An Extremely Bright QSO at z = 2.89

The epoch from reionization to the era of peak star formation ($2\lesssim z\lesssim 7$; Madau & Dickinson 2014) is a time of rapid galaxy growth and assembly (Bell 2004), as well as commensurate growth of supermassive black holes (SMBHs) in galaxies. Indeed, one of the outstanding challenges in galaxy formation is explaining the existence of the most massive SMBHs at $z\gtrsim 7$ (Volonteri 2012; Bañados et al. 2018), which are observed as quasi-stellar objects (QSOs). More generally, evolution of the bright end of the quasar luminosity function (QLF) provides an observational benchmark which must be reproduced by theoretical models (Yu & Tremaine 2002).

Beyond evolution of the QLF, ultraluminous QSOs are also of interest as physical probes of the universe. Bright, lensed QSOs can be used for time delay measurements to constrain the Hubble constant H₀, as first described by Refsdal (1964). Bright, unlensed QSOs can serve as backlights for studying the Lyα forest (Harris et al. 2016) and as tools for detecting cosmological redshift drift, as described in Loeb (1998). The latter of these is an extension of a test first proposed by Sandage (1962), in which one measures the change in the expansion velocity of the universe via the Lyα forest absorption lines superposed on the QSO spectrum. Liske et al. (2008) demonstrated that with a small sample of bright QSOs and a 30 m class telescope, direct measurements of acceleration in the expansion are possible. Ground-based Lyα forest studies and cosmological redshift drift require QSOs at z > 2 so that Lyα lies in the optical window.

Significant progress has been made in recent years in quantifying the bright end of the QLF (Wisotzki 2000; Richards et al. 2006; Ross et al. 2013; Schindler et al. 2019b), including development of machine-learning techniques and utilization of wide-area optical surveys like Pan-STARRS (Schindler et al. 2018, 2019a, 2019b). This census remains incomplete, as until recently all-sky optical data was unavailable and discrimination of QSOs from faint stars can be challenging. The combination of all-sky data from the Gaia and Wide-field Infrared Survey Explorer (WISE) missions offers a uniquely powerful means of completing this census. Our team has initiated a search to identify the brightest z > 2 quasars using these data sets. In this work, we report on the discovery of the first optically bright, ultraluminous QSO identified in this search. For pure luminosity evolution, the QLF is best described by a broken double power law consisting of a faint-end slope, a bright-end slope, the break magnitude between these slopes, and the overall density normalization (Boyle et al. 1988, 2000; Pei 1995). As shown in McGreer et al. (2013), the break magnitude evolves strongly in the redshift range 2.8 < z < 4.5, affecting both the bright-end slope and density normalization (Schindler et al. 2019b). Discoveries from high-z QSO searches such as the one presented here will provide additional constraints on the bright-end slope and break magnitude.

2.1. Data

2.2. Method

We used the Double Spectrograph (DBSP; Oke & Gunn 1982) at the 200 inch Hale telescope at Palomar Observatory to observe 2MASS J13260399 + 7023462 on the night of UT 2018 June 6. We obtained a single 300 s exposure using DBSP on a night with good seeing conditions but with a thin cloud observed at sunrise. A 15 wide slit was used with a 5500 Å dichroic beam splitter, a 600 ℓ mm⁻¹ grating in the blue channel (${\lambda }_{\mathrm{blaze}}=4000\,\mathring{\rm A} ;$ spectral resolving power $R\equiv \lambda /{\rm{\Delta }}\lambda \sim 1200$), and a 316 ℓ mm⁻¹ grating in the red channel (${\lambda }_{\mathrm{blaze}}=7500\,\mathring{\rm A} ;$ spectral resolving power $R\sim 1800$). The data were processed using standard routines found in the IRAF software package and observations were calibrated with Feige 34 (white dwarf) and BD+28 4211 (subdwarf O-star). The observed spectra are shown in Figure 1. The presence of the Lyα peak and forest along with C iv and C iii] broad emission lines confirms that this object is a QSO.

