The Contribution of Quasar Absorption Outflows to AGN Feedback
MMNRAS , 1–8 (2020) Preprint 25 September 2020 Compiled using MNRAS L A TEX style file v3.0
The Contribution of Quasar Absorption Outflows to AGNFeedback
Timothy R. Miller, Nahum Arav, (cid:63) Xinfeng Xu and Gerard A. Kriss Department of Physics, Virginia Tech, Blacksburg, VA 24061, USA Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA
Accepted XXX. Received YYY; in original form ZZZ
ABSTRACT
Determining the distance of quasar absorption outflows from the central source ( R )and their kinetic luminosity ( (cid:219) E k ) are crucial for understanding their contribution toactive galactic nucleus (AGN) feedback. Here we summarize the results for a sampleof nine luminous quasars that were observed with the Hubble Space Telescope. Wefind that the outflows in more than half of the objects are powerful enough to be themain agents for AGN feedback, and that most outflows are found at R >
100 pc. Thesample is representative of the quasar absorption outflow population as a whole andis unbiased towards specific ranges of R and (cid:219) E k . Therefore, the analysis results canbe extended to the majority of such objects, including broad absorption line quasars.We find that these results are consistent with those of another sample (seven quasars)that is also unbiased towards specific ranges of R and (cid:219) E k . Assuming that all quasarshave absorption outflows, we conclude that most luminous quasars produce outflowsthat can contribute significantly to AGN feedback. We also discuss the criterion forwhether an outflow is energetic enough to cause AGN feedback effects. Key words: galaxies: active — galaxies: kinematics and dynamics — ISM: jets andoutflows — quasars: absorption lines — quasars: general
Quasar spectra show outflowing material along the line ofsight that is propagating from the centers of quasars asblueshifted absorption troughs relative to the rest frame ofthe host quasar. Upward of 40% (Hewett & Foltz 2003; Daiet al. 2008; Ganguly & Brotherton 2008; Knigge et al. 2008)of the quasar population contains absorption outflows. Theseoutflows are candidates for producing major feedback pro-cesses within active galactic nuclei (AGNs), which includerestricting the host galaxy growth (e.g., Ciotti et al. 2009;Hopkins et al. 2009; Faucher-Gigu`ere et al. 2012; Zubovas,& King 2014; Schaye et al. 2015; Choi et al. 2017; Peirani etal. 2017; Valentini et al. 2020), explaining the mass correla-tion between the central black hole and the galaxy’s bulge(e.g. Silk & Rees 1998; Blandford & Begelman 2004; Hop-kins et al. 2009; Ostriker et al. 2010; Dubois et al. 2014;Rosas-Guevara et al. 2015; Volonteri et al. 2016; Angl´es-Alc´azar et al. 2017; Yuan et al. 2018; Nomura et al. 2020),and chemical enrichment of the intracluster and intergalac-tic medium (ICM, IGM; e.g., Scannapieco & Oh 2004; Kha-latyan et al. 2008; Tornatore et al. 2010; Barai et al. 2011; (cid:63)
E-mail: [email protected]
Taylor & Kobayashi 2015; Thompson et al. 2015; Barai etal. 2018).Quasar outflows come in many flavors: molecular (e.g.,Krips et al. 2011; Cicone et al. 2014; Feruglio et al. 2015),atomic (e.g., Kanekar & Chengalur 2008; Aditya & Kanekar2018; Morganti & Oosterloo 2018), ionized seen in emission(e.g., Rupke & Veilleux 2013; Harrison et al. 2014; Bischettiet al. 2017), ionized seen in absorption (e.g., Korista et al.2008; Dunn et al. 2010; Shen et al. 2011) and ionized seen inX-ray (e.g., Chartas et al. 2009; Nardini et al. 2015; Tombesiet al. 2015) spectra. A good comparison between the differ-ent outflow manifestations is given in Fiore et al. (2017). Seealso Figure 1 here adapted from that paper.Our aim in this paper is to determine whether absorp-tion outflows seen in the rest-frame UV of luminous quasars(with bolometric luminosity L bol > erg s − ) are capa-ble of producing AGN feedback effects. The potential quasaroutflows have to produce the above mentioned feedback pro-cesses is directly determined by their kinetic luminosity ( (cid:219) E k )(e.g., Scannapieco & Oh 2004; Hopkins & Elvis 2010). (cid:219) E k isproportional to both the distance from the central source( R ) and the total outflowing column density ( N H ). There-fore, we need to find a sample of outflows where both R and © a r X i v : . [ a s t r o - ph . GA ] S e p Miller et al. N H can be measured and where the sample is representativeof the majority of objects showing absorption outflows.Measuring N H : Ionization equilibrium in quasar out-flows is dominated by photoionization, where the outflowis characterized by its ionization parameter ( U H ) and N H .Usually, the spectral synthesis code Cloudy (Ferland et al.2017) is used to produce photoionization simulations (e.g.,Arav et al. 2013) for a range of U H and N H values. Thesesimulations predict the column density of each ion ( N ion ) inthe outflow. The best U H and N H solution is the one wherethe predicted N ion values best fit the measured N ion from theabsorption troughs of the outflow.Measuring R : The most robust way to determine R forquasar absorption outflows is by using troughs from ionicexcited states. The column density ratio between the excitedand resonance states yields the electron number density ofthe outflow. Combined with a knowledge of the outflowˆa ˘A´Zs U H , R can be determined (e.g., Arav et al. 2018). Threesubtypes of absorption outflows are used to determine R :1. The majority of R determinations arise from singlyionized species (mainly from Fe ii and Si ii , e.g., de Koolet