THE IDENTIFICATION OF z-DROPOUTS IN PAN-STARRS1: THREE QUASARS AT 6.5< z< 6.7^*

Quasars are the most luminous non-transient objects known. Their high luminosity makes quasars ideal to probe the universe at early cosmic times. Since distant ($z\gtrsim 5.7$) luminous quasars are rare, with an estimated source density of ∼1 Gpc⁻³ (e.g., Fan et al. 2004; Willott et al. 2010b), surveys covering a large area of the sky are required to uncover the distant quasar population. Over the last 15 years more than 70 quasars with redshifts between $5.5\lt z\lt 6.5$ have been discovered in various surveys (e.g., Fan et al. 2006; Jiang et al. 2008; Mortlock et al. 2009; Willott et al. 2010a; Morganson et al. 2012; Bañados et al. 2014). Most of these quasars have been found by looking for sources with a large break between the optical i and z bands (e.g., Fan et al. 2006), the so-called i-band dropouts or i-dropouts. To find quasars beyond $z\sim 6.5$, wide-field surveys with coverage beyond ∼1 μm are needed.

Currently, four quasars above $z\gt 6.5$ have been discovered in near-infrared surveys. Mortlock et al. (2011) presented a quasar at z = 7.1 discovered in the UK infrared Telescope Infrared Deep Sky Survey (UKIDSS; Lawrence et al. 2007), while Venemans et al. (2013) reported three quasars at $6.6\lt z\lt 6.9$ from the Visible and Infrared Survey Telescope Kilo-Degree Infrared Galaxy (VIKING) survey. Detailed studies of these four $z\gt 6.5$ quasars have given insight into the properties of the universe less than a gigayear after the Big Bang. For example, the optical spectrum of the z = 7.1 quasar places constraints on the fraction of neutral hydrogen (Mortlock et al. 2011; Bolton et al. 2011), while Simcoe et al. (2012) use near-infrared spectroscopy to put limits on the metal enrichment ("metallicity") of the intergalactic medium (IGM) up to $z\sim 7$. Furthermore, these quasars set a lower limit to the number density of supermassive black holes at $z\gt 6.5$: $\rho ({{M}_{{\rm BH}}}\gt {{10}^{9}}\;{{M}_{\odot }})\gt 1.1\;\times \;{{10}^{-9}}$ Mpc⁻³ (Venemans et al. 2013; De Rosa et al. 2014).

The limitation of both UKIDSS and VIKING is that they cover only small fractions of the extragalactic sky: UKIDSS has imaged ∼4000 deg², while VIKING will cover an area of 1500 deg². Overcoming this limitation, the Panoramic Survey Telescope & Rapid Response System 1 (Pan-STARRS1 or PS1; Kaiser et al. 2002, 2010) has imaged the whole sky above a declination of $-30{}^\circ$ for about four years in five filters (${{g}_{{\rm P}1}}$, ${{r}_{{\rm P}1}}$, ${{i}_{{\rm P}1}}$, ${{z}_{{\rm P}1}}$, and ${{y}_{{\rm P}1}}$; Stubbs et al. 2010; Tonry et al. 2012). The inclusion of the ${{y}_{{\rm P}1}}$ filter (central wavelength ${{\lambda }_{c}}=9620$ Å, FWHM = 890 Å; Tonry et al. 2012) enables the search for luminous quasars at $z\gt 6.5$ over more than 20,000 deg² of extragalactic sky by selecting sources with a red ${{z}_{{\rm P}1}}$–${{y}_{{\rm P}1}}$ color (z-dropouts; see Figure 1). In this paper, we present the first results from our ongoing search for $z\gt 6.5$ quasars in the PS1 data.

Zoom In Zoom Out Reset image size

All magnitudes are given in the AB system, and we adopt a cosmology with ${{H}_{0}}=70$ km s⁻¹ Mpc⁻¹, ${{{\Omega }}_{M}}=0.28$, and ${{{\Omega }}_{\lambda }}=0.72$ (Komatsu et al. 2011).