Zoom In Zoom Out Reset image size

To determine the redshift, we fit a Gaussian profile to the C iii] feature in the spectrum and compare it to its known rest-frame wavelength (1909 Å). We avoid using the Lyα line for this purpose as the multitude of absorption features bluewards of it prevent us from obtaining a reliable fit. We also avoid using the C iv feature as it is often blueshifted due to outflows (Richards et al. 2011; Netzer 2015). We determine the redshift solely from the observed C iii] broad line, taking care to avoid the telluric absorption feature on the red side of the line, and obtain z = 2.889 ± 0.002. Using this redshift, we calculate the blueshift of the C iv line to be $1740\pm 50\,\mathrm{km}\,{{\rm{s}}}^{-1}$ in the rest frame of the system after fitting a Gaussian profile, the FWHM of which is $8000\pm 150\,\mathrm{km}\,{{\rm{s}}}^{-1}$; the uncertainty of these measurements is determined by varying the wavelength range over which the line profile is fit. The line fit and blueshifting are shown in Figure 2.

Zoom In Zoom Out Reset image size

Using the zero-points specified in the Spanish Virtual Observatory (SVO) Filter Profile Service (Rodrigo et al. 2012), we calculate fluxes for each of the passbands listed in Table 1. We then redshift the composite template and scale it to match Two Micron All Sky Survey (2MASS) J-band photometry, the bluest passband at which the spectrum is devoid of strong emission lines, and overplot the fluxes to create the spectral energy distribution (SED) shown in Figure 3.

Zoom In Zoom Out Reset image size

Table 1.  Summary Information about 2MASS J13260399 + 7023462

Positional Info
DR2 Source ID	1685441172355283328
R.A. (deg)	201.5166465
Decl. (deg)	70.3961983
bgal (deg)	118.7285419
lgal (deg)	46.4484052
Photometric Info
GALEX FUV	Undetected
GALEX NUV	Undetected
Gaia G	16.070 ± 0.001
Gaia GBP	16.348 ± 0.005
Gaia GRP	15.643 ± 0.004
Pan-STARRS g	16.393 ± 0.013
Pan-STARRS r	15.997 ± 0.008
Pan-STARRS i	15.625 ± 0.009
Pan-STARRS z	15.563 ± 0.006
Pan-STARRS y	15.375 ± 0.016
2MASS J	15.089 ± 0.050
2MASS H	14.489 ± 0.050
2MASS Ks	14.009 ± 0.049
WISE W1	13.187 ± 0.024
WISE W2	12.217 ± 0.021
WISE W3	8.527 ± 0.021
WISE W4	6.364 ± 0.051
Miscellaneous Info
z	2.889 ± 0.002
DL	24.76 Gpc
${m}_{1450,\mathrm{AB}}$	16.6 ± 0.2
${M}_{1450,\mathrm{AB}}$	−28.9 ± 0.2
${M}_{\mathrm{bol}}$	−33.0 ± 0.2
Lbol	(1.2 ± 0.2)×1015${L}_{\odot }$
MBH	(2.7 ± 0.4) × 1010 ${M}_{\odot }$
${\lambda }_{\mathrm{Edd}.}$	1.3 ± 0.3

Note. All magnitudes reported in the photometric information section are in the Vega system. The luminosity distance and absolute magnitude at 1450 Å were calculated using the parameters discussed in Section 4.

Download table as:  ASCIITypeset image

Vanden Berk et al. (2004) show that for a QSO observed in the Sloan Digital Sky Survey (SDSS) r-band (the filter corresponding to rest-frame 1450 Å at z = 2.889), brightness fluctuations on the order of ∼0.2 magnitudes can be expected. In addition to this, MacLeod et al. (2012) show that the average characteristic timescale for QSO variability in the rest frame is approximately 2 yr—this is of particular relevance given that the most recent 2MASS, AllWISE, and Gaia data sets were released in the years 2003, 2013, and 2018, respectively.

To investigate the variability of 2MASS J13260399 + 7023462, we compile and plot single epoch Pan-STARRS data, the results of which we show in Figure 4. The light curve exhibits variation on a scale consistent with what is suggested by Vanden Berk et al. (2004), lending credibility to the hypothesis that some portion of the deviations from the composite are a direct consequence of the variability itself. The multi-year Pan-STARRS light curve shows a net increase in brightness from the start of observations, with the g and y-band light curves in particular showing a prominent brightening of ∼0.2 magnitudes. The light curve shows variations that far exceed the photometric uncertainties provided in Table 1. We note that a brightness increase in one set of passbands does not necessitate a similar brightness increase in other passbands. Furthermore, the dust content of any one QSO is not guaranteed to be similar to the dust content the composite template assumes.

