Eddington-Limited Accretion in z~2 WISE-selected Hot, Dust-Obscured Galaxies
Jingwen Wu, Hyunsung D. Jun, Roberto J. Assef, Chao-Wei Tsai, Edward L. Wright, Peter R. M. Eisenhardt, Andrew Blain, Daniel Stern, Tanio Díaz-Santos, Kelly D. Denney, Brian T. Hayden, Saul Perlmutter, Greg Aldering, Kyle Boone, Parker Fagrelius
DDraft version December 7, 2017
Preprint typeset using L A TEX style emulateapj v. 01/23/15
EDDINGTON-LIMITED ACCRETION IN z ∼ Jingwen Wu , Hyunsung D. Jun , Roberto J. Assef , Chao-Wei Tsai , Edward L. Wright , Peter R. M.Eisenhardt , Andrew Blain , Daniel Stern , Tanio D´ıaz-Santos , Kelly D. Denney , Brian T. Hayden , SaulPerlmutter , Greg Aldering , Kyle Boone , Parker Fagrelius Draft version December 7, 2017
ABSTRACTHot, Dust-Obscured Galaxies, or “Hot DOGs”, are a rare, dusty, hyperluminous galaxy populationdiscovered by the WISE mission. Predominantly at redshifts 2-3, they include the most luminousknown galaxies in the universe. Their high luminosities likely come from accretion onto highly obscuredsuper massive black holes (SMBHs). We have conducted a pilot survey to measure the SMBH masses offive z ∼ α emission lines, using Keck/MOSFIRE and Gemini/FLAMINGOS-2. We detect broad H α emission in all five Hot DOGs. We find substantial corresponding SMBHmasses for these Hot DOGs ( ∼ M (cid:12) ), and their derived Eddington ratios are close to unity. These z ∼ z ∼ Subject headings: galaxies: high-redshift — infrared: galaxies — galaxies: ISM — galaxies: evo-lution — quasars: supermassive black holes — galaxies: individual (WISEJ033851.33+194128.6, WISE J090439.84+394715.2, WISE J113634.29+423602.9,WISE J213655.74 − INTRODUCTION
Gravitational accretion onto super massive black holes(SMBHs) is one of the major energy production sourcesin galaxies, powering active galactic nuclei (AGN). Itsscale and intensity (generally described as the quasar lu-minosity normalized by the Eddington luminosity, lin-early proportional to the mass of the black hole, i.e. the“Eddington ratio”) is thought to be the most importantparameter to govern AGN properties (e.g., Trump et al.2011, Shen & Ho 2014). High Eddington ratios (as wellas super Eddington accretion) have only been reportedin a small fraction of unobscured quasars (e.g., Shen etal. 2008, 2011; Jun & Im 2013). The average Edding-ton ratio for all SDSS quasars is 0.26 ± National Astronomical Observatories, Chinese Academy ofSciences, 20A Datun Road, Chaoyang District, Beijing, 100012,China; [email protected] Department of Physics and Astronomy, University of Cali-fornia, Los Angeles, CA 90095, USA Jet Propulsion Laboratory, California Institute of Technol-ogy, 4800 Oak Grove Dr., Pasadena, CA 91109, USA School of Physics, Korea Institute for Advanced Study, 85Hoegiro, Dongdaemun-gu, Seoul 02455, Korea Universidad Diego Portales, Av Republica 180, Santiago,Regi´on Metropolitana, Chile Department of Physics and Astronomy, University of Leices-ter, LE1 7RH Leicester, UK Department of Astronomy, The Ohio State University, 140West 18th Avenue, Columbus, OH 43210, USA Illumination Works, LLC, 5550 Blazer Pkwy, Dublin, OH,43017, USA E.O. Lawrence Berkeley National Lab, 1 Cyclotron Road.,Berkeley, CA, 94720, USA Department of Physics, University of California Berkeley,Berkeley, CA 94720, USA at very high redshifts ( z ∼ ∼ z = 6 − ∼ Infrared Astronomical Satellite ( IRAS ,Neugebauer et al. 1984) in the local universe, then ex- a r X i v : . [ a s t r o - ph . GA ] D ec tended to the more distant universe (especially to themost rapid evolution epoch at z = 2 −
3) by later in-frared space missions and ground-based submillimeterfacilities. Milestones include the discovery of the sub-millimeter galaxies (SMGs, see Blain et al. 2002 andCasey et al. 2014 for a review) in 850 µ m and 1 mm sur-veys, and the discovery of dust-obscured galaxies (DOGs,Dey et al. 2008) by Spitzer µ m surveys (e.g., Rigby etal. 2004, Donley et al. 2007, Yan et al. 2007, Farrahet al. 2008, Lonsdale et al. 2009). Both populationsare at z ∼ Herschel SpaceTelescope (Pilbratt et al. 2010), as well as the South PoleTelescope (Carlstrom et al. 2011) and the Atacama Cos-mology Telescope (Swetz et al. 2011) have extended thearea surveyed at submillimeter wavelengths to hundredsof square degrees, highlighting dusty star forming galax-ies to even higher redshifts (e.g., Riechers et al. 2013,Dowell et al. 2014), although a large fraction of the mostluminous galaxies are lensed (e.g., Negrello et al. 2010,Vieira et al. 2013, Bussmann et al. 2015). The WISEmission (Wright et al. 2010) enabled all-sky selection ofdusty, infrared galaxies, and is more efficient in select-ing red quasars than previous surveys (e.g. Glikman etal. 2007), either by WISE colors alone (e.g., Stern et al.2012, Assef et al. 2013), or combined with other UV tonear-IR surveys (e.g., Banerji et al. 2013, 2015, Hainlineet al. 2014).Benefiting from its all-sky coverage, WISE is able toidentify the most extreme, dusty, highly obscured AGNsin the universe. One of the most successful examples isthe discovery of the hyperluminous, hot, dust-obscuredgalaxies (Hot DOGs, Eisenhardt et al. 2012, Wu et al.2012). These galaxies were selected by looking for ob-jects strongly detected by WISE at 12 and/or 22 µ m,but only faintly or not at all at 3.4 and 4.6 µ m, i.e.“W1W2drop-out” galaxies (Eisenhardt et al. 2012). Theselected galaxies lie between redshifts 1.5 and 4.6, theirdistribution peaking at z = 2 − L (cid:12) , and they donot show evidence of lensing (Wu et al. 2014, Tsai etal. 2015). The most luminous 10% even exceed 10 L (cid:12) ,comparable to the most luminous quasars known (Assefet al. 2015), and include the single most luminous galaxyor quasar on record so far, the Hot DOG WISE J2246-0526 at z = 4 .