2.1. The Pan-STARRS1 3π Survey

To search for z-dropouts, we made use of the first internal release of stacked PS1 imaging data (PV1). We selected our initial $z\gt 6.5$ quasar candidates as objects with a signal-to-noise ratio (S/N)$_{y{\rm P}1}\gt 7$ and a non-detection in the ${{g}_{{\rm P}1}}$, ${{r}_{{\rm P}1}}$, and ${{i}_{{\rm P}1}}$ bands (i.e., S/N $\lt 3$). We further required a break between the ${{z}_{{\rm P}1}}$ and ${{y}_{{\rm P}1}}$ bands (see also Figure 1 in Bañados et al. 2014):

$$\begin{eqnarray*}&&{{({\rm S}/{\rm N})}_{z{\rm P}1}}\gt 3\ \;{\rm AND}\ \;({{z}_{{\rm P}1}}-{{y}_{{\rm P}1}}\gt 1.4)\;{\rm OR}\end{eqnarray*}$$

$$\begin{eqnarray*}&&{{({\rm S}/{\rm N})}_{z{\rm P}1}}\lt 3\;{\rm AND}\;({{z}_{{\rm P}1,{\rm lim} ,3\sigma }}-{{y}_{{\rm P}1}}\gt 1.4).\end{eqnarray*}$$

To avoid extended sources, we demanded that the difference between the ${{y}_{{\rm P}1}}$-band PSF and aperture magnitudes (y_ext) be consistent within 0.3 mag. This value was determined by comparing y_ext with spectroscopically confirmed stars and galaxies from the SDSS-III database¹³ . Setting ${{y}_{{\rm ext}}}\lt 0.3$ selected the vast majority of stars ($\gt 85$%), while rejecting a large fraction ($\gt 94\%$) of galaxies. Finally, we removed sources that were marked as likely spurious in the catalogs (see Bañados et al. 2014) or that had less than 85% of the expected flux in the ${{z}_{{\rm P}1}}$ or ${{y}_{{\rm P}1}}$ images on valid pixels. For bright sources (${{y}_{{\rm P}1}}$ $\lt \;19.5$), we applied the same criteria, but we relaxed our limits in the ${{g}_{{\rm P}1}}$, ${{r}_{{\rm P}1}}$, and ${{i}_{{\rm P}1}}$ bands, requiring S/N $\lt \;5$.

The total number of z-dropouts selected from the PS1 catalogs after removing objects in the plane of the Milky Way and M31 ($\left| b \right|\lt 20{}^\circ$ and $7{}^\circ \;\lt \;$ R.A.$\;\lt 14{}^\circ$; $37{}^\circ \;\lt \;$ decl.$\lt \;43{}^\circ$) was $328,372$ (of which 13,093 had ${{y}_{{\rm P}1}}$ $\lt \;19.5$).

2.2. Public Infrared Surveys

To confirm the colors of the possible quasars and to remove lower-redshift interlopers, we imaged 194 z-dropout candidates during five observing runs. We obtained optical and infrared images between 2014 January 24 and 2014 August 13 with the MPG 2.2 m/GROND (Greiner et al. 2008), NTT/EFOSC2 (Buzzoni et al. 1984), NTT/SofI (Moorwood et al. 1998), and the Calar Alto 3.5 m/Omega2000 (Bailer-Jones et al. 2000); see Table 1 for the details of the observations.

