Subaru High-z Exploration of Low-luminosity Quasars (SHELLQs). IV. Discovery of 41 Quasars and Luminous Galaxies at 5.7 ≤ z ≤ 6.9

High-z quasars²⁴ are an important probe of the early universe in many ways. Their rest-ultraviolet (UV) spectra blueward of Lyα are very sensitive to the H i absorption, and thus indicate the progress of cosmic reionization through the neutral fraction of the intergalactic medium (IGM; Gunn & Peterson 1965; Fan et al. 2006). The very existence of such objects puts strong restrictions on, and sometimes even challenges, models of the formation and early evolution of supermassive black holes (SMBH; e.g., Volonteri 2012; Ferrara et al. 2014; Madau et al. 2014). High-z quasars are also used to probe the assembly of the host galaxies, which are thought to form in the highest density peaks of the dark matter distribution in the early phase of cosmic history (e.g., Goto et al. 2009; Decarli et al. 2017).

Previous surveys have identified more than 100 high-z quasars so far (e.g., Bañados et al. 2016, and references therein), with the most distant objects known at z = 7.54 (Bañados et al. 2018) and z = 7.085 (Mortlock et al. 2011). However, most of the known quasars have redshifts z < 6.5 and UV absolute magnitudes M₁₄₅₀ < −24 mag, while higher redshifts and fainter magnitudes remain largely unexplored. There must be numerous objects of faint quasars and active galactic nuclei (AGNs) behind the known luminous quasars; they may represent the more typical mode of SMBH growth in the early universe, and may have made a significant contribution to reionization.

For the past few years, we have been carrying out a high-z quasar survey with the Subaru 8.2 m telescope. We have already reported discoveries of 33 high-z quasars, along with 14 high-z luminous galaxies, 2 [O iii] emitters at z ∼ 0.8, and 15 Galactic brown dwarfs, in Matsuoka et al. (2016, 2018, Paper I and II, hereafter). Multi-wavelength follow-up observations of the discovered objects are ongoing, whose initial results from the Atacama Large Millimeter/submillimeter Array (ALMA) data are presented in Izumi et al. (2018, Paper III). This "Subaru High-z Exploration of Low-Luminosity Quasars (SHELLQs)" project is based on the exquisite multi-band photometry data collected by the Hyper Suprime-Cam (HSC) Subaru Strategic Program (SSP) survey (Aihara et al. 2018a). HSC is a wide-field camera on the Subaru Telescope, and has a nearly circular field of view of 15 diameter, covered by 116 2K × 4K Hamamatsu fully depleted CCDs, with the pixel scale of 017 (Miyazaki et al. 2018). The camera dewar design and on-site quality assurance design are described in Komiyama et al. (2018) and Furusawa et al. (2018), respectively. The HSC-SSP survey has three layers with different combinations of area and depth. The Wide layer is observing 1400 deg² mostly along the celestial equator, with 5σ point-source depths of (g_AB, r_AB, i_AB, z_AB, y_AB) = (26.5, 26.1, 25.9, 25.1, 24.4) mag measured in 20 aperture. The Deep and the UltraDeep layers are observing smaller areas (27 and 3.5 deg²) down to deeper limiting magnitudes (r_AB = 27.1 and 27.7 mag, respectively). The observed data are processed with the dedicated pipeline hscPipe (Bosch et al. 2018), which was developed from the Large Synoptic Survey Telescope software pipeline (Jurić et al. 2015). The HSC pipeline is evolving through the period of the survey, which sometimes gives rise to somewhat inconsistent photometry (and the quasar probability of a given source; see below) between different data releases. A full description of the HSC-SSP survey can be found in Aihara et al. (2018a).

This paper is the fourth in a series of SHELLQs publications, and reports spectroscopic identification of an additional 73 objects in the latest HSC-SSP data. We describe the photometric candidate selection briefly in Section 2, while a more complete description may be found in Papers I and II. The spectroscopic follow-up observations are described in Section 3. The quasars and other classes of objects we discovered are presented in Section 4. A summary appears in Section 5. A companion paper (Y. Matsuoka et al. 2018, in preparation) describes the quasar luminosity function at z ∼ 6 derived from the SHELLQs sample obtained so far. This paper adopts the cosmological parameters H₀ = 70 km s⁻¹ Mpc⁻¹, Ω_M = 0.3, and Ω_Λ = 0.7. All magnitudes in the optical and NIR bands are presented in the AB system (Oke & Gunn 1983), and are corrected for Galactic extinction (Schlegel et al. 1998). We use two types of magnitudes; the point-spread function (PSF) magnitude (m_AB) is measured by fitting PSF models to the source profile, while the CModel magnitude (m_CModel,AB) is measured by fitting PSF-convolved, two-component galaxy models (Abazajian et al. 2004). We use the PSF magnitude error (σ_m) measured by the HSC data processing pipeline (Bosch et al. 2018); it does not contain photometric calibration uncertainty, which is estimated to be at least 1% (Aihara et al. 2018b). In what follows, we refer to z-band magnitudes with the AB subscript ("z_AB"), while redshift z appears without a subscript.

