobjectsnearper- Spectrograph 2 (FORS2, Appenzeller et al. 1998) on the
turbingbrightobjects).UKIDSSobservationsineachwave- 8.2-m Very Large Telescope (VLT) Antu2. The candidates
band are not simultaneous and the time interval between were observed through the 600z holographic grism with a
consecutiveobservationsindifferentwavebandsisgenerally 1′.′0 wide long slit. Individual exposures were 600s dura-
intherange30to120mins.Asteroidscouldberecognizedas tion. The pixels were 2×2 binned to decrease the readout
theymovebetweentheobservationsinthedifferentUKIDSS time and noise, giving a spatial scale 0′.′25 pixel−1 and a
bands. An object that appears on more than one band and dispersion of 1.62˚Apixel−1. The nights were photometric
shiftsby&1arcsechr−1couldbeidentifiedasmoving,down with seeing around 0′.′5. Forthewavelength calibration, ex-
to theflux limit we considered. posures of He, HgCd and Ne arc lamps were obtained. The
1 Based on observations made with ESO Telescopes at the La
These steps reduced the number of candidates from
SillaObservatoryunderprogrammeID077.A-0807
>5000,themajorityofwhichwerecross-talkartifacts,down 2 Based on observations made with ESO Telescopes at the
to10,thefollow-upobservationsofwhicharedescribednext. ParanalObservatoryunderprogrammeID077.A-0310
4 B. P. Venemans, et al.
Figure 3. Finding chart for ULAS J0203+0012. The chart is
taken from the Y-band image and is 5′×5′ in size. The image
is boxcar averaged over 3×3 pixels. The quasar is marked by a
cross-hair.
Figure 2. Discovery spectrum of the z = 5.86 quasar ULAS
J0203+0012. Below the object a noise and a sky spectrum are
plotted, both in units of 10−17 ergs−1cm−2˚A−1. The spectra
areboxcaraveragedoverthreepixels.Thetwo-dimensionalspec-
trumisshownatthebottom.TheLyα(1216˚A)andNV(1240˚A)
emission lines are marked. The object shows the characteristic
continuumdecrementacrosstheLyαemissionline,withverylit-
tlefluxbluewardofthebreak.Around8500˚AaCIVλλ1548,1551
absorptiondoubletcanbeseeninthespectrum.The’absorption
lines’at8917˚A,9315˚Aand9518˚Aareskysubtractionartifacts,
andtheabsorptionnear9073 ˚Aiscausedbyabadpixel.
rms of the wavelength calibration was better than 0.2˚A.
Figure 4. Spectral energy distribution of ULAS J0203+0012
Thespectracoverthewavelengthrangeλλ7370–10200. The
(solid line) and a composite of three SDSS quasars with 5.8 <
spectrophotometric standard star GD50 (Oke 1990) was z<6.0.
observed with a 5arcsec slit for relative flux calibration,
and the absolute calibration was determined using the ob-
served Y-band flux. Total exposure times were 1200s for z=5.862.Therest-frameequivalentwidthoftheLyα+NV
ULASJ020332.38+001229.2 (hereafterULASJ0203+0012) emission line is 64˚A (uncorrected for Lyα forest absorp-
and 2400s for ULASJ022905.11+005506.5. tion), which is comparable to other z ∼6 quasars (e.g. Fan
et al. 2004). A finding chart of the quasar is shown in Fig.
3 and the magnitudes are given in Table 1. It should be
stressed that ULAS J0203+0012 is too faint in the optical
5 RESULTS
passbandstobepresentintheSDSSDR5catalogue.The3σ
Fig.2showsthespectrumofULASJ0203+0012. Thespec- z-bandmagnitudelistedinTable1isderivedusingaperture
trum is characterized by a broad, asymmetric emission photometry on theSDSS pixel image at the location of the
line and a continuum break around 8350˚A. We interpret UKIDSSsource.
these features as the Lyα emission line and the continuum The second candidate, ULAS J022905.12+005506.5,
break resulting from Lyα forest absorption at a redshift of turnedout to bea star, probably an M dwarf.
z≃5.86.TheNV1240emissionlineisblendedwiththeLyα
line.TheCIV1549emissionlinewouldlieat∼10630˚A,out-
side the wavelength range of the spectrum. We determined
6 DISCUSSION
the redshift of the quasar by fitting a Gaussian to the red
partoftheLyαline.Thisgivesaredshiftofz=5.864±0.003. In Fig. 4 we compare the optical and infrared magnitudes
The peak of the emission is at 8342˚A giving a redshift of of ULAS J0203+0012 with a composite of three 5.8 < z <
Discovery of a luminous z = 5.9 quasar in the UKIDSS LAS 5
6.0 SDSS quasars which were observed during the science
TableA1.Quasarcandidatesthatwererejectedafterthefollow-
verification of WFCAM. Both spectral energy distributions up observations. Using the zNTT-band photometry (Sect. 4.1)
(SEDs) are scaled to Fν = 1 at the centre of the Y-band theseeightobjectsaretooblueini−z (i−z<2.2)andtoored
at10305˚A.ULASJ0203+0012 issignificantlyredderinthe inz−J tobeconsideredaslikelyquasars.