Table 1.  Imaging and Spectroscopic Observations of Quasar Candidates

Objecta	Date	Telescope/Instrument	λ Range / Filters	Exposure Time	Slit Width
	2014 Jan 24–Feb 5	MPG 2.2 m/GROND	${{g}^{\prime }}$, ${{r}^{\prime }}$, ${{i}^{\prime }}$, ${{z}^{\prime }}$, J, H, K	460–960 s	⋯
	2014 Mar 2–6	NTT/EFOSC2	IN, ZN	600 s	⋯
	2014 Mar 2 and 5	NTT/SofI	J	300 s	⋯
	2014 Mar 16–19	CAHA 3.5 m/Omega2000	z, Y, J	300–600 s	⋯
	2014 Jul 23–28	NTT/EFOSC2	IN, ZN	600 s	⋯
	2014 Jul 25	NTT/SofI	J	600 s	⋯
	2014 Aug 7 and 11–13	CAHA 3.5 m/Omega2000	Y, J	600 s	⋯
P167-13	2014 Apr 26	VLT/FORS2	0.74–1.07 μm	2630 s	13
	2014 May 30–Jun 2	Magellan/FIRE	0.82–2.49 μm	12004 s	06
P036+03	2014 Jul 25	NTT/EFOSC2	0.60–1.03 μm	7200 s	12
	2014 Sep 4–6	Magellan/FIRE	0.82–2.49 μm	8433 s	06
	2014 Oct 20	Keck I/LRIS	0.55–1.03 μm	1800 sb	10
P338+29	2014 Oct 19	MMT/Red Channel	0.67–1.03 μm	1800 s	10
	2014 Oct 30	Magellan/FIRE	0.82–2.49 μm	7200 sb	06
	2014 Nov 27	LBT/MODS	0.51–1.06 μm	2700 s	12
	2014 Dec 6	LBT/LUCI	2.05–2.37 μm	3360 s	15

^aFor the full name and coordinates, see Table 2. ^bObservations in cloudy conditions.

Download table as:  ASCIITypeset image

Candidates were considered foreground interlopers if they had ${{y}_{{\rm P}1}}-J\gt 1$ (see Section 2.2). Sources with a ${{y}_{{\rm P}1}}-J\lt -1$ or undetected in J were rejected on the basis that they could be moving, varying, or a spurious object in the PS1 catalog. We reobserved objects with $-1.0\lt {{y}_{{\rm P}1}}-J\lt 1.0$ with the NTT in the filters I_N and Z_N. Only three sources remained undetected in I_N or were red with ${{I}_{N}}-{{Z}_{N}}\gtrsim 2$. These objects were targets for spectroscopy.

We obtained optical and near-infrared spectroscopy of all the three candidates that had good quasar colors after the follow-up imaging. We carried out spectroscopic observations between 2014 April 26 and 2014 December 6 using the following instruments: VLT/FORS2 (Appenzeller et al. 1998); Megallan/FIRE (Simcoe et al. 2008, 2010); NTT/EFOSC2; Keck/LRIS (Oke et al. 1995); MMT/Red Channel Spectrograph; and LBT/MODS (Pogge et al. 2010) and LBT/LUCI (Seifert et al. 2003). Details of the observations are listed in Table 1.

We reduced the data following standard reduction steps (e.g., Venemans et al. 2013; Bañados et al. 2014). We show the merged spectra in Figure 1.

All three z-dropouts for which we obtained optical spectroscopy showed a strong continuum break in their optical spectrum (Figure 1) and were identified as quasars at redshifts $6.5\lt z\lt 6.7$. We fitted the continuum with three components: a power law with slope β (${{f}_{\lambda }}\sim {{\lambda }^{\beta }}$), a Balmer continuum, and an Fe ii template (see, e.g., De Rosa et al. 2014). In all cases, the Balmer continuum was found to be negligible at the wavelengths we considered (${{\lambda }_{{\rm rest}}}\lesssim 3000$ Å). A single-Gaussian fit of the emission lines (most prominently C iv and Mg ii) provided a sufficiently good model of the line profiles given the S/N of our spectra. Only in the spectrum of the brightest quasar, PSO J036.5078 + 03.0498, were we able to constrain on the Fe ii emission (Section 4.2). The near-infrared spectra of the other two quasars did not have sufficient S/N. This did not significantly affect the fit of the Mg ii lines in these quasars.