593 (Tsai et al. 2015, Diaz-Santos et al.2016). This discovery achieved one of the primary WISEscience goals: finding the most luminous galaxies in theuniverse.Having a consistent spectral energy distribution (SED)shape (Wu et al. 2012, Tsai et al. 2015, Tsai et al. inprep), which is characterized by an unusually high mid-IR to submm ratio, Hot DOGs are quite different from most other well-studied IR luminous populations: theycontain greater proportions of hot dust, with a charac-teristic dust temperature of 60-100 K (Wu et al. 2012,Bridge et al. 2013, Fan et al. 2016a), significantly hot-ter than the typical 30-40 K observed in SMGs or DOGs(e.g., Melbourne et al. 2012, Magnelli et al. 2012). Theyare much brighter and rarer than DOGs (there are about0.03 deg − vs. 300 deg − ). Although not required, forselection using WISE colors, they all satisfy the DOGselection criteria (Dey et al. 2008) but are much hotter;therefore we have dubbed them as ’Hot DOGs’ (Wu etal. 2012).Hot DOGs likely host very powerful AGNs, as indi-cated by their high dust temperatures and SEDs, andthe AGN dominates their luminosities (Eisenhardt et al.2012, Assef et al. 2015, Tsai et al. 2015, Fan et al.2016a, Farrah et al. 2017). These AGNs have very highextinction, typically A V ∼
20 and up to A V ∼
60 (As-sef et al. 2015), and are close to Compton-thick in theX-ray bands (Stern et al. 2014, Piconcelli et al. 2015,Assef et al. 2016, Vito et al. in prep.). They are likelyexperiencing very strong feedback: a significant numberof such galaxies present extended Lyman- α blob (LABs)structures extended over 10’s of kpc (Bridge et al. 2013).A recent high spatial resolution [CII] observation withALMA of the most luminous Hot DOG W2246-0526 re-veals an extended, uniform, highly turbulent ISM, in-dicative of an isotropically expelling galaxy-scale event(D´ıaz-Santos et al. 2016).Their high luminosities, hot dust temperatures, andstrong feedback suggest Hot DOGs may be in transitionbetween the obscured and unobscured phases of lumi-nous quasars, when the surrounding dust and gas arebeing heated and blown out, just before visible quasarsemerge (Wu et al. 2012, Bridge et al. 2013, Assef et al.2015, D´ıaz-Santos et al. 2016, Fan et al. 2016b). Giventhe high luminosities for Hot DOGs, the inferred BHmass must be well above the local BH mass-host galaxycorrelation if a typical AGN accretion rate is assumed,or, alternatively, the Eddington ratio must be very high,even above the Eddington limit (Assef et al. 2015, Tsaiet al. 2015). Potentially both factors may be present.Therefore, a measurement of their BH masses is a keystep to understand Hot DOGs.Here we present a pilot survey to measure the BH massof five Hot DOGs at z ∼ α line redshifted into the near-IR. A descriptionof the observations and data reduction is given in Section2, and in Section 3 we fit the line width of the spectra.The estimated BH masses, luminosities, and Eddingtonratios are presented in Section 4. In Section 5, we com-pare Hot DOGs to other galaxy populations at z ∼
2, aswell as to quasars at z ∼
6. Conclusions are summarizedin Section 6. Throughout this paper, we assume a ΛCDMcosmology with H = 70 km s − Mpc − , Ω m = 0 .
3, andΩ Λ = 0 . OBSERVATIONS AND DATA REDUCTION
We observed five Hot DOGs with secure redshifts andwell sampled SEDs. All selected targets have redshifts z ∼ α line is observable with near-IR spec-troscopy. Except for requiring coordinates that madethem accessible during the observing runs, no other se-lection criteria were applied when choosing these targets.Their extinctions have been calculated from a modelof their rest-frame UV-to-mid-IR SEDs that includes astarburst, evolved stars, and reddened AGN components(Assef et al. 2015). The derived extinctions range from A V ∼ Keck/MOSFIRE
We obtained near-IR spectroscopy for fourHot DOGs (WISE J033851.33+194128.6, WISEJ090439.84+394715.2, WISE J113634.29+423602.9 andWISE J213655.74 − − K -band spectra forthree Hot DOGs and H -band spectra for two Hot DOGswere obtained during three runs in 2014 and 2015 (onetarget was observed in both bands). The MOSFIRE K -band filter is centered at 2.162 µ m with a full-widthat half-maximum (FWHM) of 0.483 µ m. The H -bandfilter is centered at 1.637 µ m with a FWHM of 0.341 µ m.The data were acquired using masks with a slit widthof 0.7 (cid:48)(cid:48) , giving a velocity resolution of ∆ v ∼
80 km s − ( R ∼ − ∼ (cid:48)(cid:48) for W2136-1631, and 1 (cid:48)(cid:48) -1.2 (cid:48)(cid:48) for the rest ofthe targets.The data were reduced using the MOSFIRE data re-duction pipeline (DRP), which performs flat fielding,sky subtraction and wavelength calibration, and outputsrectified two-dimensional (2D) spectra, from which one-dimensional spectra were extracted. Telluric correctionand flux calibration were performed using the spectra ofA0 standard stars. Gemini/FLAMINGOS-2
WISE J221648.05+072353.6 (hereafter W2216+0723)was observed using FLAMINGOS-2 (Eikenberry et al.2012) at the Gemini-South Observatory on the night ofUT 2014 November 7. It was observed simultaneouslyin the J and H bands using the JH grating with the 2pixel-wide (0.36 (cid:48)(cid:48) ) longslit, providing a spectral resolv-ing power of approximately R ∼ IRAF toolswith the sky emission lines used for wavelength calibra-tion. The star HIP106817 was used for telluric and fluxcalibration using the XTellCorr General routine of the
Spextool package (Vacca, Cushing & Rayner 2003). LINE WIDTH FITTING
We detected broad H α lines in all five Hot DOGs. Wemodeled the lines using the IDL routine MPFIT (Mark-wardt 2009). The aim of the fitting is to measure theline width of the H α line originating from the BLR whichis broadened by virial motion, in order to estimate theblack hole mass (see section 4.1). There are two majorconcerns that can affect the fitting result significantly:one is the blending with emission lines from the narrowline regions (NLRs), the other is any contribution fromoutflows. We discuss them separately next. Fitting spectra with broad lines and narrow lines
The broad H α line is blended with narrow H α and[NII] λ α , and narrow lines H α , [NII] λ λ λ λ α and [NII]doublets at relatively low SNR. We fit the rest-frame6000-7000 ˚A spectra including a power-law continuumand Gaussian lines, where all the components are lim-ited to have non-negative fluxes and reported line widthsare corrected for instrumental resolution. The modeledspectra are plotted in Figure 1. We iterated the fit onceto include just the inner 98.76% of the data sorted inabsolute value of the residual (2.5 σ Gaussian rejection),and minimized χ ν over the 6000-7000 ˚A range.In this Section, we focus on the question of whetherthe broad line and/or narrow line components are nec-essary to fit the near-IR spectra. We explored threestrategies to select the best parametrization to fit thespectra. In strategy 1, we only fit one broad Gaussian(FWHM > − ) for H α , assuming the narrowH α and [NII] are negligible compared to the broad H α (“1B alone” strategy hereafter).In strategy 2, we fit a single narrow Gaussian (FWHM < − ) in addition to a single broad Gaussian toH α (“1B+1N” hereafter). We kept all narrow line widthsthe same and allowed the broad component’s center torange within ± − from the narrow H α redshift(e.g., Bonning et al. 2007, Shen et al. 2011), determinedfrom the peak of the narrow line model fit.In strategy 3, we allow two broad Gaussians and onenarrow Gaussian to fit H α (“2B+1N” hereafter), withthe reported FWHM of H α derived by the combinationof the two broad Gaussians. The second broad compo-nent is assumed to have the same redshift as the nar-row lines. Combining multiple Gaussians to obtain thebroad FWHM has been used in many other works (e.g.,Greene & Ho 2005, Assef et al. 2011, Jun et al. 2017).We present the fitted spectra in Figure 1. The fittingparameters including FWHM, the errors from MPFIT, http://iraf.noao.edu χ ν (cid:48) ( χ ν calculated only covering H α wavelengths: 6450-6650 ˚A), and the degrees of freedom (DOF) of the bestfit model for each strategy are listed in Table 2.To determine which strategy works best, we appliedan F-test to each spectrum over rest-frame 6450-6650 ˚A.The results are listed in Table 3. The values in Table 3show the probability p that the χ ν of the fit with thelarger number of components was consistent with be-ing drawn from the same distribution as the fit withthe smaller number of components. Generally, when p < .