Our photometric candidate selection starts from the HSC-SSP source catalog. We used the survey data in the three layers, observed before 2017 May, i.e., a newer data set than that contained in the latest public data release (Aihara et al. 2018b). We required that our quasar survey field has been observed in the i, z, and y bands (not necessarily to the full depths; see below), but impose no requirement on the g- or r-band coverage. All sources meeting the following criteria, without critical quality flags (see Paper I), are selected:

$$\begin{eqnarray}&&({z}_{\mathrm{AB}}\lt 24.5\ \mathrm{and}\ {\sigma }_{z}\lt 0.155\ \mathrm{and}\ {i}_{\mathrm{AB}}-{z}_{\mathrm{AB}}\gt 1.5\\ &&\quad \mathrm{and}\ {z}_{\mathrm{AB}}-{z}_{\mathrm{CModel},\mathrm{AB}}\lt 0.15)\end{eqnarray} \tag{ 1 }$$

or

$$\begin{eqnarray}&&({y}_{\mathrm{AB}}\lt 24.0\ \mathrm{and}\ {\sigma }_{y}\lt 0.155\ \mathrm{and}\ {z}_{\mathrm{AB}}-{y}_{\mathrm{AB}}\gt 0.8\\ &&\quad \mathrm{and}\ {y}_{\mathrm{AB}}-{y}_{\mathrm{CModel},\mathrm{AB}}\lt 0.15).\end{eqnarray} \tag{ 2 }$$

Here, we use the difference between the PSF and CModel magnitudes (m_AB–m_CModel,AB) to reject extended sources. With the present threshold value (0.15), our completeness of point-source selection is >80% at z_AB < 24.0 mag (Paper II). The completeness decreases toward fainter magnitude, which should be accounted for in statistical measurements (e.g., luminosity function) of the discovered quasars. We further remove sources with more than 3σ detection in the g- or r-band, if these bands are available, as such sources are most likely low-z interlopers.

Next, we process the HSC images of the above sources through an automatic image-checking procedure, in all the available bands. As detailed in Papers I and II, this procedure uses Source Extractor (Bertin & Arnouts 1996), and removes those sources whose photometry is not consistent (within 5σ significance) between the stacked and all the pre-stacked, individual images. We also eliminate sources with too-compact, diffuse, or elliptical profiles to be celestial point sources on the stacked images. The sources removed at this stage are mostly cosmic rays, moving or transient sources, or image artifacts.

The candidates selected above are matched, within 10 separation, to the public near-IR catalogs from the United Kingdom Infrared Telescope Infrared Deep Sky Survey (Lawrence et al. 2007), the Visible and Infrared Survey Telescope for Astronomy (VISTA) Kilo-degree Infrared Galaxy survey, the VISTA Deep Extragalactic Observations Survey (Jarvis et al. 2013), and the UltraVISTA survey (McCracken et al. 2012). The choice of the 10 matching radius is rather arbitrary, but is at least sufficiently larger than the astrometric uncertainties of the above surveys. Then, using the i, z, and y-band magnitudes along with all the available J, H, and/or K magnitudes, we calculate Bayesian quasar probability (${P}_{{\rm{Q}}}^{{\rm{B}}}$ ) for each candidate, based on spectral energy distribution (SED) models and estimated surface densities of high-z quasars and contaminating brown dwarfs as a function of magnitude. Our algorithm does not contain galaxy models at present. We created the quasar SED models by stacking the SDSS spectra of 340 bright quasars at z ≃ 3, where the quasar color selection is fairly complete (Richards et al. 2002; Willott et al. 2005), and correcting for the effect of IGM absorption (Songaila 2004). The quasar surface density was modeled with the luminosity function of Willott et al. (2010). The color models of brown dwarfs were computed with a set of observed spectra compiled in the SpeX prism library,²⁵ and the CGS4 library,²⁶ while the surface densities were calculated following Caballero et al. (2008). A more detailed description of our Bayesian algorithm may be found in Paper I. We reject sources with P_Q^B < 0.1, and keep those with higher quasar probabilities in the sample of candidates.