The redshifts were determined by the peak of the Mg ii line. Other bright emission lines, such as Si iv λ 1397 and C iv λ 1549, are blueshifted by 300–2000 km s⁻¹ with respect to Mg ii. Such shifts are similar to those measured in spectra of other distant luminous quasars (e.g., Richards et al. 2002; De Rosa et al. 2014).

We estimated black hole masses using the local scaling relation based on the Mg ii line (Equation (1) in Vestergaard & Osmer 2009), which has a systematic uncertainty of a factor of ∼3. The black hole mass uncertainties quoted in Sections 4.1–4.3 and in Table 2 represent only statistical errors.

Table 2.  Photometric Properties and Derived Parameters of the New $z\gt 6.5$ Quasars from PS1

	PSO J167.6415-13.4960	PSO J036.5078+03.0498	PSO J338.2298+29.5089
Abbreviated name	P167-13	P036+03	P338+29
R.A.(J2000)	11h10m3398	02h26m0188	22h32m5515
Decl.(J2000)	–13°29'456	+03°02'594	+29°30'322
IN	$\gt 24.68$ a	23.62 ± 0.18	$\gt 24.37$ a
${{z}_{{\rm P}1}}$	$\gt 22.54$ a	21.51 ± 0.24	$\gt 22.25$ a
ZN	22.08 ± 0.09	20.46 ± 0.04	21.92 ± 0.09
${{y}_{{\rm P}1}}$	20.49 ± 0.12	19.37 ± 0.04	20.13 ± 0.08
J	21.21 ± 0.09	19.51 ± 0.03	20.74 ± 0.09
${{W}_{1,{\rm AB}}}$	21.13 ± 0.39b	19.43 ± 0.08	20.51 ± 0.21
${{M}_{1450}}$	−25.58 ± 0.13	−27.36 ± 0.03	−26.04 ± 0.09
β	−1.0 ± 0.1	−1.70 ± 0.05	$-1.85_{-0.05}^{+0.08}$
zMgII	6.508 ± 0.001	6.527 ± 0.002	6.658 ± 0.007
FWHMMgII (km s−1)	2350 ± 470	$3500_{-740}^{+1010}$	$6800_{-900}^{+1200}$
${{M}_{{\rm BH}}}$ (${{M}_{\odot }}$)	$(4.9\pm 2.0)\times {{10}^{8}}$	$(1.9_{-0.8}^{+1.1})\times {{10}^{9}}$	$(3.7_{-1.0}^{+1.3})\times {{10}^{9}}$
${{L}_{{\rm Bol},3000\,\overset{\circ}{\rm A} }}/{{L}_{{\rm Edd}}}$	1.2 ± 0.5	$0.96_{-0.35}^{+0.70}$	$0.13_{-0.04}^{+0.05}$
RNZ (proper Mpc)	1.5 ± 0.7	3.1 ± 0.7	5.2 ± 0.7
${{R}_{{\rm NZ},{\rm corrected}}}$ (proper Mpc)	2.3 ± 1.1	2.8 ± 0.6	6.9 ± 0.9

^aNon-detections listed as $3\sigma$ upper limits. ^bListed in the AllWISE Reject table.

Download table as:  ASCIITypeset image

We computed bolometric luminosities by applying the bolometric correction obtained by Shen et al. (2008) to the monochromatic luminosity density measured at 3000 Å. The Eddington luminosity is defined as ${{L}_{{\rm Edd}}}=1.3\times {{10}^{38}}\;({{M}_{{\rm BH}}}/{{M}_{\odot }})$ erg s⁻¹.

A summary of the photometric properties and the parameters derived from the spectra is provided in Table 2. Below, we describe the new quasars in more detail.