05, we can assume that an extra component isnecessary in the fitting. Note that when p = 0, it meansthat the number is so small that it is below the numericalprecision of the integrator we used.Based on Table 2 and Table 3, the option 1 “1B alone”strategy is ruled out by the F-test analysis, for all fivetargets, implying the narrow line component is not neg-ligible. [A “1 narrow line alone” (1N alone) case vs. “1B+ 1N” is also rejected by the F-test]. Both broad lineand narrow line components are necessary to representthe spectra.As seen in Figure 1, W0338+1941 and W0904+3947have poorer signal to noise ratios (SNRs) than the otherthree targets. According to the F-test, the 2B+1N modelis strongly preferred in two of the three high SNR tar-gets. Since all Hot DOGs are arguably physically similar,we assumed that a consistent model should apply to alltargets. In this paper we use the 2B+1N model to fit allsources. For the one high SNR Hot DOG (W2216+0723)where the 2B+1N model is not preferred over the 1B+1Nmodel by the F-Test, the resulting FWHMs and impliedBH masses for the two models are consistent (see Table2).We adopt the FWHM and χ ν from the 2B+1N modelin MPFIT as best-fit values for each target, then we es-timate the error of this best-fit FWHM using a MonteCarlo approach similar to that of Assef et al. (2011).For a given spectrum, we first scale the uncertainty ofeach pixel such that the χ ν of the best fit model is equalto 1. We then create 1,000 simulated spectra with thesame pixel size as the observed spectrum. The value ofeach pixel in the simulated spectra is randomly drawnfrom a Gaussian distribution centered at the flux of thebest-fit model in that pixel, with a dispersion equal toits scaled uncertainty. We then fit each of the simulatedspectra in the same way as described above. The 68.3%confidence interval of the FWHM is obtained from thedistribution of the best-fit FWHM to the simulated spec-tra. Specifically, the range contains 68.3% of the FWHMvalues below and above the median of the distribution.This uncertainty is listed in Table 4 as the asymmetric 1 σ error of the FWHM. Can outflows explain the broad H α line? Massive outflows have been discovered in some high-redshift, dust-obscured quasars, revealed by the high ve-locity dispersion of forbidden lines (e.g., Liu et al. 2013a,2013b; Zakamska et al. 2016), sometimes accompaniedby blue-shifted wings. These features have been ex-plained as a manifestation of strong AGN feedback (e.g.,Spoon & Holt 2009; Mullaney et al. 2013; Zakamska &Greene 2014; Brusa et al. 2015). In this section, we testif the broad H α lines detected in these Hot DOGs can be explained by outflow combined with emission fromNLRs.We obtained H -band spectroscopy for W1136+4236,and detected broad [OIII] λ J -band spectroscopy for W2216+0723, with broad[OII] λ λ − . For W2216+0723, we fit a single Gaussian to the[OII] λ − . Both these forbidden lines are very broad, andpresent blue-shifted wing features, implying there is out-flow in both targets.The broad H α we observe in these Hot DOGs can ei-ther be caused by virial motion in BLRs right aroundthe massive blackholes, or from more extended NLRscombined with outflows. We made the following teststo see which scenario is preferred by the data. In test1, we fit a single Gaussian line profile to both the oxy-gen and H α lines, fixing the linewidth to be the sameas [OIII] λ λ χ ν s are notably worse than the corresponding mod-els in the 1B+1N and/or 2B+1N models shown in Fig-ure 1. In addition, the resulting line profiles of the [OI]and [SII] lines in this model are too wide to fit thedata. In test 2, we used the asymmetric line profile of[OII] λ α in W2216+0723 (a similar fitto the [OIII] λ χ ν .In the discussion above, we assume the outflow occursin the NLRs. This is a plausible scenario to test if outflowfrom NLRs can cause the observed H α line. Although wecannot fully exclude the possibility that outflows con-tribute to the observed broad H α , our analysis showsthe 1B+1N and 2B+1N models, where the H α lines arebroadened by the virial motion from BLRs, provide abetter fit to our data than the two test models, wherethe H α profiles are broadened by NLR outflows.Even if strong NLR outflows dominate the observedFWHM, preventing us from constraining the BH massand Eddington ratios in some Hot DOGs, this is stillconsistent with the picture that Hot DOGs mark a tran-sitional stage in AGN evolution. The detection of verybroad [OII] and [OIII] lines is intriguing, as the intenseoutflows this suggests also imply high accretion rates(e.g. Zakamska et al. 2016). Both are expected for the“blow-out” phase between the obscured and unobscuredquasar stages. While we do not consider BLR outflows inthis discussion, we note that strong BLR outflows wouldlikely imply that the virialized BLR component is nar-rower, which would translate into lower BH masses and,hence, higher, true Eddington ratios. BLACK HOLE MASSES, LUMINOSITIES, ANDEDDINGTON RATIOS
Black hole masses
Fig. 1.— K -band spectroscopy for three Hot DOGs and H -band spectroscopy for two Hot DOGs with model fits. We binned the spectrato their spectral resolution ( ∼ ∼ λ α λ λ λ Fig. 2.—
Upper panels: H - and K -band spectra for W1136+4236 with model fits. We use Gaussians to fit the H β and [OIII] λ H -band spectrum (upper left). In the K -band spectrum (upper right), we use Gaussians to fit [OI] λ α λ λ H -band. Lower panels: J - and H -band spectroscopy forW2216+0723 with model fits. We fit one Gaussian for the [OII] λ α in the H -band (lower right). All fitted lines are shown in thick black lines, data are plotted in dark grey, and errors are presented inlight grey. The residual spectra are shown in the panels below the spectral panels. Vertical lines below the spectra show the positions ofthe H β λ λ λ α λ λ λ Black hole mass can be estimated by assuming virialequilibrium in the broad line region: M BH ∝ v R BLR / G ,where R BLR is the radius of the BLR, and G is the grav-itational constant. The BLR radius can be measuredthrough the reverberation mapping (RM) technique (e.g.,Blandford & McKee 1982, Peterson 1993), which esti-mates the radius of the BLR from the lag between thevariability in the AGN continuum and the correspondingvariability in the broad permitted lines. This method hasbeen successful in measuring black hole masses in the lo-cal universe (e.g., Peterson et al. 2004, Vestergaard &Peterson 2006). For higher redshift galaxies, BH massmeasurements rely on an empirical relation found be-tween the BLR size and the AGN luminosity (the R − L relation) discovered and calibrated by RM studies (i.e. R BLR ∝ L . , where L = λL λ at λ = 5100˚A, Kaspiet al. 2000, Bentz et al. 2006, 2009, 2013), which arecalled as “single-epoch” black hole mass measurements.Thus, by measuring rest-frame optical broad line widthand adjacent continuum luminosity, one can estimate thesingle-epoch BH mass.Typically used broad lines to calculate high-redshiftAGN BH mass include H α , H β , Mg II , and C IV , withH α and H β generally regarded as more reliable. The H β line has primarily been used to calibrate the R − L rela-tion based on RM measurements, and the H α emissionlinewidths and luminosities correlate well with those ofH β over a wide range of total AGN luminosities (e.g.,Greene & Ho 2005; Jun et al. 2015), supporting the useof H α whenever available. Since H α emission is severaltimes stronger than H β (Shen & Liu 2012; Jun et al.2015), and is less blended with broad Fe II emission thanH β , it is especially preferred for spectroscopy of fainterAGNs. At z (cid:38)