al. 2001; Hamann et al. 2001; Moe et al. 2009; Aoki etal. 2011; Lucy et al. 2014; Choi et al. 2020). However, mostoutflows show absorption troughs only from more highly ion-ized species. Therefore, the applicability of R derived fromsingly ionized species to the majority of outflows is model-dependent (see discussion in Section 1 of Dunn et al. 2010).Furthermore, this is a heterogeneous sample where many ofthe objects were selected for unique features and/or indica-tions that R and/or (cid:219) E k would have large values. Therefore,this is not a representative sample even for the low ionizationoutflows.2. For ground-based observations, the main high-ionization species with a measurable trough arising from anexcited state is S iv , which has resonance and excited leveltransitions at 1062.66 and 1072.97 ˚A, respectively. There-fore, R determinations for high-luminosity quasars usinghigh-ionization diagnostics from the ground concentrated onusing the above S iv diagnostic (e.g., Borguet et al. 2012,2013; Chamberlain & Arav 2015; Xu et al. 2018). Of specialinterest is the seven quasar sample of Xu et al. (2019), whichwas designed to be unbiased towards particular R ranges(elaboration on this sample is given in Section 3.1).3. Using the Hubble Space Telescope/Cosmic OriginSpectrograph (HST/COS), the 500–1050 ˚A rest-frame re-gion (hereafter EUV500) can be observed in quasars at red-shift z ∼ . The EUV500 contains an order of magnitudemore diagnostic troughs (see Figure 1 in Arav et al. 2020)than are available in ground-based data, which cover the λ > viii , Na ix , Mg x , andSi xii ) whose ionization phase carries most of the outflowing N H (e.g., Arav et al. 2013). Recently, we observed a sampleof 10 EUV500 objects that show outflows. As described inArav et al. (2020) (and summarized in Section 2 here), thissample is the most suitable to learn about the properties ofabsorption outflows in high luminosity quasars. First, theseobjects are representative of the majority of absorption out-flows. Second, they are unbiased towards specific R , N H , andvelocity ranges. Third, they give a census of the dominantvery high-ionization phase (VHP) described above. In this paper we summarize the results of this sample and comparethem to the S iv sample of Xu et al. (2019) as well as toother types of quasar outflows.We note that the value of R is also important to thefeasibility of AGN feedback for another reason. R is a directmeasure of how widespread the impact of the outflow is. Ifthe outflows are confined to the AGN nuclear region, theymay not do much to impact gas accretion and star formationthroughout the galaxy as a whole. As we show below, mostof the outflows in our sample have R of several hundred to afew thousand parsecs. That is, they are on a galactic scale.This paper is structured as follows. Section 2 describeshow the EUV500 sample was obtained, discusses which out-flow properties are unbiased and describes the relationshipbetween this sample and the population of quasar absorp-tion outflows as a whole. Section 3 starts by describing twocomparison samples. It then details the physical properties,distances, and energetics of both the individual outflows andhost quasars in the EUV500 sample as well as compares thesamples. In the Discussion (Section 4) we: a) advocate thatthe criterion for whether an outflow is energetic enough tocause AGN feedback effects should be based on the ratio (cid:219) E k L Edd (where L Edd is the Eddington luminosity) and not of (cid:219) E k L bol ; b) Extrapolate the results of the sample to the major-ity of luminous quasars; and c) discuss the full census of (cid:219) E k .We summarize our results in section 5. The EUV500 sample gets its name from the fact that the ob-servations primarily cover portions of the EUV500 for eachquasar. Distances are determined from excited state transi-tions from various ionic species, but mostly from Ne v andNe vi .The sample is comprised of nine quasars that werefirst observed in programs where the scientific goals wereto use the quasar light to probe intervening absorption fromthe IGM, the circumgalactic medium (CGM), galaxy ha-los, or high-velocity clouds. Eight of the nine quasars arefrom the aforementioned dedicated survey for quasar out-flows observed during program GO-14777 (PI: N. Arav). Theninth is HE 0238-1904, whose rest-frame spectra cover theEUV500 and was observed during program GO-11541 (PI:J. Green). We note that there were two additional quasarsobserved during program GO-14777 that are not included inthis sample. The first, 2MASS J1436+0727, has one outflowand it shows only VHP troughs, which do not yield den-sity/distance diagnostics. Therefore, its R and (cid:219) E k are un-determined, and we opted to remove this object from thesample. The second, LBQS 1206+1052, has a small redshiftsuch that the rest-frame spectra covers primarily lambda > × − erg cm − s − ˚A − was requiredfor each quasar. An outflow was identified by matching atleast two troughs with the same velocity that arise from res- MNRAS , 1–8 (2020) uasar Absorption Outflows onance transitions of either the high-ionization phase (HP;e.g., O iv , N iv , and S iv ) or the VHP (i.e., Ne viii andMg x ). This identification scheme prevented biases towardsa particular phase (either phase was chosen), a particular R scale (searched for only resonance lines), or a particularvelocity (identified outflows at any velocity).These outflows do not show troughs from abundantsingly ionized species with strong lines (e.g., N ii , O ii