Finally, we inspect the images of all the candidates by eye and remove additional problematic sources. About 80% of the remaining candidates were rejected at this last stage, and were mostly cosmic rays but also included transient/moving objects and sources close to bright stars. The latest HSC-SSP (internal) data release covers 650 deg², when we limit to the field where at least a single exposure in the i and two exposures each in the z and y bands were obtained. From the final sample of candidates, we put the highest priority for spectroscopic identification on a subsample with relatively bright magnitudes (z_AB < 24.0 mag or y_AB < 23.5 mag), red colors (i_AB − z_AB > 2.0 or z_AB − y_AB > 0.8), and detection in more than a single band (for i-dropouts) or a single exposure (for z-dropouts). The number of photometric candidates is changing throughout the survey, due to the continuous arrival of new HSC-SSP survey data, a change in the photometry and quasar probability (${P}_{{\rm{Q}}}^{{\rm{B}}}$) with updates of the data processing pipeline, and the progress of our follow-up spectroscopy.

The present HSC survey footprint includes 15 high-z quasars discovered by other surveys, as summarized in Table 1. We recovered 10 of these quasars with ${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$, while three quasars were not selected due to their relatively low redshifts (z < 5.9) and bluer HSC colors (i_AB − z_AB < 1.5) than the HSC color selection threshold described above. The remaining two quasars are missing due to nearby bright sources; VIK J0839+0015 is assigned with an HSC pixelflags_bright_objectcenter flag (indicating that the pipeline measurements of this source may be affected by a nearby bright source) and therefore does not meet the HSC source selection criteria described above, while SDSS J1602+4228 is rejected in our automatic image-checking procedure, since its i-band photometry is not consistent between stacked and pre-stacked images, likely due to blending with a foreground galaxy.

Table 1.  Recovery of Known High-z Quasars

Name	R.A.	Decl.	Redshift	Recovered?	Comment
CFHQS J0210−0456	02h10m13s19	−04°56'209	6.43	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
CFHQS J0216−0455	02h16m2781	−04°55'341	6.01	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
CFHQS J0227−0605	02h27m4329	−06°05'302	6.20	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
SDSS J0836+0054	08h36m4386	+00°54'533	5.81	N	iAB−zAB < 1.5
VIK J0839+0015	08h39m5536	+00°15'542	5.84	N	Close to a bright source
VIK J1148+0056	11h48m3318	+00°56'423	5.84	N	iAB−zAB < 1.5
VIK J1215+0023	12h15m1687	+00°23'247	5.93	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
PSO J184.3389+01.5284	12h17m2134	+01°31'425	6.20	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
SDSS J1602+4228	16h02m5398	+42°28'249	6.09	N	Blended with a foreground galaxy
IMS J2204+0012	22h04m1792	+01°11'448	5.94	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
VIMOS 2911001793	22h19m1722	+01°02'489	6.16	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
SDSS J2228+0110	22h28m4354	+01°10'322	5.95	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
CFHQS J2242+0334	22h42m3755	+03${}^{^\circ }$34'216	5.88	N	iAB−zAB < 1.5
SDSS J2307+0031	23h07m3535	+00°31'494	5.87	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$
SDSS J2315−0023	23h15m4657	−00°23'581	6.12	Y	${P}_{{\rm{Q}}}^{{\rm{B}}}=1.000$

Note. The naming convention follows Bañados et al. (2016).

Download table as:  ASCIITypeset image

Our previous papers reported the results of follow-up spectroscopy carried out before the autumn of 2016. Since that time, we have observed 73 additional quasar candidates, using the Optical System for Imaging and low-Intermediate-Resolution Integrated Spectroscopy (OSIRIS; Cepa et al. 2000) mounted on the 10.4 m Gran Telescopio Canarias (GTC), and the Faint Object Camera and Spectrograph (FOCAS; Kashikawa et al. 2002) mounted on Subaru. We observed roughly the brightest one-third of the candidates with OSIRIS, and the remaining candidates with FOCAS. The observations were scheduled in such a way that the targets with brighter magnitudes and higher ${P}_{{\rm{Q}}}^{{\rm{B}}}$ were observed at earlier opportunities. The journal of these discovery observations is presented in Table 2. We present the details of these observations below.

3.1. GTC/OSIRIS

3.2. Subaru/FOCAS

Figures 1–9 present the reduced spectra of the 73 candidates we observed. They include 31 high-z quasars, 10 high-z galaxies, 4 strong [O iii] emitters at z ∼ 0.8, and 28 cool dwarfs (low-mass stars and brown dwarfs). Their photometric properties are summarized in Table 3. The astrometric accuracy of the HSC-SSP data is estimated to be ≲01 (root mean square; Aihara et al. 2018b). Ten of these objects are detected in the J, H, and/or K bands, as summarized in Table 4.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 3.  Photometric Properties