1, Balmer emission is redshifted out of theoptical window, so rest-frame UV emission (Mg II , C IV )has been used to measure BH masses. However, concernsabout the scatter of the C IV -derived BH masses com-pared to those derived from Balmer lines (e.g., Assef etal. 2011; Shen & Liu 2012), as well as strong Fe II blend-ing near Mg II , favor using H α if possible. Moreover, forheavily obscured AGNs like Hot DOGs, UV broad linesare hard to detect. The stronger line and redder wave-length of H α make it a better choice than other lines toprobe BLRs in Hot DOGs.Following Assef et al. (2011), we calculate the BH massbased on the FWHM of broad line H α : M BH = 7 . × f ( F W HM Hα km s − ) . × ( L erg s − ) . M (cid:12) , where f is a scale factor of order unity that dependson the structure, kinematics, and inclination of the BLR(e.g., Collin et al. 2006). Here we adopt the best-fitfixed f factor of 1.17 following the arguments in Assefet al. (2011). Due to the very high UV/optical extinc-tion of Hot DOGs, we can’t use direct measurementsof L . Instead, we adopt the rest-frame UV-to-mid-IR SED models for Hot DOGs reported in Assef et al.(2015), which are constrained by WISE and follow-upoptical and near-IR photometry, fitting the contributionsof a starburst, evolved stars, and reddened AGN compo-nents, and derive the obscuration-corrected AGN lumi-nosity. The derived extinctions ( A V ) from this modeland the resulting BH masses for the five Hot DOGs with broad H α line measurements are presented in Table 4.Due to the low SNR for the two H β lines detected inW1136+4236 and W2216+0723, we can’t decompose thebroad and narrow components for H β , complicating theuse of the Balmer decrement to estimate extinction. Inaddition, the possibility that the broad lines are observeddue to scattering (Assef et al. 2016) makes the Balmerdecrement method potentially problematic. If we assumethe same broad to narrow component ratio for H α andH β , the broad line region extinction for the two sourcespredicts a much lower AGN luminosity than what weestimated from the observed SED, implying the Balmerdecrement method is not a reliable way to estimate theextinction for our sample. Uncertainties of BH masses
We estimate the measurement uncertainties in ourBH mass calculations by propagating the errors in theFWHM and in the continuum luminosities. The uncer-tainty in FWHM varies from 8% to 66% due to differingdata quality. We assume a consistent 50% uncertaintyin calculating L from the SED model (Assef et al.2015). We assume a 20% uncertainty on the bolometricluminosities (Tsai et al. 2015). The resulting measure-ment uncertainties are presented in Table 4 and is foldedinto the error bars in all the Figures.Some authors (e.g., Jun et al. 2015) also consider mea-surement errors from the f factor (43% following Collinet al. 2006) and the R − L relation (7%, Bentz et al.2013). This leads to an additional 0.16 dex uncertaintyfor BH masses and 0.18 dex uncertainty for Eddingtonratios. There are also systematic errors for the constants,which is about 0.3-0.4 dex for f factor (Kormendy & Ho2013; McConnell et al. 2013) and 0.11 dex for the R − L relation (Peterson et al. 2010).The uncertainties discussed above do not include theuncertainty introduced by the possible contribution ofoutflows to the FWHM of H α . As described in Sec-tion 3.3, including an outflow component based on theobserved [OIII] and [OII] lines for W1136+4236 andW2216+0723, respectively, in modeling their broad H α line profiles leads to a poorer χ ν . To quantitatively esti-mate the contribution of outflows to FWHM values willrequire higher signal-to-noise spectra, but is not expectedto significantly affect the results presented here. Bolometric luminosities
Using extensive follow-up photometry (Griffith et al.2012, Wu et al. 2012, Jones et al. 2014, Assef et al.2015, Tsai et al. 2015, Tsai et al. in prep.), we con-structed complete SEDs for the Hot DOGs and calcu-lated their bolometric luminosities. The method we usedis described in detail in Tsai et al. (2015). In brief,we simply integrated over the detected photometric datapoints from the optical to far-IR bands, using power-lawinterpolations between measurements and extrapolatedto 20% beyond the shortest and longest wavelengths.This method is more secure than bolometric luminosi-ties simply scaled up from one or a few wavelength mea-surements using templates, but it is conservative sinceit may miss flux between data points and beyond thelongest/shortest wavelength data. If the best-fit SEDtemplates or spline-smoothed SEDs are considered, theluminosities typically increase by 20% (Tsai et al. 2015).We list the derived bolometric luminosities of the fiveHot DOGs in Table 4. The photometry of the Hot DOGs(including the five reported in this paper) observed from
Spitzer (Griffith et al. 2012), WISE and
Herschel , andmore details on the method used to calculate their bolo-metric luminosities (in which we consider that BH accre-tion is dominating the luminosity) is reported in Tsai etal. (in prep).
Eddington ratios
The Eddington luminosity is defined as L Edd = 3 . × ( M BH M (cid:12) ) L (cid:12) , and the Eddington ratio is η = L AGN /L Edd . For HotDOGs, we can attribute most of the bolometric lumi-nosity to the AGN (Eisenhardt et al. 2012; Jones et al.2014; D´ıaz-Santos et al. 2016, Farrah et al. 2017, Tsaiet al. in prep.). Hence we calculated Eddington ratiosas L bol /L Edd for the five Hot DOGs (see Table 4). Wefound that η from high SNR targets are greater than 0.5and close to unity. Considering that L bol is calculatedconservatively, we conclude that the derived Eddingtonratios are close to or above the Eddington limit for theseHot DOGs. DISCUSSION
Based on their SEDs and far-IR to mid-IR luminosityratios, Wu et al. (2012) speculated that Hot DOGs mayrepresent a transitional phase following the SMG andDOG phase and preceding the regular quasar phase. Pro-gression along such a sequence is consistent with galaxyevolution scenarios based on major mergers (e.g., Hop-kins et al. 2008) and is likely driven by the growth of thecentral SMBH.The primary objective of this study is to measure theBH masses of Hot DOGs, and to test if such an evolu-tionary sequence is reasonable by comparing their BHmasses to those of other populations. The project is alsointended to find out if the high luminosities of Hot DOGsresult from hosting black holes with masses well abovethe local BH-host galaxy relation, and/or if their SMBHsare accreting at or even above the Eddington limit. Toexplore these topics, we compare Hot DOGs to othergalaxy populations with measurements of BH masses andEddington ratios. The discussion is organized as follows:We summarize z ∼ z ∼ z ∼ Comparison samples at z ∼ z ∼ − α ,H β , Mg II , and C IV ) when they are redshifted into theSDSS spectroscopic range. We are most interested inquasars around z ∼ II rather than fromC IV over this redshift range, since Mg II is the less com-plicated line to use as a tool for measuring BH masses(e.g., Shen et al. 2008, 2011, Assef et al. 2011, Jun et al.2015).The BH mass estimate for SMGs is complicated, partlydue to the large uncertainty in determining the AGN lu-minosity in SMGs. Unlike Hot DOGs, the luminositiesof SMGs are generally dominated by starbursts ratherthan AGNs, though some SMGs are actually 850 µ m-selected dusty quasars with large extinction. In this pa-per, we simply adopt average values for general SMGs(whose luminosities are still dominated by starbursts)from Alexander et al. (2008) of log ( M BH /M (cid:12) ) ∼ . η ∼ . α measurements.Two of these (DOG 1 and DOG 4) have reliable bolo-metric (8-1000 µ m) luminosities from SED integrationincluding Herschel data (Melbourne et al. 2012). Sincethe AGN contribution is believed to dominate the SEDof power-law DOGs, we approximate their AGN lumi-nosities using their bolometric luminosities to calculateEddington ratios, as we did for Hot DOGs. Melbourneet al. (2011) note that their derived BH masses are lowerlimits due to the uncertainty in dust correction. In addi-tion, AGNs in DOGs are less dominant than those in HotDOGs, and the starburst contribution to the luminositymay not be negligible. Therefore, the derived Eddingtonratios for these DOGs are upper limits.There is growing interest in the study of red quasars,which are thought to be a linking population betweenobscured and unobscured quasars (e.g., Glikman et al.2012; Ross et al. 2015). Initially, red quasars wereselected using optical/near-IR colors (e.g., Richards etal. 2003), and some cross-matching to radio surveys(e.g., Glikman et al. 2007, 2015). With the large area,deep mid-IR imaging surveys available from
Spitzer andWISE, additional mid-IR color cuts have been found to
Fig. 3.—
Black hole masses and Eddington ratios for the five HotDOGs reported in this paper, compared to typical values for z ∼ f factor and the R − L relations are considered. be very efficient in selecting red quasars (e.g., Banerjiet al. 2013, 2015, Ross et al. 2015, Hamann et al.2017). These red quasars normally have moderate ex-tinctions ( E ( B − V ) < . A V ∼ −
6) at z ∼ z ∼
2, and BH mass measurements obtained frombroad Balmer lines, except for the Type-1 SDSS quasarsfrom Shen et al. (2011). Figure 3 shows that the HotDOGs reported here have higher BH masses than generalSMGs or DOGs, based on the limited information avail-able, and are comparable to quasars. The Hot DOGsshow systematically higher Eddington ratios than otherpopulations at z ∼
2, with values close to unity. Ourpilot project supports the idea that Hot DOGs are atransitional, high accreting phase between obscured andunobscured quasars.