infraredthantheSDSSquasarcomposite.Ifweapproximate
theSEDsintheinfraredbyapowerlawwithslopeα(Fν ∝ UKIDSSname
να)thentheslopesforULASJ0203+0012 areα=−1.44±
0.17 between the Y and K and α = −1.25±0.21 between UULLAASSJJ003003385522..5254+−000035333244..10
J and K. The SDSS quasars have average slopes of α = ULASJ022119.09−010905.9
−0.52±0.09 and α = −0.56±0.12 between Y and K and ULASJ223845.14+011132.8
J and K respectively. This is consistent with the average ULASJ021753.93+003906.9
slope of α=−0.57 thatwas found for asample of 45 SDSS ULASJ223457.11+005136.6
quasars with 3.6<z<5.0 byPentericci et al. (2003). ULASJ011650.20+011532.2
ULAS J0203+0012 is significantly fainter than the ULASJ021047.00−001419.6
SDSS quasars. The absolute magnitude at 1450˚A in the
rest-frame is M1450 = −26.2, nearly a magnitude fainter
than that of the SDSS composite (M1450 = −27.1). How- ACKNOWLEDGMENTS
ever,becauseULASJ0203+0012isredder,theextrapolated WeacknowledgethecontributionsofthestaffofUKIRT,in
absolute B-band magnitude is similar to that of the SDSS particular Andy Adamson, to the implementation UKIDSS
composite, MB =−27.9 and MB =−27.7 respectively. survey and the Cambridge Astronomical Survey Unit and
theWideFieldAstronomyUnitinEdinburghforprocessing
the UKIDSS data. We thank G. Miley and F. Maschietto
fortheirhelp inobtaining theVLTspectra,and thereferee
7 CONCLUSIONS
M. Strauss for his valuable comments. This work is based
We have reported initial results of a survey program us- in part on data obtained as part of the UKIRT Infrared
ing UKIDSS LAS data to discover high-redshift quasars. Deep Sky Survey. The United Kingdom Infrared Telescope
Our results from 106 deg2 demonstrate that the federated isoperatedbytheJoint AstronomyCentreonbehalfof the
UKIDSSandSDSSdatacanbeusedtocarryoutsuchasur- U.K.Particle Physics and Astronomy Research Council.
vey. The analysis and follow-up strategy we have described
is relatively efficient and should be applicable to the entire
UKIDSSLASsurvey area of 4000deg2. REFERENCES
Follow-up observations of the UKIDSS DR1 quasar
Adelman-McCarthy J. K.et al., 2006, submitted to ApJS
candidates in the SDSS Stripe 82 region are now com-
Appenzeller I.et al. 1998, The Messenger, 94, 1
plete. It is difficult to draw any strong conclusions regard-
Casali M. et al., 2006, A&A in press
ing the space density of quasars from one object but we
DekkerH.,Delabre B., Dodorico S., 1986, SPIE, 627, 339
can estimate the number of quasars we would expect in
the UKIDSS survey by extrapolating the z = 6 luminos- DyeS. et al. 2006, MNRAS,372, 1227
Fan X. et al., 2000, AJ, 120, 1167
ity function of Fan et al. (2004) to fainter magnitudes. The
Fan X. et al., 2003, AJ, 125, 1649
density of high-redshift quasars as found in the SDSS is
Φ(M1450 <−26.7)=6×10−10 Mpc−3 (Fanetal.2004)ata Fan X. et al., 2004, AJ, 128, 515
redshiftofz∼6.1.Weassumethatthecomovingdensityof Fan X. et al., 2006, AJ, 132, 117
quasars evolves with redshift as log(Φ)∝ −0.43z (Schmidt, Gnedin N. Y., Fan X., 2006, ApJ,648, 1
Hewett P. C., Warren S. J., Leggett S. K., Hodgkin S. T.,
Schneider& Gunn 1995).
2006, MNRAS,367, 454
Number density predictions are calculated for the Y-
band,corresponding toarest-frame wavelength of ∼1450˚A Kauffmann G., Haehnelt M., 2000, MNRAS,311, 576
at z = 6.1, for which no correction for the presence of Lyα Lawrence A. et al., 2006, submitted to MNRAS
(astro-ph/0604426)
emission or the Lyα forest is required. Integrating down to
OkeJ. B., 1990, AJ, 99, 1621
YAB=20.4(themedian7σlimitoftheDR1data)weexpect
onequasarin∼130deg2 intheredshiftrange5.8<z<6.5. PageL.etal.,2006, submittedtoApJ(astro-ph/0603450)
Pentericci L.et al., 2003, A&A,410, 75
This is consistent with the one object, ULAS J0203+0012,
that was found in 106 deg2. We can estimate in the same Sako M. et al. 2005, astro-ph/0504455
Schmidt M., SchneiderD., Gunn J., 1995, AJ, 110, 68
way the surface density of z > 6.5 quasars in the LAS and
wefindoneobject per∼390 deg2 downtoYAB =20.4. The SkrutskieM. F. et al., 2006, AJ, 131, 1163
Storrie-LombardiL.J.,IrwinM.J.,McMahonR.G.,Hook
UKIDSS LAS two-year plan (Dye et al. 2006) will cover
around2100deg2.Weexpectthisareatocontain∼22new I. M., 2001, MNRAS,322, 933
Warren S. J. et al., 2006, MNRAS in press
z > 5.8 quasars (excluding six quasars already discovered
(astro-ph/0610191)
by SDSS),of which about six would beat z>6.5.
Abovearedshiftofz∼7.5theLyαlineshiftsoutofthe York D.G. et al., 2000, AJ, 120, 1579
Y-band.Potentiallyquasarsabovethisredshiftcouldbedis-
coveredasY-banddropoutobjects. However,thepredicted
number density for such quasars is very low (<1 quasar in
APPENDIX A:
2100 deg2).