andS ii ), and therefore would not show troughs from Mg ii ,Si ii or Fe ii if observed at rest-frame wavelengths ( λ rest )greater than 1050 ˚A. Thus, these are not low-ionization out-flows. In contrast, All nine objects have outflows that showthroughs from high ionization species (e.g., O iv , Ne v ). Asdetailed in Arav et al. (2020) section 4.1, such outflows wouldhave a detectable C iv λ λ rest > ˚A.Therefore, our EUV500 sample is representative of high ion-ization outflows, which are the large majority of observedquasar outflows. We note that due to their relatively lowredshift ( z < . ) the SDSS spectra of these objects do notcover the spectral region of the C iv λ ii , Fe ii and sometimes Al iii for the EUV500sample. As expected, we do not see the absorption featuresfor these low ionization species.Detailed analysis of each outflow appears in Arav et al.(2013), Miller et al. (2020a,b,c), and Xu et al. (2020a,b,c).Arav et al. (2020) give an example of how R and (cid:219) E k areextracted from the data and discuss several issues relatedto the EUV500 outflows: the many advantages of studyingquasar outflows using EUV500 data, including: a) measuringthe dominant VHP of the outflow, b) determining the total N H and ionization structure of the outflows, and c) out-flow distance determinations; comparison with X-Ray ob-servations of Seyfert and quasar outflows; comparison withearlier EUV500 observations of quasar outflows; and BroadAbsorption Line (BAL) definition for the EUV500.Here we give a summary of the results for the wholesample. First, in the Appendix we give the derived resultsfor both individual outflows (Table 1: Velocities, VelocityWidths, and Distances of Each Outflow System) and for thehost quasar (Table 2: Quasar Properties and Total OutflowEnergetics). These tables also contain the results for ourmain comparison sample (the S iv sample). Second, our fig-ures show some of the results graphically, while comparingthem to those in the comparison samples. The main comparison sample is a collection of seven quasarsand is known in this work as the S iv sample since all of theoutflow distances are determined from the excited and reso-nance state transitions of S iv , specifically S iv* iv iv* and S iv were not known a priori, preventingan R scale bias just like the EUV500 sample. Detailed anal-ysis of each outflow appears in Borguet et al. (2012, 2013), Figure 1.
Right panel of Figure 1 from Fiore et al. (2017) where (cid:219) E kin (OF) = (cid:219) E k . Molecular, ionized, and X-ray outflows are blue,green, and red respectively. The black stars are their BAL out-flows. Our EUV500 and S iv samples are overlaid in mostly filled,black circles and squares, respectively. The dotted, dashed, andsolid lines denote where (cid:219) E k = 0.01, 0.1, 1.0 L bol , respectively. Chamberlain & Arav (2015), Miller et al. (2018), and Xu etal. (2018, 2019).The secondary comparison sample is from Fiore et al.(2017), where the entire collection consists of over 80 AGNsand includes molecular and ionized emission outflows in ad-dition to absorption outflows observed in the ultraviolet andX-ray.
For these three samples, the total (cid:219) E k for each quasar can bedetermined (see Section 3.3.2 for the EUV500 and S iv sam-ples and Table B.1 of Fiore et al. (2017) for the last sample)and compared. Figure 1 makes such a comparison and showsthe right panel of Figure 1 from Fiore et al. (2017), whereoverlaid on top in mostly filled black circles is the EUV500sample and in mostly filled squares is the S iv sample. Ascan be seen in figure 1 (see also the numerical values in table2), the objects from both the EUV500 and the S iv span theluminosity range × − × erg s − . That is, theyare all luminous quasars. This selection effect arises fromthe need to observe bright targets where a reasonable ex-posure time yield high enough signal-to-noise to enable theanalysis. This is true for both the HST data of the EUV500quasars at redshifts . < z < . , and for the S iv quasarsat redshifts < z < . . Six of the EUV500 and one of theS iv objects comprise half of the 13 most energetic outflowsshown in figure 1. The EUV500 object SDSS J1042+1646has the largest (cid:219) E k across all samples. MNRAS000
For these three samples, the total (cid:219) E k for each quasar can bedetermined (see Section 3.3.2 for the EUV500 and S iv sam-ples and Table B.1 of Fiore et al. (2017) for the last sample)and compared. Figure 1 makes such a comparison and showsthe right panel of Figure 1 from Fiore et al. (2017), whereoverlaid on top in mostly filled black circles is the EUV500sample and in mostly filled squares is the S iv sample. Ascan be seen in figure 1 (see also the numerical values in table2), the objects from both the EUV500 and the S iv span theluminosity range × − × erg s − . That is, theyare all luminous quasars. This selection effect arises fromthe need to observe bright targets where a reasonable ex-posure time yield high enough signal-to-noise to enable theanalysis. This is true for both the HST data of the EUV500quasars at redshifts . < z < . , and for the S iv quasarsat redshifts < z < . . Six of the EUV500 and one of theS iv objects comprise half of the 13 most energetic outflowsshown in figure 1. The EUV500 object SDSS J1042+1646has the largest (cid:219) E k across all samples. MNRAS000 , 1–8 (2020)
Miller et al. [0.1,1) [1,10) [10,100) [1000,3500) [0.1,1) [1,10) [10,100) [1000,3500) F r e qu e n cy R Bins (pc)
EUV500 Outflows S IV Outflows [0.1,1) [1,10) [10,100) [10 ,10 ) [10 ,10 ) Figure 2.