Name	R.A.	Decl.	iAB (mag)	zAB (mag)	yAB (mag)	${P}_{{\rm{Q}}}^{{\rm{B}}}$
Quasars
J2210+0304	22h10m2724	+03°04'285	27.36 ± 0.78	>25.03	22.94 ± 0.06	1.000
J0213−0626	02h13m1694	−06°26'152	26.19 ± 0.30	>25.05	21.69 ± 0.03	1.000
J0923+0402	09h23m4712	+04°02'544	26.45 ± 0.24	22.64 ± 0.02	20.21 ± 0.01	1.000
J0921+0007	09h21m2056	+00°07'229	26.28 ± 0.22	21.83 ± 0.01	21.24 ± 0.01	1.000
J1545+4232	15h45m0562	+42°32'116	25.74 ± 0.12	22.45 ± 0.01	22.16 ± 0.04	1.000
J1004+0239	10h04m0136	+02°39'307	26.41 ± 0.11	22.76 ± 0.01	22.58 ± 0.01	1.000
J0211−0203	02h11m4453	−02°03'039	27.47 ± 1.05	24.04 ± 0.06	23.66 ± 0.10	0.999
J2304+0045	23h04m2297	+00°45'054	26.85 ± 0.33	23.08 ± 0.02	22.75 ± 0.04	1.000
J2255+0251	22h55m3804	+02°51'266	>25.72	23.44 ± 0.05	22.95 ± 0.07	1.000
J1406−0116	14h06m2912	−01°16'112	25.78 ± 0.13	22.64 ± 0.02	21.93 ± 0.03	1.000
J1146−0005	11h46m5889	−00°05'377	>26.43	23.76 ± 0.04	24.78 ± 0.27	1.000
J1146+0124	11h46m4842	+01°24'201	26.76 ± 0.42	23.01 ± 0.03	23.05 ± 0.06	1.000
J0918+0139	09h18m3317	+01°39'233	26.70 ± 0.30	23.17 ± 0.03	23.18 ± 0.04	1.000
J0844−0132	08h44m0861	−01°32'165	>26.24	23.31 ± 0.03	23.73 ± 0.15	1.000
J1146−0154	11h46m3266	−01°54'383	26.90 ± 0.55	23.60 ± 0.06	23.76 ± 0.14	1.000
J0834+0211	08h34m0088	+02°11'469	26.27 ± 0.32	22.97 ± 0.03	22.85 ± 0.05	1.000
J0909+0440	09h09m2150	+04°40'429	25.02 ± 0.06	22.18 ± 0.02	21.91 ± 0.03	1.000
J2252+0225	22h52m0544	+02°25'319	26.78 ± 0.37	23.80 ± 0.06	23.88 ± 0.11	1.000
J1406−0144	14h06m4688	−01°44'026	26.78 ± 0.37	23.33 ± 0.04	23.40 ± 0.13	1.000
J1416+0147	14h16m5301	+01°47'022	26.60 ± 0.55	22.89 ± 0.04	23.46 ± 0.14	1.000
J0957+0053	09h57m4039	+00°53'336	>26.91	23.66 ± 0.01	23.58 ± 0.04	1.000
J2223+0326	22h23m0951	+03°26'203	24.07 ± 0.03	21.31 ± 0.01	21.41 ± 0.02	1.000
J1400−0125	14h00m2999	−01°25'210	25.16 ± 0.08	22.88 ± 0.02	22.82 ± 0.06	1.000
J1400−0011	14h00m2879	−00°11'515	26.15 ± 0.19	23.62 ± 0.04	24.16 ± 0.18	1.000
J1219+0050	12h19m0534	+00°50'375	24.90 ± 0.05	22.81 ± 0.02	23.13 ± 0.06	1.000
J0858+0000	08h58m1351	+00°00'571	24.28 ± 0.04	21.29 ± 0.01	21.43 ± 0.01	1.000
J0220−0432	02h20m2972	−04°32'040	26.29 ± 0.18	24.42 ± 0.10	24.01 ± 0.11	0.000
J1422+0011	14h22m0023	+00°11'030	26.22 ± 0.14	23.85 ± 0.05	23.56 ± 0.08	1.000
J2231−0035	22h31m4889	−00°35'475	26.52 ± 0.24	24.27 ± 0.07	24.35 ± 0.18	1.000
J1209−0006	12h09m2399	−00°06'465	26.69 ± 0.23	24.13 ± 0.05	24.13 ± 0.12	1.000
J1550+4318	15h50m0093	+43°18'028	25.52 ± 0.08	23.71 ± 0.03	23.54 ± 0.10	1.000
Galaxies
J1428+0159	14h28m2471	+01°59'344	26.05 ± 0.43	22.90 ± 0.05	22.83 ± 0.07	1.000
J0917−0056	09h17m0028	−00°56'581	26.06 ± 0.18	23.69 ± 0.05	23.81 ± 0.10	1.000
J0212−0315	02h12m3667	−03°15'176	26.98 ± 0.50	23.85 ± 0.05	23.90 ± 0.13	1.000
J0212−0532	02h12m4946	−05°32'380	>26.22	24.40 ± 0.08	24.31 ± 0.19	1.000
J2311−0050	23h11m4286	−00°50'207	26.83 ± 0.41	24.23 ± 0.09	23.91 ± 0.20	0.356