Biases due to source selection and line fitting
Possible selection effects for the pilot sample
Our pilot project to calculate BH masses and Edding-ton ratios is based on five Hot DOGs. We did not imposeany selection criteria apart from redshift, but the small
Fig. 4.—
The distribution of redshifts and luminosities of thefive Hot DOGs reported in this paper (red stars), compared tothe sample of Hot DOGs with
Herschel measurements (black dots)presented in Figure 1 of Tsai et al. (2015). sample size raises the question of how well they representthe BH masses and Eddington ratios for the whole HotDOG population.In Figure 4 we compare the distribution of redshiftsand luminosities of the five Hot DOGs in this paper toall Hot DOGs with redshifts and
Herschel measurements.Their luminosities are near or somewhat below the me-dian values of Hot DOGs at similar redshifts, and arelower than most Hot DOGs. Hence the Eddington ratiosof these five Hot DOGs may be somewhat lower than fortypical Hot DOGs with similar BH masses, and it is un-likely there is a selection bias for a high-Eddington ratiosubsample because of higher luminosity.Compared with the obscuration values reported in As-sef et al. (2015), two of the five targets are close to thelower limit of extinction values, two are close to the me-dian value, and one is close to the upper limit. Overall,there is no obvious bias in extinction.Although we did not select targets to have broad UVlines, two (W0904+3947 and W1136+4236) out of thefive targets do show broad rest-UV spectra features,which may suggest the direct detection of the BLR inthe rest-UV. This 40% ratio is somewhat higher thanthe 12%-16% fraction of Hot DOGs in general with broadrest-UV lines (Eisenhardt et al. in prep). The relation-ship between the detection of BLR in rest-optical andrest-UV bands is not clear, but if we assume a higherchance of detecting broad H α for sources with broadrest-UV lines, our pilot sample may be biased againstHot DOGs with narrower or less obvious broad UV-lines.But for a fixed AGN continuum, this means a bias againstsmaller FWHMs, lower BH masses, and higher Edding-ton ratios.In summary, the possible selection effect of the five HotDOGs in this paper, including luminosities, obscurations,and existing rest-UV spectra, either have little influence,or suggest even higher Eddington ratios for typical HotDOGs. These considerations are unlikely to change ourconclusion that Hot DOGs represent a high-Eddingtonratio population. Influence of different line fitting approaches
The SNR of our five spectra vary significantly. Ourline fitting strategy of using a consistent multi-Gaussian0fit takes advantage of the high SNR spectra, and our ap-proach of using spectral information as a prior in Monte-Carlo simulation with an F-test investigation gives a ro-bust and sensitive way to deal with the low SNR spectraand variations between models for some targets. How-ever, some authors prefer to use a simpler approach forlower SNR spectra, namely selecting the simplest Gaus-sian model with a small enough χ . How would our re-sults change if we were to take this simpler approach?From Table 2 and Figure 1, the simplest Gaussianmodel with a reasonable χ would select the 1B model forall targets except for W2136-1631, for which the 1B+1Nmodel is required. If these models are selected, two tar-gets (W0338+1941, W2136-1631) have larger BH massesand lower Eddington ratios, and the other three targetshave lower BH masses and higher Eddington ratios. Therange of Eddington ratios hanges from 0.25 -1.46 to 0.45- 1.31, with the median value changing from 0.74 to 0.88.The overall effect is to move the derived Eddington ra-tios of Hot DOGs towards unity. Again, this does notchange our major conclusion that Hot DOGs are a high-Eddington ratio population. Why are Hot DOGs so luminous: comparing toSDSS quasars
In the previous section, we showed that Hot DOGstend to have higher Eddington ratios than other IR lu-minous (active) galaxy populations at z ∼
2. Here weargue that an unusually high accretion rate is character-istic of Hot DOGs, and is not a selection effect. In Figure5, we compare the black hole masses and bolometric lu-minosities for Hot DOGs to SDSS quasars at all redshifts( z = 0 −
5) from Shen et al. (2011), which is the largestand least biased quasar sample with BH masses in theliterature. We also include a lower redshift ( z = 1 . z = 1 . − . z ∼
6) quasars,as we discuss in Section 5.5.
A quasar accretion history at z ∼ z ∼
2, connecting obscured and unobscuredquasars. This scenario is consistent with the expecta-tion of a popular galaxy model in which AGN feedbacksweeps out the surrounding material that is also the fuelfor BH accretion (e.g., Hopkins et al. 2008, Somervilleet al. 2008). We should expect to see less obscuredquasars associated with larger SMBHs, and with fallingaccretion rates and fading quasar luminosities. A similarevolutionary connection was also proposed in Assef et al.(2015), based on the luminosity function of Hot DOGsand quasars, that Hot DOGs can be the progenitors ofmore massive type-1 quasars, in the case that they areexperiencing enhanced BH accretion.We can test these predictions using our comparisonsamples. Although we do not know how long each evo-lutionary stage lasts during this evolution, we can takea snapshot of the z ∼ . < z < .
5, which includes the Hot DOGs inthis paper. The peak epoch of quasar density and ma-jor merger activities is z ∼
2, and observational data atthese redshifts is relatively rich. We focus on BH massesderived from Balmer lines if possible, which are observ-able from ground-based near-IR spectroscopy at theseredshifts. For the SDSS quasars that only have opticalspectra, we choose BH masses measured from Mg II in-stead of C IV, as discussed in Section 5.1. This sets aupper limit of z = 2 .
25 for Mg II to be in the SDSS DR7spectra. Therefore we have a slightly smaller redshiftrange of 1 . < z < .