The distribution of distances ( R ) for the outflows where[x,y) = x ≤ R < y. The majority of R lie between 100 and 1000pc. Given the heterogeneous nature of the Fiore et al. (2017)BAL outflows, the remaining comparisons and results areonly on the EUV500 and S iv samples. Table 1 in the Appendix contains the velocities, velocitywidths, and distances of the individual outflow systems. Thevelocity centroid marks the deepest part of the troughs as-sociated with a given outflow system. The velocity of thewidest trough is the midpoint of said trough (or blend oftroughs between outflow systems) where continuous absorp-tion below a residual intensity of 0.9 is observed. The widthof the widest trough is for a single transition from the ionlisted in the table and classifies each outflow as either a BALor mini-BAL (Arav et al. 2020). Only the systems with R constraints are listed.The distribution of R is shown in Figure 2 for the indi-vidual outflow distances listed in Table 1. We note that the R distribution for the two independent samples are consis-tent with each other. The distances for outflow systems 4in UM 425 and 1 in SDSS J1135+1615 are excluded sincethey are upper limits. Also, only the lower limits for the dis-tances of outflow systems 1 in VV2006 J1329+5405 and 2 inSDSS J1512+1119 are included. The majority of R lie above100 pc for both samples, and the maximum and minimumdistances (excluding upper/lower limits) for the samples are3400 pc and 0.15 pc, respectively. Table 2 in the Appendix contains L bol , central black holemass ( M BH ), and Eddington ratio ( L bol L Edd ) of each quasar, ob-tained from the listed references. We note that the M BH for [0.0001,0.5) [0.5,5) [5,55) [0.0001,0.5) [0.5,5) [5,55) F r e qu e n cy E (cid:215) k /L bol Bins (%)
EUV500 Quasars S IV Quasars
Figure 3.
The distribution for the ratio of the total kinetic lu-minosity ( (cid:219) E k ) with respect to the bolometric luminosity ( L bol ) forthe quasars where [x,y) = x ≤ (cid:219) E k L bol (%) < y. HE 0238-1904 was calculated in this work by using the Mg ii -based black hole mass equation from Bahk et al. (2019). Fol-lowing their methodology, we measured the Mg ii FWHMand local continuum level from Figure 1 of Muzahid et al.(2012). The dominant uncertainty in all M BH values is asystematic uncertainty of about 0.3 dex.Based on the velocity width of the widest trough (seeTable 1), for the EUV500 sample, five objects are classifiedas BALQSOs, while four are classified as mini-BALs. Forthe S iv sample five are BALQSOs and two are mini-BALs.The total outflow energetics, i.e., the sum of the individ-ual outflow energetics (excluding outflows with only upperlimits) are also listed in Table 2. This includes the mass flux( (cid:219) M ), momentum flux ( (cid:219) P ), and kinetic luminosity ( (cid:219) E k ). Notethat, for consistency, we use the M BH determined by Xu etal. (2019) for SDSS J0831+0354 instead of Chamberlain &Arav (2015) in the calculation of (cid:219) E k L Edd .The distribution for the ratio of (cid:219) E k with respect to L bol is shown in Figure 3. This ratio is commonly used withinthe literature when investigating the feedback potential ofoutflows (e.g., Crenshaw & Kraemer 2012; Harrison et al.2018). However, in Section 4.1 we argue that the ratio (cid:219) E k L Edd is the meaningful physical comparison for assessing AGNfeedback potential. The divisions at 0.5% and 5% arise fromtheoretical predictions by Hopkins & Elvis (2010) and Scan-napieco & Oh (2004), respectively, where those values markthe minimum ratio required for significant AGN feedback.Nearly half of the quasars have a ratio meeting at least oneof these two thresholds. We note that this distribution doesnot change when L Edd is substituted in for L bol . This occurssince most of the quasars have L bol within a factor of two of L Edd , and the bins span an order of magnitude or more.Figure 4 shows the distribution of (cid:219) P (cid:219) P AGN (total momen-tum load) where (cid:219) P AGN = L bol / c is the radiation momentum MNRAS , 1–8 (2020) uasar Absorption Outflows [0.001,0.01) [0.1,1) [1,10) [10,100) [0.001,0.01) [0.1,1) [1,10) [10,100) F r e qu e n cy P (cid:215) /P (cid:215) AGN
Bins
EUV500 Quasars S IV Quasars [10 - ,10 - ) [10 - ,0.1) [0.1,1) [1,10) [10,100) Figure 4.