4.1. PSO J167.6415-13.4960

4.2. PSO J036.5078+03.0498

4.3. PSO J338.2298+29.5089

Quasar near zones are the regions surrounding distant quasars where the UV radiation of the central source has ionized the H i. The size of the near zone (R_NZ) is a function of, among others, the quasar age, the flux of ionizing photons, and the fraction of neutral H (f_Hi ) in the IGM. A study of near zones of $5.7\lt z\lt 6.4$ quasars was performed by Carilli et al. (2010), measuring R_NZ around 27 quasars. To compare the near zone sizes as function of redshifts, Carilli et al. (2010) scaled the measured R_NZ to an absolute magnitude of ${{M}_{1450}}=-27$: ${{R}_{{\rm NZ},{\rm corrected}}}={{R}_{{\rm NZ}}}\times {{10}^{0.4(27+M1450)/3}}$. They found that the ionized region around quasars decreases with increasing redshift and follows the relation ${{R}_{{\rm NZ},{\rm corrected}}}=(7.4\pm 0.3)-(8.0\pm 1.1)\times (z-6)$. This signals an increase in f_Hi close to quasars at higher redshifts, although it is not straightforward to translate a change in R_NZ to a change in f_Hi (e.g., Bolton & Haehnelt 2007).

We measured the near zones following the method described in Fan et al. (2006) and also employed by Carilli et al. (2010). The results are shown in Figure 2(a). We derive near-zone radii of 1.5, 3.1, and 5.2 Mpc (proper) for P167-13, P036 + 03, and P338 + 29, respectively. The uncertainty in the computed near zone (including the uncertainty in the quasar's systemic redshift derived from Mg ii) is about 0.7 Mpc (Carilli et al. 2010). The sizes scaled to ${{M}_{1450}}=-27$ are ${{R}_{{\rm NZ},{\rm corrected}}}=2.3$, 2.8, and 6.9 Mpc, respectively.

Zoom In Zoom Out Reset image size

In Figure 2(b), we compare the near zones of the PS1 quasars with those of $5.7\lt z\lt 7.1$ quasars from the literature. The PS1 quasars roughly follow the trend of smaller near zones at higher redshifts. A weighted linear fit results in a relation ${{R}_{{\rm NZ},{\rm corrected}}}=(7.2\pm 0.2)-(6.1\pm 0.7)\times (z-6)$ Mpc. Interpreting the decrease in R_NZ as an increase in f_Hi , ${{R}_{{\rm NZ}}}\propto {{(1+z)}^{-1}}f_{{\rm HI}}^{-1/3}$ (e.g., Fan et al. 2006; but see Bolton & Haehnelt 2007), the decrease in ${{R}_{{\rm NZ},{\rm corrected}}}$ by a factor of 6.5 between z = 6 and z = 7 implies an increase in the neutral fraction of a factor of ∼180. Combined with a measured ${{f}_{{\rm HI}}}\approx 2\times {{10}^{-4}}$ at $z\sim 6$ (e.g., Fan et al. 2006), this suggests ${{f}_{{\rm HI}}}\approx 0.04$ at z = 7, confirming the rapid evolution of f_Hi at $z\gt 6$ (e.g., Fan et al. 2006; Bolton et al. 2011). The large spread (a factor of ∼3) in (corrected) near-zone sizes between individual quasars indicates a wide range in quasar ages and/or a large variation in f_Hi along different lines of sight.

We identified three new quasars at redshifts $6.5\lt z\lt 6.7$ in PS1, nearly doubling the number of known $z\gt 6.5$ quasars from four to seven. The newly discovered quasars have a wide range of properties (Table 2). The rest-frame UV luminosities are between ${{M}_{1450}}=-25.6$ and ${{M}_{1450}}=-27.4$. The brightest of the PS1 quasars is the most luminous quasar discovered at $z\gt 6.5$ so far (Figure 3), with a luminosity at 1450 Å close to that of the bright SDSS quasar J1148 + 5251 at z = 6.42 (Fan et al. 2003). The faintest PS1 quasar is only marginally brighter than the faintest $z\gt 6.5$ quasar found in the VIKING survey (J0109–3047; Venemans et al. 2013). Since the areal coverage of PS1 is more than 10× larger than that of VIKING, this is very promising for our continuing PS1 z-dropout search.