J1609+5515	16h09m5279	+55°15'485	25.89 ± 0.08	24.15 ± 0.03	24.20 ± 0.09	1.000
J1006+0300	10h06m3353	+03°00'052	25.79 ± 0.10	23.70 ± 0.02	23.65 ± 0.05	1.000
J0914+0442	09h14m3647	+04°42'317	25.46 ± 0.12	23.29 ± 0.06	23.07 ± 0.08	0.980
J0219−0132	02h19m3094	−01°32'072	25.47 ± 0.24	24.27 ± 0.07	24.36 ± 0.13	0.820
J0915−0051	09h15m4502	−00°51'360	25.61 ± 0.13	24.04 ± 0.08	24.19 ± 0.15	0.316
[O iii] Emitters
J1000+0211	10h00m1246	+02°11'274	24.65 ± 0.04	22.87 ± 0.02	25.16 ± 0.28	1.000
J0154−0116	01h54m3169	−01°16'195	>23.08	22.75 ± 0.04	23.19 ± 0.07	1.000
J2226+0237	22h26m4968	+02°37'540	24.25 ± 0.03	22.61 ± 0.02	24.57 ± 0.24	1.000
J0845−0123	08h45m1654	−01°23'216	23.59 ± 0.02	21.89 ± 0.01	24.31 ± 0.19	1.000
Cool Dwarfs
J0158−0033	01h58m0710	−00°33'321	>23.65	23.80 ± 0.07	23.68 ± 0.07	0.119
J0207−0052	02h07m0105	−00°52'250	25.34 ± 0.32	22.22 ± 0.01	21.21 ± 0.01	0.831
J0213−0334	02h13m5082	−03°34'452	25.99 ± 0.20	23.10 ± 0.03	22.10 ± 0.03	0.105
J0216−0207	02h16m1718	−02°07'193	26.14 ± 0.26	23.99 ± 0.06	23.18 ± 0.06	0.000
J0220−0134	02h20m4741	−01°34'506	25.73 ± 0.31	24.41 ± 0.08	24.27 ± 0.12	0.042
J0225−0351	02h25m0018	−03°51'468	25.33 ± 0.08	23.50 ± 0.05	22.96 ± 0.05	0.000
J0837−0000	08h37m1718	−00°00'210	23.18 ± 0.01	20.24 ± 0.01	19.11 ± 0.01	0.000
J0856+0248	08h56m0910	+02°48'511	24.63 ± 0.12	23.24 ± 0.03	22.79 ± 0.05	0.000
J0900+0424	09h00m1840	+04°24'155	27.46 ± 0.45	24.11 ± 0.08	23.21 ± 0.08	0.021
J0902−0030	09h02m2231	−00°30'404	26.02 ± 0.22	24.07 ± 0.05	23.55 ± 0.07	0.000
J0906−0206	09h06m5082	−02°06'102	25.06 ± 0.14	22.93 ± 0.05	22.29 ± 0.04	0.000
J0906+0431	09h06m5510	+04°31'319	25.05 ± 0.06	23.56 ± 0.06	23.12 ± 0.08	0.000
J0912−0121	09h12m1099	−01°21'029	>25.77	23.74 ± 0.07	23.17 ± 0.07	0.994
J1359+0134	13h59m4500	+01°34'123	24.60 ± 0.07	23.32 ± 0.06	23.31 ± 0.10	0.533
J1400+0106	14h00m1538	+01°06'237	27.91 ± 0.85	25.10 ± 0.15	23.50 ± 0.07	0.000
J1415−0113	14h15m3220	−01°13'143	25.08 ± 0.06	22.30 ± 0.01	21.22 ± 0.01	0.000
J1432+0045	14h32m0578	+00°45'317	25.59 ± 0.12	24.01 ± 0.07	23.89 ± 0.11	0.027
J1434−0204	14h34m3736	−02°04'167	25.33 ± 0.18	22.36 ± 0.03	21.42 ± 0.02	0.000
J1435+0040	14h35m2969	+00°40'353	25.17 ± 0.06	22.12 ± 0.01	21.02 ± 0.01	0.000
J1607+5417	16h07m3284	+54°17'245	26.46 ± 0.12	24.23 ± 0.03	23.63 ± 0.05	0.295
J1620+4438	16h20m4691	+44°38'397	25.64 ± 0.16	23.39 ± 0.09	23.16 ± 0.10	0.154
J1629+4233	16h29m4812	+42°33'385	25.77 ± 0.09	23.72 ± 0.04	23.55 ± 0.08	1.000
J2236+0006	22h36m1242	+00°06'326	26.27 ± 0.18	24.77 ± 0.13	24.27 ± 0.18	0.000
J2239−0048	22h39m5343	−00°48'026	24.23 ± 0.08	23.00 ± 0.04	22.55 ± 0.05	0.000
J2248+0103	22h48m3469	+01°03'122	26.73 ± 0.38	23.63 ± 0.04	22.93 ± 0.04	0.997
J2253−0117	22h53m5780	−01°17'053	25.65 ± 0.21	23.01 ± 0.08	22.42 ± 0.04	0.347
J2305−0051	23h05m0575	−00°51'326	25.71 ± 0.13	23.58 ± 0.09	23.47 ± 0.10	0.719
J2315−0041	23h15m1406	−00°41'011	25.69 ± 0.13	22.72 ± 0.02	21.63 ± 0.02	0.195