The distribution for the ratio of the total momentumflux ( (cid:219) P ) with respect to the radiation momentum flux of the blackhole ( (cid:219) P AG N ) for the quasars where [x,y) = x ≤ (cid:219) P (cid:219) P AGN < y. flux of the quasar. Momentum conserving outflows wouldhave a value of 1 (Fiore et al. 2017). As can be seen, the ma-jority of the EUV500 sample have values above 1 whereasthe majority of the S iv sample have values below 1. Thelower values (on average) of (cid:219) P and (cid:219) E k for the S iv samplecompared to the EUV500 one are plausibly because the dom-inant outflow component measured by the VHP lines isn’tmeasured in the S iv sample (see section 4.3 for elaboration). As noted in the previous section, several works use a specificpercentage of the ratio (cid:219) E k L bol as a threshold for whether anoutflow is energetic enough to cause AGN feedback effects.Here we make the case that the specific percentage shouldbe of the ratio (cid:219) E k L Edd and not of (cid:219) E k L bol . For luminous quasarswhere L bol ∼ L Edd (as is the case for our samples) thispoint wouldn’t matter much. However, for Seyfert galaxieswhere the Eddington ratio L bol L Edd < . , the difference willbe more than an order of magnitude. For example, usingthe criterion (cid:219) E k L bol > . , Crenshaw & Kraemer (2012) findthat 30% of their Syfert outflows have enough energy toproduce AGN feedback. However, if instead (cid:219) E k L Edd > . isused, none of these Seyferts have enough energy to produceAGN feedback.We outline our rationale for why a comparison with L Edd is more fundamental by first showing that theoreticalworks that attempt to derive this ratio are often assumingthat L bol = L Edd , and therefore their (cid:219) E k L bol fraction is im-plicitly an (cid:219) E k L Edd fraction. One example is Scannapieco & Oh (2004) who state: “The greatest uncertainty is (cid:15) k ≡ (cid:219) E k L bol ,the fraction of the total bolometric luminosity (assumed tobe the Eddington luminosity) that appears as kinetic lumi-nosity.” Therefore, their (cid:15) k is actually (cid:219) E k L Edd . They find thatwhen (cid:15) k ≥ strong feedback effects are possible.Another influential theoretical paper is Hopkins & Elvis(2010) which gives the lower limit for the amount of energyneeded for AGN feedback. In the abstract they write: ” ∼ L bol or L Edd . The relevant por-tion of the paper is:“Another way of stating this is, for accretion with anEddington ratio (cid:219) m and black hole mass M BH relative to theexpectation (cid:104) M BH (cid:105) from the M BH ˆa ˘A¸S σ relation, the relevantoutflows will be driven (and star formation suppressed) when η (cid:219) m M BH (cid:104) M BH (cid:105) ∼ . f hot ( ) where η is the feedback efficiency ( (cid:219) E = η L ) ” (Note that f hot ∼ . is the mass fraction in the hot diffuse ISM.)By substituting η = (cid:219) E / L and (cid:219) m = L / L Edd into theirequation (9), we obtain the criterion (cid:219) EL Edd M BH (cid:104) M BH (cid:105) ∼ . f hot which clearly depends on the fraction (cid:219) E k L Edd and not on (cid:219) E k L bol .Di Matteo, Springel, & Hernquist (2005) state: “We fur-ther assume that a small fraction, f , of the radiated lumi-nosity couples thermodynamically to the surrounding gas.”’However they also state: ˆa ˘AIJOwing to the enhanced gasdensity, the black holes, which also merge to form one ob-ject, experience a rapid phase of accretion close to the Ed-dington rate, resulting in significant mass growth.ˆa ˘A˙I Thissuggests that they need L bol to be close to the Eddingtonrate. See also their figure 2. Silk & Rees (1998) describe intheir Section 2.1 the mechanical (i.e. outflow) luminosity asa fraction of L Edd and detail the conditions required to expelgas from the host protogalaxy.Two physical considerations:a) L bol (the total amount of electromagnetic energy emittedper unit of time) can vary by orders of magnitude over timein the same object. Therefore, if it is (cid:219) E k L bol that is important,the ratio will vary similarly with changes of L bol . This doesnot seem to be physically plausible as the feedback effectsare connected to the (slowly changing) parameters of thehost galaxy.b) Physically, AGN feedback works if a fraction of the restmass energy of the forming black hole is entrained back inthe host galaxy (e.g., Loeb 2005). This fraction is related to L Edd and not to L bol , which can be arbitrarily smaller.Therefore, the potential for AGN feedback from an out-flow should be conditional on a percentage of L Edd ratherthan a percentage of L bol . In section 2.1 we demonstrated that the EUV500 sampleis representative of high ionization outflows, which are thelarge majority of observed quasar outflows. Here we argue
MNRAS000
MNRAS000 , 1–8 (2020)