25 for the SDSS quasar sample,which includes 27761 quasars.The BH masses and Eddington ratios of Hot DOGsand other comparison samples are plotted in Figure 7.Hot DOGs have among the highest Eddington ratios ofall quasars with similar black hole masses, much greaterthan SMGs or DOGs. Red quasars in Banerji et al.(2012, 2015) have overlapping but generally lower Ed-dington ratios than Hot DOGs, but higher Eddingtonratios than SDSS quasars, given similar BH masses. Theoverall average Eddington ratio is 0.37 ± z ∼ Comparison to z ∼ quasars Fig. 5.—
Comparison of BH masses and bolometric luminosities of Hot DOGs and SDSS quasars at z < z ∼ z ∼ f factor and the R − L relations. Fig. 6.—
Comparison of BH masses and bolometric luminositiesof Hot DOGs and SDSS quasars (Shen et al. 2011) in the redshiftrange (1 . < z < . The consistently high Eddington ratios of z ∼ z ∼ − L (cid:12) ), and comparable black holemasses to z ∼ M Sph ) relation in Figure 7 of their paper,assuming a fixed Eddington ratio of 0.3 that is typicalfor SDSS quasars. They estimated the stellar massesof Hot DOGs by multiplying the rest-frame luminosityof the host component in the K -band by the mass tolight ( M/L ) ratio in that band, which depends on manyparameters, including star formation history, metallicity,stellar initial mass function, and the contribution of ther-mally pulsating AGB stars. They used maximum
M/L ratios for these parameters to estimate the upper limitsof the stellar masses. Their results suggested Hot DOGsmay lie well above the local M BH − M Sph relation. Basedon the present work, we can reasonably change the fixedEddington ratio to unity, and update the M BH − M Sph relation for Hot DOGs, as presented in Figure 9. The2
Fig. 7.—
Hot DOGs may trace the highest Eddington ratio stage before the red quasar (Banerji et al. 2012, 2015) and SDSS quasarphases (Shen et al. 2011) at redshift z ∼
2. The likely range of BH mass and Eddington ratios for general SMGs (Alexander et al. 2008)and the limits for two DOGs (Melbourne et al. 2011, 2012, Shen et al. 2011) are also marked. A z ∼ z ∼ f factor and the R − L relations are considered. Fig. 8.—
Eddington ratio and extinction for z ∼ Hot DOG stellar masses shown in the plot are again up-per limits. We mark the three Hot DOGs with highSNR H α spectra as red stars in Figure 9, using the BH We also note here that Figure 9 corrects an error in Assef et al.(2015). Assef et al. (2015) stated they used an IMF from Conroyet al. (2013) with a higher
M/L ratio in the K -band than the M/L for a Chabrier (2003) IMF, but mistakenly used the Chabrier(2013) IMF in their Figure 7. Hence the stellar masses of HotDOGs shown in Figure 9 in this paper are about a factor of 2higher than those shown in Figure 7 of Assef et al. (2015). masses from this work. After these updates, Hot DOGsare closer to the local M BH − M Sph relation, though thesestellar mass estimates are upper limits and their pointsmay move to the left on Figure 9.Wang et al. (2010) have estimated the BH-host massrelation for z ∼ z ∼ M BH − M Sph plot, suggestingthey may be hosted by similar kinds of galaxies. The keyelements for AGN systems are their black holes masses,host galaxies, and accretion rates. It seems Hot DOGsand z ∼ z ∼ z ∼ z = 2 − z ∼ z ∼ z ∼ Fig. 9.— M BH and M Sph values, with Hot DOG data taken fromAssef et al. (2015), but assuming a fixed Eddington ratio of 1.0based on the present work. The high SNR Hot DOG detectionsfrom this work are marked with red stars. The bulge masses ofhost galaxies are constrained by using the best-fit SED templatemodel of Assef et al. (2011, 2015), which are upper limits. A localrelation of active galaxies determined by Bennert et al. (2011), aswell as the values for quasars at z ∼ z ∼ z ∼ . might have been close to unity (e.g., Willott et al. 2010,Johnson et al. 2012). This is thought to be fundamen-tally different from lower redshift SDSS quasars, whoseSMBHs have passed their peak accretion, and whose hostgalaxy gas either will not cool or has been cleared outby AGN feedback (e.g., Di Matteo et al. 2008). Thusthe duty cycle of lower redshift quasars is much lower.However, we know little about the immediate surround-ing environment of SMBHs at z ∼
6, except that thehost galaxies are dusty. It is possible that SMBHs inHot DOGs accrete in a similar way to z ∼ z ∼ z ∼ z ∼ z ∼ z ∼ A Hot DOG stage at different cosmic epochs?
We believe that Hot DOGs are heavily obscured AGNsat a special evolutionary stage, characterized by high lu-minosity due to high BH accretion rates, and likely withstrong AGN feedback. Current work focuses on the stud-ies of z ∼ − z ∼ z ∼ z ∼ − z ∼ Spitzer , and
Herschel may help to reveal moredusty, higher obscured z ∼ z < z = 1 .
009 (Ricci et al. 2017), whose SEDmatches well with Hot DOGs, and whose BH mass andEddington ratio agree with the trends we find in z ∼ z = 0 .
184 (e.g., Reeves et al. 2000, 2003,Nardini et al. 2015), with a comparable BH mass andluminosity to the z ∼ SUMMARY
4A population of hyperluminous, dusty galaxies hasbeen discovered by WISE, which we call “hot, dust-obscured galaxies” or “Hot DOGs”. Their extreme lu-minosities and hot dust temperatures suggest they ei-ther host very massive black holes well above the localBH mass-stellar mass relation, or are accreting at veryhigh rates. We have conducted a pilot survey to mea-sure the BH masses of five Hot DOGs at z ∼
2, usingMOSFIRE at Keck and FLAMINGOS-2 at Gemini. Theprimary results from this study are summarized below:1. Broad H α lines were detected in all five targets.Spectral fits imply they are broadened by BLRs aroundSMBHs. We estimate their BH masses to be ∼ M (cid:12) ,and their Eddington ratios are close to unity.2. The BH masses are greater than those of typicalSMGs and DOGs, and comparable to those of unob-scured quasars. This is consistent with the model whereHot DOGs represent a transitional stage between ob-scured and unobscured quasars. Although not preferredby our spectral fitting, even if strong outflows contributeto the broad line width of H α , the implied strong feed-back still supports Hot DOGs’ role in the overall evolu-tionary picture.3. Hot DOGs have high luminosities compared toquasars with similar black hole masses, which impliesthey are accreting at the highest possible rates for theirSMBH masses, i.e. they have the highest Eddington ra-tios observed for quasars and SMBHs.4. Our results are consistent with a “Hot DOG-redquasar-optical quasar” evolutionary sequence.5. Hot DOGs and z ∼ M BH − M Sph rela-tions. Their SMBHs both accrete at the maximum ob-served rates, close to the Eddington limit, making themthe most luminous persistent objects in their own cosmicepochs.JW is supported by the National Key Program forScience and Technology Research and Development ofChina (grant 2016YFA0400702) and Project 11673029supported by NSFC. HDJ is supported by Basic ScienceResearch Program through the National Research Foun- dation of Korea (NRF) funded by the Ministry of Ed-ucation (NRF-2017R1A6A3A04005158). RJA was sup-ported by FONDECYT grant number 1151408. T.D.-S.acknowledges support from ALMA-CONICYT project31130005 and FONDECYT regular project 1151239.This material is based upon work supported by the Na-tional Aeronautics and Space Administration under Pro-posal No. 13-ADAP13-0092 issued through the Astro-physics Data Analysis Program. This publication makesuse of data obtained at the W.M. Keck Observatory,which is operated as a scientific partnership among Cal-tech, the University of California and NASA. The KeckObservatory was made possible by the generous finan-cial support of the W.M. Keck Foundation. The au-thors wish to recognize and acknowledge the very sig-nificant cultural role and reverence that the summit ofMauna Kea has always had within the indigenous Hawai-ian community. We are most fortunate to have theopportunity to conduct observations from this moun-tain. This research was partly based on observationsobtained at the Gemini Observatory, which is operatedby the Association of Universities for Research in As-tronomy, Inc., under a cooperative agreement with theNSF on behalf of the Gemini partnership: the NationalScience Foundation (United States), the National Re-search Council (Canada), CONICYT (Chile), Ministe-rio de Ciencia, Tecnolog´ıa e Innovaci´on Productiva (Ar-gentina), and Minist´erio da Ciˆencia, Tecnologia e In-ova¸c˜ao (Brazil). This work uses data products fromthe Wide-field Infrared Survey Explorer, which is a jointproject of the University of California, Los Angeles, andthe Jet Propulsion Laboratory/California Institute ofTechnology, funded by the National Aeronautics andSpace Administration. This work uses data obtainedfrom the
Spitzer Space Telescope , which is operated bythe Jet Propulsion Laboratory, California Institute ofTechnology under contract with NASA. This work usesdata from
Herschel . Herschel is an ESA space observa-tory with science instruments provided by European-ledPrincipal Investigator consortia and with important par-ticipation from NASA.