Zoom In Zoom Out Reset image size

The PS1 quasars are powered by black holes with estimated masses of $(0.5-4)\times {{10}^{9}}$ ${{M}_{\odot }}$, based on the Mg ii line widths and the quasar luminosities. The black holes are accreting in the range of 0.13–1.2 times the Eddington limit. Black hole masses, accretion rates, and (when estimated) Mg ii/Fe ii ratio are similar to those derived for other $z\gt 6$ quasars (e.g., Willott et al. 2010a; De Rosa et al. 2014).

We derived the ionized region around the quasars and found (luminosity-corrected) near zones between 2.3 and 6.9 Mpc, in line with the sizes measured around $5.7\lt z\lt 6.4$ quasars. By comparing the near-zone radii of quasars between $5.7\lt z\lt 7.1$, we derive that the average size of the quasar ionization region decreases by a factor of ∼6.5 between z = 6 and z = 7. This implies a neutral hydrogen fraction in the IGM of a few percent at z = 7, although the scatter in R_NZ at all redshifts (a factor of 3 between the new quasars) suggests large variations in f_Hi along different lines of sight.

We thank the referee for carefully reading the manuscript and providing constructive comments and suggestions.

B.P.V., E.P.F., and F.W. acknowledge funding through ERC grant "Cosmic Dawn." E.B. thanks the IMPRS for Astronomy & Cosmic Physics at the University of Heidelberg. X.F. and I.D.M. acknowledge support from US NSF grant AST 11-07682, and R.S. and D.M. from US NSF grant AST-1109915.

The Pan-STARRS1 Surveys have been made possible through contributions of the Institute for Astronomy, University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, Max-Planck-Institute for Astronomy, Heidelberg and Max-Planck-Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, University of Edinburgh, Queens University Belfast, Harvard-Smithsonian Center for Astrophysics, the Las Cumbres Observatory Global Telescope Network Incorporated, the National Central University of Taiwan, the Space Telescope Science Institute, the National Aeronautics and Space Administration under grant No. NNX08AR22G issued through the Planetary Science Division of the NASA Science Mission Directorate, the National Science Foundation under grant AST-1238877, the University of Maryland, and Eotvos Lorand University (ELTE).

Part of the funding for GROND was granted from the Leibniz-Prize to Prof. G. Hasinger (DFG grant HA 1850/28-1).

This publication makes use of data products from the Wide-field Infrared Survey Explorer, a joint project of the University of California, Los Angeles, and the Jet Propulsion Laboratory/California Institute of Technology, funded by the National Aeronautics and Space Administration.

The LBT is an international collaboration among institutions in the USA, Italy, and Germany. The partners are The University of Arizona; Istituto Nazionale di Astrofisica, Italy; LBT Beteiligungsgesellschaft, Germany, representing the Max-Planck Society, the Astrophysical Institute Potsdam, and Heidelberg University; The Ohio State University; The Research Corporation, on behalf of The University of Notre Dame, University of Minnesota, and University of Virginia.

Facilities: PS1 (GPC1) - Panoramic Survey Telescope and Rapid Response System Telescope #1 (Pan-STARRS), UKIRT (WFCAM) - United Kingdom Infrared Telescope, ESO:VISTA (VIRCAM) - , WISE - Wide-field Infrared Survey Explorer, NTT - New Techology Telescope (EFOSC2 SOFI), Max Planck:2.2 m (GROND) - , CAO:3.5 m (OMEGA2000) - , VLT:Antu (FORS2) - , Magellan:Baade (FIRE) - , Keck:I (LRIS) - , MMT (Red Channel Spectrograph) - MMT at Fred Lawrence Whipple Observatory, LBT - Large Binocular Telescope Observatory (MODS, LUCI).