Zoom In Zoom Out Reset image size

Although we do not utilize variability information in the current implementation of our search method, it can be used to distinguish QSOs from contaminants. Schmidt et al. (2010) used intrinsic variability to select quasars with high completeness and purity comparable to existing color selection methods for the redshift range 2.5 < z < 3.

A standard measure of QSO luminosity is the absolute magnitude at 1450 Å (M₁₄₅₀) due to the lack of emission features that overlap this wavelength. We estimate this quantity using the Hernán-Caballero et al. (2016) composite after scaling it to match the Gaia photometry for this QSO. We find F_ν at 5623 Å, the observed wavelength of rest-frame 1450 Å at z = 2.889, to be F_ν,1450 =  0.89 ± 0.15 mJy from which we calculate m₁₄₅₀ = 16.6 ± 0.2. Using the cosmology specified in Planck Collaboration et al. (2018) for a flat universe, the implied absolute magnitude is M₁₄₅₀ = −28.9 ± 0.2. The most luminous, unlensed QSO at this epoch, HS 1946 + 7658, is of a comparable absolute magnitude with ${M}_{1450}=-29.2$, indicating that 2MASS J13260399 + 7023462 is one of the most luminous QSOs known. Using the aforementioned ${F}_{\nu ,1450}$ and luminosity distance, we infer a monochromatic luminosity of νL_ν,1450 = (1.2 ± 0.2)×10⁴⁸ erg s⁻¹.

We next compute the bolometric luminosity for this system. We find L_bol = (4.7 ± 0.9)×10⁴⁸ erg s⁻¹, equivalent to (1.2 ± 0.2)×10¹⁵${L}_{\odot }$, assuming a bolometric correction of 3.8 at 1450 Å as in Wolf et al. (2018). This luminosity corresponds to a bolometric magnitude for this QSO of −33.0 ± 0.2. We measure the FWHM of the C iv line in the spectrum to be $8000\pm 150\,\mathrm{km}\,{{\rm{s}}}^{-1}$, from which we obtain a corrected FWHM of $6000\pm 360\,\mathrm{km}\,{{\rm{s}}}^{-1}$ using Equation (4) from Coatman et al. (2017) to correct for non-virial C iv emission. We estimate the monochromatic luminosity at 1350 Å from our scaled composite to be 1.5 × 10⁴⁸ erg s⁻¹ at 1350 Å. Using these quantities with Equation (6) from Coatman et al. (2017), we obtain a value of (2.7 ± 0.4) × 10¹⁰ ${M}_{\odot }$ for the black hole mass. For comparison, the most massive black hole currently known is TON 618 with a mass of 6.6 × 10¹⁰ ${M}_{\odot }$, indicating that the black hole powering 2MASS J13260399 + 7023462 is among the most massive currently known (Shemmer et al. 2004). Using L_bol and ${L}_{\mathrm{Edd}.}=1.3\times {10}^{38}\left(\tfrac{{M}_{\mathrm{BH}}}{{M}_{\odot }}\right)$ erg s⁻¹ from Rybicki & Lightman (1979), we determine the Eddington ratio, ${\lambda }_{\mathrm{Edd}.}={L}_{\mathrm{bol}}/{L}_{\mathrm{Edd}.}$ to be 1.3 ± 0.3, which indicates 2MASS J13260399 + 7023462 is currently undergoing accretion near its Eddington limit.

4.1. 2MASS J13260399 + 7023462 in Context

To further investigate the nature of this QSO, we compare its properties with other known luminous QSOs. With ${G}_{\mathrm{BP}}=16.348$ and G = 16.070, it is the third brightest QSO at $z\gt 2.75$ in Figure 5. As such, this QSO can serve as a valuable backlight for future Lyα forest studies. To better place this QSO in context in terms of intrinsic properties, in Figure 6 we compare it to bright QSOs at 2.8 < z < 5 from SDSS and Pan-STARRS searches from Schindler et al. (2018, 2019a, 2019b). At this epoch, 2MASS J13260399 + 7023462 currently is among the dozen most luminous known. Of those that are potentially more luminous, three are lensed QSOs (green hexagons), and hence their intrinsic luminosities are overestimated in the figure. We can further estimate the total number of brighter QSOs at this epoch (including those yet to be discovered) based on the published parameterizations of the bright end of the QLF. Using the QLF from Schindler et al. (2019b), we estimate there should exist a total of 37 QSOs brighter than ${M}_{1450,\mathrm{AB}}=-28.9$ at 2.8 < z < 4.5. It is therefore expected there remain a number of brighter QSOs to discover at this epoch.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