REFERENCESAbazajian, K. N., Adelman-McCarthy, J. K., Ag¨ueros, Marcel A.Alexander, D. M., Brandt, W. N., Smail, I., et al. 2008, AJ, 135,1968Antonucci, R. R. J., & Miller, J. S. 1985, ApJ, 297, 621Antonucci, R. 1993, Ann. Rev. Ast. & Astrophys., 31, 473Assef, R. J., Kochanek, C. S., Ashby, M. L. N. et al. 2011, ApJ,728, 56Assef, R. J., Stern, D., Kochanek, C. S., et al. 2013, ApJ, 772, 26Assef, R. J., Eisenhardt, P. R. M., Stern, D., et al. 2015, ApJ,804, 27Assef, R. J., Walton, D. J., Brightman, M., et al. 2016, ApJ, 819,111Bachetti, M., Harrison, F. A., Walton, D. J. et al. 2014, Nature,514, 202Banerji, M., McMahon, R. G., Hewett, P. C., et al. 2012,MNRAS, 427, 2275Banerji, M., McMahon, R. G., Hewett, P. C., et al. 2013,MNRAS, 429, 55Banerji, M., Alaghband-Zadeh, S., Hewett, P. C., McMahon, R.G. 2015, MNRAS, 447, 3368Bennert, V. N., Auger, M. W., Treu, T., Woo, J.-H., & Malkan,M. A. 2011, ApJ, 726, 59 Bentz, M. C., Peterson, B. M., Pogge, R. W., Vestergaard, M.,Onken, C. A. 2006, ApJ, 644, 133Bentz, M. C., Peterson, B. M., Netzer, H., Pogge, R. W.,Vestergaard, M. 2009, ApJ, 697, 160Bentz, M. C., Denney, K. D., Grier, C. J. et al. 2013, ApJ, 767,149Blain, A. W., Smail, I., Ivison, R. J., Kneib, J.P., Frayer, D. T.2002, Physics Reports, 369, Issue 2, 111Blain, A. W., Assef, R., Stern, D., et al. 2013, ApJ, 778, 113Blandford, R. D. & McKee, C. F. 1982, ApJ, 255, 419Bondi, H. & Hoyle, F. 1944, MNRAS, 104, 273Borys, C., Smail, I., Chapman, S. C., et al. 2005, ApJ, 635, 853Bridge, C., Blain, A., Borys, C. et al. 2013, ApJ, 768, 91Brusa, M., Feruglio, C., Cresci, G. et al. 2015, A&A, 578, 11Bussmann, R. S., Dey, A., Borys, C., Desai, V., Jannuzi, B. T.,LeFloc´h, E., Melbourne, J., Sheth, K., Soifer, B. T. 2009, ApJ,705, 184Bussmann, R. S., Riechers, D., Fialkov, A., et al. 2015, ApJ, 812,43Carlstrom, J. E., Ade, P. A. R., Aird, K. A., et al. 2011, PASP,123, 568 Carnall, A. C., Shanks, T., Chehade, B., et al. 2015, MNRAS,451, 16Casey, C. M., Narayanan, D., Cooray, A. 2014, Physics Reports,541, 45Chapman, S. C., Blain, A., Smail, Ian, Ivison, R. J. 2005, ApJ,622, 772Chabrier, G. 2003, PASP, 115, 763Collin, S., Kawaguchi, T., Peterson, B. M., Vestergaard, M. 2006,A&A, 456, 75Conroy, C., Dutton, A. A., Graves, G. J., Mendel, J. T., vanDokkum, P. G. 2013, ApJ Letters, 776, 26Coppin, K. E. K., Swinbank, A. M., Neri, R., et al. 2008,MNRAS, 389, 45De Rosa, G, Decarli, R., Walter, F., 2011, ApJ, 739, 56Dey, A., Soifer, B. T., Desai, V., Brand, K., Le Floc´h, E., Brown,M. J. I., Jannuzi, B. T., Armus, L., Bussmann, S., Brodwin,M., Bian, C., Eisenhardt, P., Higdon, S. J., Weedman, D.,Willner, S. P. 2008, ApJ, 677, 943D´ıaz-Santos, T., Assef, R. J., Blain, A. W. et al. 2016, ApJLetters, 816, 6Dietrich, M., Hamann, F., Appenzeller, I., Vestergaard, M. 2003,ApJ, 596, 817Di Matteo, T., Colberg, J., Springel, V., Hernquist, L.,; Sijacki,D. 2008, ApJ, 676, 33Donley, J. L., Rieke, G. H., P´erez-Gonz´alez, P. G., Rigby, J. R.,Alonso-Herrero, A. 2007, ApJ, 660, 167Dowell, C. D., Conley, A., Glenn, J., et al. 2014, ApJ, 780, 75Eikenberry, S., Bandyopadhyay, R., Bennett, J. G. et al. 2012,SPIE, 8446, 0IEisenhardt, P.R.M., Wu, J., Tsai, C. et al. 2012, ApJ, 755, 173Fan, L., Han, Y., Fang, G., et al. 2016a, ApJ, 822, 32Fan, L., Han, Y., Nikutta, R. et al. 2016b, ApJ, 823, 107Fan, X., Hennawi, J. F., Richards, G. T., et al. 2004, AJ, 128, 515Fan, X., Carilli, C. L. Keating, B. 2006, Ann. Rev. Ast. &Astrophys., 44, 415Farrah, D., Lonsdale, C. J., Weedman, D. W., Spoon, H. W. W.,Rowan-Robinson, M., Polletta, M., Oliver, S., Houck, J. R.,Smith, H. E. 2008, ApJ, 677, 957Farrah, D., Petty, S et al., Connolly, B. et al. 2017, ApJ, 844, 106Feng, H., Soria, R. 2011, New Astronomy Reviews, 55, 166F¨urst, F., Walton, D. J., Harrison, F. A. et al. 2016, ApJ, 831, 14Glikman, E., Helfand, D. J., White, R. L., et al. 2007, ApJ, 667,673Glikman, E., Urrutia, T., Lacy, M., et al. 2012, ApJ, 757, 51Glikman, E., Simmons, B., Mailly, M., et al. 2015, ApJ, 806, 218Greene, J. E., Ho, L. C. 2005, ApJ, 630, 122Greene, J. E., Alexandroff, R., Strauss, M. A., et al. 2014, ApJ,788, 91Griffith, R. L., Kirkpatrick, J. D., Eisenhardt, P. R. M., et al.2012, AJ, 144, 148Hainline, K. N., Hickox, R. C., Carroll, C. M., et al. 2014, ApJ,795, 124Hamann, F., Zakamska, N. L., Ross, N., et al. 2017, MNRAS,464, 3431Hopkins, P F., Hernquist, L., Cox, T. J., Di Matteo, T.,Robertson, B., Springel, V. 2006, ApJS, 163, 1Hopkins, P. F., Hernquist, L., Cox, T. J., Keres, D. 2008, ApJS,175, 356Hopkins, P. F, Hernquist, L. 2010, MNRAS, 402, 985Israel, G. L., Papitto, A., Esposito, P., Stella, L. et al. 2017a,MNRAS, 466, 48Israel, G. L., Belfiore, A., Stella, L. et al. 2017b, Science, 355, 817Jakobsen, P., Jansen, R. A., Wagner, S., Reimers, D. 2003, A&A,397, 891Jiang, L., Fan, X., Vestergaard, M., et al. 2007, AJ, 134, 1150Jiang, L., Fan, X., Brandt. W. N., et al. 2010, Nature, 464, 