Note. Coordinates are at J2000.0. We took magnitudes from the latest HSC-SSP data release, and recalculated P_Q^B for objects selected with the older data releases (this is why the quasar J0220−0432 has ${P}_{{\rm{Q}}}^{{\rm{B}}}=0.000$, which used to be higher in the old data release; see the text). Magnitude upper limits are placed at 5σ significance.

Download table as:  ASCIITypeset images: 1 2

We identified 31 new quasars at 5.8 < z ≤ 6.9, as displayed in Figures 1–4 and listed in the first section of Table 5. The two highest-z quasars, J2210+0304 and J0213−0626 at z = 6.9 and z = 6.72, respectively, are observed as complete z-band dropouts on the HSC images. J0923+0402 has a very red z − y color, but is clearly visible in the z-band. All the remaining quasars are i-band dropouts. It is worth mentioning that we discovered two quasars at z ∼ 6.5 (J0921+0007 and J1545+4232), where quasars have similar optical colors to Galactic brown dwarfs and are thus hard to identify (see Figure 1 of Paper II). These two quasars have very strong Lyα + N v λ1240 lines, which allowed us to separate them from brown dwarfs due to their unusually blue z − y colors.

The majority of the objects in Figures 1–4 exhibit characteristic spectral features of high-z quasars, namely strong and broad Lyα and in some cases N v λ 1240, blue rest-UV continua, and sharp continuum breaks just blueward of Lyα. However, for some objects, the quality and wavelength coverage of our spectra are not sufficient to provide robust classification. In Papers I and II, we reported objects with luminous and narrow Lyα emission, whose quasar/galaxy classification is still ambiguous. Our new quasar sample also includes two such objects, namely J0220−0432 and J1209−0006. Following Papers I and II, we classify these objects as possible quasars, due to their high Lyα luminosity (L > 10⁴³ erg s⁻¹). This is based on the fact that, at z ∼ 2, the majority of Lyα emitters with L > 10⁴³ erg s⁻¹ are associated with AGNs, based on their X-ray, UV, and radio properties (Konno et al. 2016).

We estimated the redshifts of the discovered quasars from the Lyα lines, assuming that the observed line peaks correspond to the intrinsic Lyα wavelength (1216 Å in the rest-frame). This assumption is not always correct, due to the strong H i absorption from the IGM. It is more difficult to measure redshifts of quasars without clear Lyα lines, as is the case for J2210+0304, J0923+0402, and several other quasars in our sample. In these cases we estimated rough redshifts from the wavelengths of the onset of the Gunn–Peterson trough. Therefore, the redshifts presented here (Table 5) are accompanied by relatively large uncertainties (up to Δz ∼ 0.1), and must be interpreted with caution. Future follow-up observations at other wavelengths, e.g., near-IR covering Mg ii λ2800 or sub-mm covering [C ii] 158 μm, are crucial for secure redshift measurements.

The absolute magnitude M₁₄₅₀ and Lyα line properties were measured in the same way as described in Paper II. For every object, we defined a continuum window at wavelengths relatively free from strong sky emission lines, and extrapolated the measured continuum flux to estimate M₁₄₅₀. A power-law continuum model with a slope α = −1.5 (F_λ ∝ λ^α; e.g., Vanden Berk et al. 2001) was assumed. Since the continuum windows (falling in the range of λ_rest = 1250–1350 Å) are close to λ_rest = 1450 Å, these measurements are not sensitive to the exact value of α. The Lyα properties (luminosity, FWHM, and rest-frame equivalent width [EW]) of an object with weak continuum emission, such as J1146−0005, were measured with a local continuum defined as an average flux of all the pixels redward of Lyα. For the remaining objects with relatively strong continuum, we measured the properties of the broad Lyα + N v complex, with a local continuum defined by the above power-law model. We did not assume any line profile models, but used the continuum-subtracted flux counts directly to measure these line properties. Due to the difficulty in defining accurate continuum levels, these line measurements should be regarded as only approximate. The resultant line properties are summarized in Table 5. More detailed descriptions of the above procedure may be found in Paper II.