4.2. Lensing

Given that three of the brighter QSOs at this epoch are strongly lensed, it is relevant to consider whether 2MASS J13260399 + 7023462 is itself lensed, in which case its luminosity will be overestimated. To constrain any magnification due to lensing, we obtained Hubble Wide Field Camera 3 (WFC3) imaging (Program 15950) on 2020 January 20 in F475W within a 512 × 512 pixel subarray of UVIS2. The QSO does not resolve into multiple sources at this high resolution. To better constrain lensing, we compare the brightness profile of the QSO with the WFC3 point-spread function (PSF). We use TinyTim (Hook & Stoehr 2008; Krist et al. 2011) to generate a model PSF for WFC3/UVIS2 F475W for each individual exposure. We then scale the model PSFs to match the amplitudes in the individual WFC3 images and combine the individual images in the same way as the actual data to derive the appropriate composite PSF. In Figure 7, we show the comparison of the surface brightness profiles for 2MASS J13260399 + 7023462 and the composite TinyTim PSF. In Figure 8, we show the combined F475W image and composite TinyTim PSF.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The QSO is unresolved at Hubble Space Telescope (HST) resolution, consistent with it not being multiply imaged. We can further place an upper limit on the spatial extent by convolving the PSF with a Gaussian. A comparison of the QSO with the Gaussian-convolved PSF constrains σ < 0017 at 3σ. If we take this to be an approximate limit on the Einstein radius, ${\theta }_{{\rm{E}}}$, this corresponds to a 3σ mass limit on any lensing galaxy of $M\lt 4.0\times {10}^{8}$ ${M}_{\odot }$. For halos with M ≲ 10⁹ the cross section for galaxy lensing is so small (e.g., Hilbert et al. 2008) that the probability of lensing is negligible. We therefore can exclude the possibility that this source is significantly magnified by a foreground galaxy.

4.3. Radio and X-Ray Emission

Objects like 2MASS J13260399 + 7023462 are rare and historically have been difficult to find. The release of DR2 from ESA's Gaia mission combined with data from WISE has enabled searches for ultraluminous, high-z QSOs by allowing one to easily filter out stellar contaminants. Using a combination of parallax, proper motion, and color selection criteria, we are able to identify 2MASS J13260399 + 7023462 as a bright QSO at $z=2.889$. With G = 16.070 at $z=2.889$, it is the third brightest QSO known at $z\gt 2.75$, and one of the dozen most luminous QSOs known at $z\gt 2$.

With the inferred ${M}_{1450}=-28.9\,\pm \,0.2$ and bolometric magnitude of ${M}_{\mathrm{bol}}=-33.0\,\pm \,0.2$ in addition to its brightness relative other QSOs at this epoch, this QSO is extremely bright and follow-up observations with Hubble WFC3 demonstrate that this QSO is not lensed. The spectroscopic confirmation of the first QSO from our search suggests that the selection criteria we employ are able to identify interesting QSO candidates at z > 2. With further refinement of this method and acquisition of spectroscopic data for additional candidates, we expect to find more bright, high-z QSOs like 2MASS J13260399 + 7023462.

S.E. was supported in part by a University of Florida Research Foundation Professorship.

This research is based on observations made with the NASA/ESA Hubble Space Telescope obtained from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 526555. These observations are associated with program(s) 15950.

This publication makes use of data products from the Wide-field Infrared Survey Explorer, which is 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. This publication makes use of data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. The Pan-STARRS1 Surveys (PS1) and the PS1 public science archive have been made possible through contributions by 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, the 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 grant No. AST-1238877, the University of Maryland, Eotvos Lorand University (ELTE), the Los Alamos National Laboratory, and the Gordon and Betty Moore Foundation.

This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/Gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/Gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement.

This research has made use of the SVO Filter Profile Service (http://svo2.cab.inta-csic.es/theory/fps/) supported from the Spanish MINECO through grant AyA2014-55216. This research has made use of Astropy, a community-developed core Python package for Astronomy (The Astropy Collaboration et al. 2018).