380Jones, S. F., Blain, A. W., Stern, D., et al. 2014, MNRAS, 443,146Jun, H. D., Im, M. 2013, ApJ, 779, 104Jun, H. D., Im, M., Lee, H. M., et al. 2015, ApJ, 806, 109Jun, H. D., Im, M., Kim, D., Stern , D. 2017, ApJ, 838, 41Kaspi, S. Smith, P. S., Netzer, H. et al. 2000, ApJ, 533, 631Kleinmann, D. E. & Low, F. J. 1970, ApJ, 159, 165Kormendy, J., & Ho, L. C. 2013, Ann. Rev. Ast. & Astrophys.,51, 511 Leipski, C., Meisenheimer, K., Walter, F. et al. 2014, ApJ, 785,154Liu, G., Zakamska, N. L., Greene, J. E., Nesvadba, N. P. H., Liu,X. 2013, MNRAS, 436, 2576Liu, G., Zakamska, N. L., Greene, J. E. 2014, MNRAS, 442, 1303Lonsdale, C. J. et al. 2009, ApJ, 692, 442Makishima, K., Kubota, A., Mizuno, T. et al. 2000, ApJ, 535, 632bibitem[Magnelli et al. (2012)]Magnelli2012 Magnelli, B., Lutz,D. Santini, P. et al. 2012, A&A, 539, 155bibitem[Markwardt (2009)]Markwardt09 Markwardt, C. B.2009, ASPC, 411, 251McConnell, N. J., & Ma, C.-P. 2013, ApJ, 764, 184Melbourne, J., Peng, C. Y., Soifer, B. T. et al. 2011, AJ, 141, 141Melbourne, J., Soifer, B., Desai, V. et al. 2012, AJ, 143, 125Miller, M. C. & Colbert, E. J. M., 2004, International Journal ofModern Physics, 13, 1Mullaney, J. R., Alexander, D. M., Fine, S., Goulding, A. D.,Harrison, C. M., et al. 2013, MNRAS, 433, 622Narayanan, D., Dey, A., Hayward, C. C., Cox, T.J., Bussmann,R. S., Brodwin, M., Jonsson, P., Hopkins, P. F., et al. 2010,MNRAS, 407, 1701Nardini, E., Reeves, J. N., Gofford, J., et al. 2015, Science, 347,860Negrello, M., Hopwood, R., De Zotti, G., et al. 2010, Science, 330,800Neugebauer, G., Habing, H. J., van Duinen, R., et al. 1984, ApJLetters, 278, 1Ohsuga, K., Mineshige, S., Mori, M., Umemura, M. 2002, ApJ,574, 315Ohsuga, K., Mori, M., Nakamoto, T., Mineshige, S. 2005, ApJ,628, 368Peterson, B. M. 1993, PASP, 105, 24Peterson, B. M., Ferrarese, L., Gilbert, K. M., et al. ApJ, 613, 682Peterson, B. M. 2010, in IAU Symp. 267, Co-Evolution of CentralBlack Holes and Galaxies, ed. B. Peterson, R. Somerville, & T.Storchi-Bergmann (Cambridge: Cambridge Univ. Press), 151Piconcelli, E., Vignali, C., Bianchi, S., et al. 2015, A&A, 574, 9Pilbratt, G. L., Riedinger, J. R., Passvogel, T., et al. 2010, A&A,518, 1Reeves, J. N., O’Brien, P. T., Vaughan, S., et al. 2000, MNRAS,312, 17Reeves, J. N., O’Brien, P. T., Ward, M. J. 2003, ApJ, 593, 65Ricci, C., Assef, R. J., Stern, D., et al. 2017, ApJ, 835, 105Richards, G. T., Hall, P. B., Vanden Berk, D. E. et al. 2003, AJ,126, 1131Riechers, Dominik A.; Bradford, C. M.; Clements, D. L. et al.2013, Nature, 496, 329Rieke, G. H. & Low, F. J. 1972, 1972, ApJ Letters, 176, 95Rigby, J. R., Rieke, G. H., Maiolino, R. et al. 2004, ApJS, 154,160.Ross, N. P., Hamann, F., Zakamska, N. L., et al. 2015, MNRAS,453, 3932Schneider, D. P., Richards, G. T., Hall, P. B., et al. 2010, AJ,139, 2360Shen, Y., Greene, J. E., Strauss, M. A., Richards, G. T.,Schneider, D. P. ApJ, 680, 169Shen, Y., Richards, G. T., Strauss, M. A. et al. 2011, ApJ, 194, 45Shen, Y. & Liu, X. 2012, ApJ, 753, 125Shen, Y., Ho, L. C. 2014, Nature, 513, 210Somerville, R. S., Hopkins, P. F., Cox, T. J., Robertson, B. E.Hernquist, L. 2008, MNRAS, 391, 481Spoon, H. W. W. & Holt, J. 2009, ApJ Letters, 702, 42Stern, D., Assef, R. J., Benford, D. J., et al. 2012, ApJ, 753, 30Stern, D., Lansbury, G. B., Assef, R. J., et al. 2014, ApJ, 794, 102Swetz, D. S., Ade, P. A. R., Amiri, M., et al. 2011, ApJS, 194, 41Trump, J. R., Impey, C.D., Kelly, B. C. 2011, ApJ, 733, 60Tsai, C., Eisenhardt, P. R. M., Wu, J., et al. 2015, ApJ, 805, 90Vacca, W. D., Cushing, M. C., Rayner, J. T. 2003, PASP, 115, 389Vestergaard, M. & Peterson, B. M. 2006, ApJ, 641, 689Vieira, J. D., M. D. P.; Chapman, S. C., et al. 2013, Nature, 495,344Volonteri, M., Capelo, P. R., Netzer, H., et al. 2015, MNRAS,452, 6Wang, F., Wu, X., Fan, X., et al. 2015, ApJ Letters, 807, 9Wang, R., Carilli, C. L., Wagg, J., et al. 2008, ApJ, 687, 848Wang, R., Carilli, C. L., Neri, R. et al., 2010, ApJ, 714, 699 TABLE 1Target list
Source R.A. Dec. Instrument Filter UT Date Int time (min)W0338+1941 03:38:51.33 +19:41:28.6 Keck/MOSFIRE K K K , H − − H J , H TABLE 2Fitting H α with models − ) a χ ν (cid:48) b DOF c FWHM (km s − ) a χ ν (cid:48) b DOF c FWHM (km s − ) a χ ν (cid:48) b DOF c W0338+1941 3087(197) 1.766 271 4761(434) 1.438 268 1742(320) 1.422 266W0904+3947 2070(106) 1.289 271 4516(652) 1.069 268 4516(652) 1.074 268W1136+4236 2278(37) 3.442 303 6275(321) 2.618 300 3291(319) 2.509 298W2136 − a Numbers in parentheses are 1 σ uncertainties. b χ ν (cid:48) are calculated within H α wavelengths (6450-6650 ˚A); they are different from χ ν marked in Figure 2 that are calculated over the entire fittingregion. c Degrees of freedom in fitting.
TABLE 3F-Test result between different spectral fitting models
Source 1N vs 1B+1N 1B vs 1B+1N 1B+1N vs 2B+1NW0338+1941 3.3e-6 2.8e-7 0.24W0904+3947 3.9e-5 3.7e-9 1.00W1136+4236 1.8e-9 1.0e-12 8.8e-3W2136 − TABLE 4Derived parameters of Hot DOGs