The 10 objects without a broad or luminous (>10⁴³ erg s⁻¹) Lyα line, as presented in Figure 5, are most likely galaxies at z ∼ 6. We have now spectroscopically identified 24 such high-z luminous galaxies, when combined with the similar objects presented in the previous papers. The redshifts of these objects were estimated from the observed positions of Lyα emission, the interstellar absorption lines of Si ii λ1260, Si ii λ1304, C ii λ1335, and/or the continuum break caused by the IGM H i absorption. This is not always easy with our limited S/N, hence the redshifts reported here must be regarded as only approximate (with uncertainties up to Δz ∼ 0.1). The absolute magnitude M₁₄₅₀ and Lyα properties were measured in the same way as for quasars, except that we assumed a continuum slope of β = −2.0 (F_λ ∝ λ^β; Stanway et al. 2005).

This paper also adds four [O iii] emitters at z ∼ 0.8 (see Figures 6) to the two similar objects reported in Paper II. We measured the line properties of Hγ, Hβ, [O iii] λ4959 and λ5007, and listed the results in Table 5. Since they have very weak continua, we estimated the continuum levels by summing up all the available pixels except for the above emission lines. This still gives relatively large uncertainties in the EWs for some of the objects. As we discussed in Paper II, their very high [O iii] λ5007/Hβ ratios may indicate that these are galaxies with sub-solar metallicity and high ionization state of the interstellar medium (e.g., Kewley et al. 2016), and/or AGN contribution. J1000+0211 and J0845−0123 have extremely large [O iii] λ5007 EWs (≳5000 Å), which even exceeds those found in the so-called "green pea" galaxies characterized by strong [O iii] λ5007 (EW ≲ 1000 Å; Cardamone et al. 2009). These objects could be an interesting subject for future follow-up studies.

Finally, we found 28 new Galactic cool dwarfs (low-mass stars and brown dwarfs), as presented in Figures 7–9. We estimated their rough spectral classes by fitting the spectral standard templates of M4- to T8-type dwarfs, taken from the SpeX Prism Spectral Library (Burgasser 2014; Skrzypek et al. 2015), to the observed spectra at λ = 7500–9800 Å. The results are summarized in Table 6 and plotted in the figures. Due to the low S/N and limited wavelength coverage of the spectra, the classification presented here is accompanied by large uncertainties, and should be regarded as only approximate.

In Figure 10, we plot the HSC i_AB − z_AB and z_AB − y_AB colors of all the spectroscopically identified objects in Papers I and II and this work. We also include the 10 previously known quasars recovered by SHELLQs. The figure demonstrates that the discovered quasars broadly follow the expected colors from our SED model, while there are outliers with very blue z_AB − y_AB colors, due to exceptionally large Lyα EWs. On the other hand, we are finding many cool dwarfs that are bluer than our fiducial models, due to the nature of our selection criteria. It may be worth noting a clump of brown dwarfs at (i_AB − z_AB, z_AB − y_AB) ∼ (3, 1); they lie exactly on the quasar SED track in the color space, and are thus assigned with high ${P}_{{\rm{Q}}}^{{\rm{B}}}$ values as candidates of quasars at z ∼ 6.5. The z ∼ 6 galaxies occupy colors in between quasars and cool dwarfs, due to their weaker Lyα emission than quasars and bluer continua than cool dwarfs. Finally, the six [O iii] emitters are well separated from the quasars on this diagram, and do not satisfy our latest selection condition of i_AB − z_AB > 2.0.

Zoom In Zoom Out Reset image size

Figure 11 displays the redshifts versus the absolute magnitudes M₁₄₅₀ of the SHELLQs quasars and galaxies, along with those of high-z quasars discovered by other surveys. We continue to probe a new parameter space at z > 5.7 and M₁₄₅₀ > −25 mag, to which other surveys have only limited sensitivities.