Miller et al. that these results can be extended to the majority of allluminous quasars.Quasar absorption outflows are seen in only a portionof all quasar spectra. For example, high-ionization BALs aredetected in roughly 20% of all quasars (e.g., Hewett & Foltz2003). The most common explanation for this percentage isthat all quasars have BAL outflows, which cover only 20% ofthe solid angle around the object (e.g., Weymann et al. 1991;Reichard et al. 2003). Therefore, in only 20% of the cases ourline of sight towards the quasar intercepts a BAL outflow.Our calculation of the kinetic luminosity takes this consid-eration into account by multiplying the applicable result forthe full solid angle by the appropriate detection-fraction forthe studied population. That is, for high-ionization BALQ-SOs we multiply the full solid angle result by 20% in orderto derive the actual energy such an outflow has on average.This procedure assumes implicitly that such outflows residein all luminous quasars. Assuming that all quasars have ab-sorption outflows, we conclude that most luminous quasarsproduce outflows that can create significant AGN feedback. (cid:219) E k Are we accounting for most of the (cid:219) E k associated with all theoutflows in a given object?For the EUV500 objects there are two issues that cancause underestimation of (cid:219) E k . First, we have shown that themajority of the measured outflowing N H resides in the VHP.However, our highest ionization potential (IP) diagnostic isSi xii with IP=523 eV. X-ray outflows in Seyfert galaxies(the so called warm-absorbers) show diagnostics at muchhigher energy (e.g., IP=9278 eV for Fe xxvi ). Modeling suchspectra show that the ionization phase above what can beprobed with Si xii often carries a larger amount of N H thanthe Si xii and lower ionization phases (e.g., Kaastra et al.2014). It is reasonable to assume that the absorption out-flows in luminous quasars have the same ionization distri-bution as the warm absorbers. If that is the case, we aremissing most of the N H of the EUV500 outflows. Second,out of the 29 outflows we detect in the EUV500 survey, sixare only detected by their VHP phase and they lack R di-agnostics (Arav et al. 2020; Miller et al. 2020c). Therefore,their (cid:219) E k are undetermined and are missing from the total (cid:219) E k census.The underestimation of (cid:219) E k is probably larger for theS iv objects. First, in these objects, we do not cover thespectral region of the VHP diagnostics (e.g., Ne viii andMg x ). Therefore, we cannot measure the VHP of the out-flow which is shown to carry the majority of the outflowing N H in the EUV500 objects. Second, most S iv objects showseveral outflows in C iv , but only one system where an S iv trough is associated with the C iv trough. It is only thatsystem from which we extract (cid:219) E k . Each of the other C iv systems has an (cid:219) E k that we cannot determine. Therefore our (cid:219) E k estimate for the object is a lower limit to the total (cid:219) E k inthat object. Third, like in the case of the EUV500 objects,we may be missing material with even higher ionization thanwe can detect through the VHP. As can be seen in Figure 1, the small dynamical range ofour samples in L bol prevents meaningful correlations to bemade with the outflow properties (see also the left panels ofFigures 1 and 2 in Fiore et al. (2017) for other correlationsthey observed). However, we draw the same conclusions asCrenshaw & Kraemer (2012) about the correlations betweenindividual outflow properties, namely there are no majorcorrelations (see their first four figures). In this paper we summarize the results from our sampleof EUV500 quasar absorption outflows and compared themwith other samples. Our main results are:(i) The outflows in more than half of the EUV500 objectsare powerful enough to be the main agents for AGN feed-back, and most of the outflows in these objects are found at R >
100 pc (see tables 1 and 2).(ii) The sample is representative of the quasar absorp-tion outflow population as a whole. Specifically, it representsthe high ionization outflows, which are the large majority ofquasar UV absorption outflows. Furthermore, the sample isunbiased towards specific ranges of R and (cid:219) E k . Therefore, theanalysis results can be extended to the majority of such ob-jects, including broad absorption line quasars (see section2).(iii) We find that these results are consistent with those ofanother sample: seven ground based observed quasars, whichshow absorption troughs from S iv and S iv * transitions.This comaparison sample is also unbiased towards specificranges of R and (cid:219) E k (see section 3.3) .(iv) We compare our results with a large heterogeneoussample of different types of AGN outflows (molecular, ion-ized seen in emission, ionized seen in absorption, and X-ray).Six of the EUV500 and one of the S iv objects comprise halfof the 13 most energetic outflows shown in figure 1 (see sec-tion 3.2).(v) It is generally assumed that all quasars have absorp-tion outflows, where the detection fraction of absorption out-flows is similar to the fraction of solid angle subtended bythe outflow around the source. Under this assumption, weconclude that most luminous quasars produce outflows thatcan contribute significantly to AGN feedback (see section4.2).(vi) We also discuss the criterion for whether an outflowis energetic enough to cause AGN feedback effects, and con-clude that instead of using a specific percentage of the ratio (cid:219) E k L bol the criterion should be based on a percentage of theratio (cid:219) E k L Edd (see section 4.1).
ACKNOWLEDGEMENTS
T.M., N.A., and X.X. acknowledge support from NASAgrants
HST
GO 14777, 14242, 14054, and 14176 as well as
HST
AR 15786. This support is provided by NASA througha grant from the Space Telescope Science Institute, whichis operated by the Association of Universities for Research
MNRAS , 1–8 (2020) uasar Absorption Outflows in Astronomy, Incorporated, under NASA contract NAS5-26555. T.M. and N.A. also acknowledge support from NASAADAP 48020 and NSF grant AST 1413319. We thank TiagoCosta and Chris Harrison for illuminating discussions. Wethank Fabrizio Fiore for letting us use the right panel of fig-ure 1 from his Fiore et al. (2017) paper, as the base for ourfigure 1. DATA AVAILABILITY STATEMENT
The data underlying this article are available in the article,especially in tables 1 and 2.
REFERENCES
Aditya, J. N. H. S., & Kanekar, N. 2018, MNRAS, 481, 1578Angl´es-Alc´azar, D., Dav´e, R., Faucher-Gigu`ere, C.-A., ¨Ozel, F.,& Hopkins, P. F. 2017, MNRAS, 464, 2840Aoki, K., Oyabu, S., Dunn, J. P., et al. 2011, PASJ, 63, 457Arav, N., Borguet, B., Chamberlain, C., Edmonds, D., & Dan-forth, C. 2013, MNRAS, 436, 3286Arav, N., Liu, G., Xu, X., et al. 2018, ApJ, 857, 60Arav, N., Xu, X., Miller, T. R., et al. 2020, ApJS, 247, 37Bahk, H., Woo, J.-H., & Park, D. 2019, ApJ, 875, 50.Barai, P., Proga, D., & Nagamine, K. 2011, MNRAS, 418, 591Barai, P., Gallerani, S., Pallottini, A., et al. 2018, MNRAS, 473,4003Bischetti, M., Piconcelli, E., Vietri, G., et al. 2017, A&A, 598,A122Blandford, R. D., & Begelman, M. C. 2004, MNRAS, 349, 68Borguet, B. C. J., Edmonds, D., Arav, N., Benn, C., & Cham-berlain, C. 2012, ApJ, 758, 69Borguet, B. C. J., Arav, N., Edmonds, D., Chamberlain, C., &Benn, C. 2013, ApJ, 762, 49Chamberlain, C., & Arav, N. 2015, MNRAS, 454, 675Chartas, G., Saez, C., Brandt, W. N., et al. 2009, ApJ, 706, 644Choi, E., Ostriker, J. P., Naab, T., et al. 2017, ApJ, 844, 31Choi, H., Leighly, K. M., Terndrup, D. M., et al. 2020, ApJ, 891,53Cicone, C., Maiolino, R., Sturm, E., et al. 2014, A&A, 562, A21Ciotti, L., Ostriker, J. P., & Proga, D. 2009, ApJ, 699, 89Crenshaw, D. M., & Kraemer, S. B. 2012, ApJ, 753, 75Dai, X., Shankar, F., & Sivakoff, G. R. 2008, ApJ, 672, 108-114de Kool, M., Arav, N., Becker, R. H., et al. 2001, ApJ, 548, 609Di Matteo T., Springel V., Hernquist L., 2005, Natur, 433, 604Dubois, Y., Volonteri, M., & Silk, J. 2014, MNRAS, 440, 1590Dunn, J. P., Bautista, M., Arav, N., et al. 2010, ApJ, 709, 611Faucher-Gigu`ere, C.-A., Quataert, E., & Murray, N. 2012, MN-RAS, 420, 1347Ferland, G. J., Chatzikos, M., Guzm´an, F., et al. 2017, Rev. Mex.Astron. Astrofis., 53, 385Feruglio, C., Fiore, F., Carniani, S., et al. 2015, A&A, 583, A99Fiore, F., Feruglio, C., Shankar, F., et al. 2017, A&A, 601, A143Ganguly, R., & Brotherton, M. S. 2008, ApJ, 672, 102-107Hamann, F. W., Barlow, T. A., Chaffee, F. C., Foltz, C. B., &Weymann, R. J. 2001, ApJ, 550, 142Harrison, C. M., Alexander, D. M., Mullaney, J. R., et al. 2014,MNRAS, 441, 3306Harrison, C. M., Costa, T., Tadhunter, C. N., et al. 2018, NatureAstronomy, 2, 198Hewett, P. C., & Foltz, C. B. 2003, AJ, 125, 1784Hopkins, P. F., Murray, N., & Thompson, T. A. 2009, MNRAS,398, 303Hopkins, P. F., & Elvis, M. 2010, MNRAS, 401, 7. Kaastra, J. S., Kriss, G. A., Cappi, M., et al. 2014, Science, 345,64Kanekar, N., & Chengalur, J. N. 2008, MNRAS, 384, L6Kelly, B. C., Becker, A. C., Sobolewska, M., et al. 2014, ApJ, 788,33Khalatyan, A., Cattaneo, A., Schramm, M., et al. 2008, MNRAS,387, 13Knigge, C., Scaringi, S., Goad, M. R., & Cottis, C. E. 2008, MN-RAS, 386, 1426Korista, K. T., Bautista, M. A., Arav, N., et al. 2008, ApJ, 688,108-115Krips, M., Mart´ın, S., Eckart, A., et al. 2011, ApJ, 736, 37Laurenti, M., Vagnetti, F., & Middei, R. 2020, Journal of PhysicsConference Series, 012015Loeb A., 2005, gbha.conf, 18Lucy, A. B., Leighly, K. M., Terndrup, D. M., Dietrich, M., &Gallagher, S. C. 2014, ApJ, 783, 58Miller, T. R., Arav, N., Xu, X., et al. 2018, ApJ, 865, 90.Miller, T. R., Arav, N., Xu, X., et al. 2020a, ApJS, 247, 39Miller, T. R., Arav, N., Xu, X., et al. 2020b, ApJS, 247, 41Miller, T. R., Arav, N., Xu, X., et al. 2020c, ApJS, 245, 15Moe, M., Arav, N., Bautista, M. A., & Korista, K. T. 2009, ApJ,706, 525Morganti, R., & Oosterloo, T. 2018, A&ARv, 26, 4Murray, N., Chiang, J., Grossman, S. A., & Voit, G. M. 1995,ApJ, 451, 498Muzahid, S., Srianand, R., Savage, B. D., et al. 2012, MNRAS,424, L59Nardini, E., Reeves, J. N., Gofford, J., et al. 2015, Science, 347,860Nomura, M., Ohsuga, K., & Done, C. 2020, MNRAS, 494, 3616Ostriker, J. P., Choi, E., Ciotti, L., Novak, G. S., & Proga, D.2010, ApJ, 722, 642Peirani, S., Dubois, Y., Volonteri, M., et