Zoom In Zoom Out Reset image size

Figure 12 presents a histogram of the Bayesian quasar probability (${P}_{{\rm{Q}}}^{{\rm{B}}}$), for all the spectroscopically identified objects in Papers I and II and this work. The ${P}_{{\rm{Q}}}^{{\rm{B}}}$ values have a clear bimodal distribution, with the higher peak at ${P}_{{\rm{Q}}}^{{\rm{B}}}\sim 1.0$ being dominated by high-z quasars. All but one quasar have ${P}_{{\rm{Q}}}^{{\rm{B}}}\gt 0.8$, which implies that their spectral diversity is reasonably covered by the quasar model in our Bayesian algorithm. The lower peak at ${P}_{{\rm{Q}}}^{{\rm{B}}}\sim 0.0$ is populated mostly by cool dwarfs. Many of these dwarfs lie below the quasar selection threshold (${P}_{{\rm{Q}}}^{{\rm{B}}}=0.1$), due to the improvement of HSC photometry with updates of the data reduction pipeline; they were selected for spectroscopic identification because they had ${P}_{{\rm{Q}}}^{{\rm{B}}}\gt 0.1$ in the older data releases. This is also the case for the quasar J0220−0432, which had z_AB − y_AB ∼ 0.0 and P_Q^B = 1.0 previously, but happens to have z_AB − y_AB ∼ 0.4 and ${P}_{{\rm{Q}}}^{{\rm{B}}}=0.0$ in the latest data release. Because no additional z- or y-band data were taken for this object since the previous data release, and because this object is located within a few arcseconds of a brighter galaxy, this discrepancy in color measurements may be due to different deblender treatment in the different versions of the HSC pipeline.

Zoom In Zoom Out Reset image size

This paper is the fourth in a series presenting the results from the SHELLQs project, a search for low-luminosity quasars at z ≳ 6 based on the deep multi-band imaging data produced by the HSC-SSP survey. We continue to use the quasar selection procedure we described in Papers I and II, and here report spectroscopy of additional objects that roughly double the number of identifications compared to previous papers. Through the SHELLQs project, we have so far identified 137 extremely red HSC sources over about 650 deg², which include 64 high-z quasars, 24 high-z luminous galaxies, 6 [O iii] emitters at z ∼ 0.8, and 43 Galactic cool dwarfs. Our discovery now exceeds, in number, the final SDSS sample of 52 high-z quasars (Jiang et al. 2016). The new quasars span luminosities from M₁₄₅₀ ∼ −26 to −22 mag, and continue to probe luminosities a few magnitudes lower than those probed by previous wide-field surveys. Our companion paper will present the quasar luminosity function established over an unprecedentedly wide range of M₁₄₅₀ ∼ −28 to −21 mag, using the SHELLQs and other survey outcomes.

Our project will continue to identify high-z quasars in the HSC data, as the SSP survey continues toward its goal of observing 1400 deg² in the Wide layer, and 27 and 3.5 deg² in the Deep and UltraDeep layers, respectively. At the same time, we are carrying out multi-wavelength follow-up observations of the discovered objects. Black hole masses are being measured with Mg ii λ2800 lines obtained with near-IR spectrographs on Subaru, the Gemini telescopes, and the Very Large Telescope (M. Onoue et al. 2018, in preparation). We are also using ALMA, in order to probe the stellar and gaseous properties of the host galaxies (partly published in Izumi et al. 2018). The results of these observations will be presented in forthcoming papers.

This work is based on data collected at the Subaru Telescope, which is operated by the National Astronomical Observatory of Japan (NAOJ). We appreciate the staff members of the telescope for their support during our FOCAS observations. The data analysis was in part carried out on the open use data analysis computer system at the Astronomy Data Center of NAOJ.

This work is also based on observations made with the Gran Telescopio Canarias (GTC), installed at the Spanish Observatorio del Roque de los Muchachos of the Instituto de Astrofísica de Canarias, on the island of La Palma. We thank Stefan Geier and other support astronomers for their help during preparation and execution of our observing program.

Y.M. was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI grant No. JP17H04830.

The Hyper Suprime-Cam (HSC) collaboration includes the astronomical communities of Japan and Taiwan, and Princeton University. The HSC instrumentation and software were developed by NAOJ, the Kavli Institute for the Physics and Mathematics of the Universe (Kavli IPMU), the University of Tokyo, the High Energy Accelerator Research Organization (KEK), the Academia Sinica Institute for Astronomy and Astrophysics in Taiwan (ASIAA), and Princeton University. Funding was contributed by the FIRST program from Japanese Cabinet Office, the Ministry of Education, Culture, Sports, Science and Technology (MEXT), the Japan Society for the Promotion of Science (JSPS), Japan Science and Technology Agency (JST), the Toray Science Foundation, NAOJ, Kavli IPMU, KEK, ASIAA, and Princeton University.

This paper makes use of software developed for the Large Synoptic Survey Telescope (LSST). We thank the LSST Project for making their code available as free software at http://dm.lsst.org.

The Pan-STARRS1 Surveys (PS1) have been made possible through contributions of the Institute for Astronomy, the University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, the Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, the University of Edinburgh, Queen's University Belfast, the 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 No. AST-1238877, the University of Maryland, Eötvös Lorand University (ELTE), and the Los Alamos National Laboratory.

IRAF is distributed by the National Optical Astronomy Observatory, which is operated by the Association of Universities for Research in Astronomy (AURA) under a cooperative agreement with the National Science Foundation.