al. 2017, MNRAS, 472,2153Proga, D., Stone, J. M., & Kallman, T. R. 2000, ApJ, 543, 686Proga, D., & Kallman, T. R. 2004, ApJ, 616, 688Reichard, T. A., Richards, G. T., Hall, P. B., et al. 2003, AJ, 126,2594Rosas-Guevara, Y. M., Bower, R. G., Schaye, J., et al. 2015, MN-RAS, 454, 1038Rumbaugh, N., Shen, Y., Morganson, E., et al. 2018, ApJ, 854,160Rupke, D. S. N., & Veilleux, S. 2013, ApJ, 768, 75S´anchez-S´aez, P., Lira, P., Cartier, R., et al. 2019, ApJS, 242, 10Savage, B. D., & Sembach, K. R. 1991, ApJ, 379, 245Scannapieco, E., & Oh, S. P. 2004, ApJ, 608, 62Schaye, J., Crain, R. A., Bower, R. G., et al. 2015, MNRAS, 446,521Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, ApJS, 194,45Silk, J., & Rees, M. J. 1998, A&A, 331, L1Taylor, P., & Kobayashi, C. 2015, MNRAS, 452, L59Thompson, T. A., Fabian, A. C., Quataert, E., et al. 2015, MN-RAS, 449, 147Tombesi, F., Mel´endez, M., Veilleux, S., et al. 2015, Nature, 519,436Tornatore, L., Borgani, S., Viel, M., & Springel, V. 2010, MNRAS,402, 1911Valentini, M., Murante, G., Borgani, S., et al. 2020, MNRAS, 491,2779Volonteri, M., Dubois, Y., Pichon, C., & Devriendt, J. 2016, MN-RAS, 460, 2979Weymann, R. J., Morris, S. L., Foltz, C. B., & Hewett, P. C. 1991,ApJ, 373, 23Xu, X., Arav, N., Miller, T., & Benn, C. 2018, ApJ, 858, 39Xu, X., Arav, N., Miller, T., et al. 2019, ApJ, 876, 105Xu, X., Arav, N., & Miller, T. 2020a, ApJS, 247, 38MNRAS000
Aditya, J. N. H. S., & Kanekar, N. 2018, MNRAS, 481, 1578Angl´es-Alc´azar, D., Dav´e, R., Faucher-Gigu`ere, C.-A., ¨Ozel, F.,& Hopkins, P. F. 2017, MNRAS, 464, 2840Aoki, K., Oyabu, S., Dunn, J. P., et al. 2011, PASJ, 63, 457Arav, N., Borguet, B., Chamberlain, C., Edmonds, D., & Dan-forth, C. 2013, MNRAS, 436, 3286Arav, N., Liu, G., Xu, X., et al. 2018, ApJ, 857, 60Arav, N., Xu, X., Miller, T. R., et al. 2020, ApJS, 247, 37Bahk, H., Woo, J.-H., & Park, D. 2019, ApJ, 875, 50.Barai, P., Proga, D., & Nagamine, K. 2011, MNRAS, 418, 591Barai, P., Gallerani, S., Pallottini, A., et al. 2018, MNRAS, 473,4003Bischetti, M., Piconcelli, E., Vietri, G., et al. 2017, A&A, 598,A122Blandford, R. D., & Begelman, M. C. 2004, MNRAS, 349, 68Borguet, B. C. J., Edmonds, D., Arav, N., Benn, C., & Cham-berlain, C. 2012, ApJ, 758, 69Borguet, B. C. J., Arav, N., Edmonds, D., Chamberlain, C., &Benn, C. 2013, ApJ, 762, 49Chamberlain, C., & Arav, N. 2015, MNRAS, 454, 675Chartas, G., Saez, C., Brandt, W. N., et al. 2009, ApJ, 706, 644Choi, E., Ostriker, J. P., Naab, T., et al. 2017, ApJ, 844, 31Choi, H., Leighly, K. M., Terndrup, D. M., et al. 2020, ApJ, 891,53Cicone, C., Maiolino, R., Sturm, E., et al. 2014, A&A, 562, A21Ciotti, L., Ostriker, J. P., & Proga, D. 2009, ApJ, 699, 89Crenshaw, D. M., & Kraemer, S. B. 2012, ApJ, 753, 75Dai, X., Shankar, F., & Sivakoff, G. R. 2008, ApJ, 672, 108-114de Kool, M., Arav, N., Becker, R. H., et al. 2001, ApJ, 548, 609Di Matteo T., Springel V., Hernquist L., 2005, Natur, 433, 604Dubois, Y., Volonteri, M., & Silk, J. 2014, MNRAS, 440, 1590Dunn, J. P., Bautista, M., Arav, N., et al. 2010, ApJ, 709, 611Faucher-Gigu`ere, C.-A., Quataert, E., & Murray, N. 2012, MN-RAS, 420, 1347Ferland, G. J., Chatzikos, M., Guzm´an, F., et al. 2017, Rev. Mex.Astron. Astrofis., 53, 385Feruglio, C., Fiore, F., Carniani, S., et al. 2015, A&A, 583, A99Fiore, F., Feruglio, C., Shankar, F., et al. 2017, A&A, 601, A143Ganguly, R., & Brotherton, M. S. 2008, ApJ, 672, 102-107Hamann, F. W., Barlow, T. A., Chaffee, F. C., Foltz, C. B., &Weymann, R. J. 2001, ApJ, 550, 142Harrison, C. M., Alexander, D. M., Mullaney, J. R., et al. 2014,MNRAS, 441, 3306Harrison, C. M., Costa, T., Tadhunter, C. N., et al. 2018, NatureAstronomy, 2, 198Hewett, P. C., & Foltz, C. B. 2003, AJ, 125, 1784Hopkins, P. F., Murray, N., & Thompson, T. A. 2009, MNRAS,398, 303Hopkins, P. F., & Elvis, M. 2010, MNRAS, 401, 7. Kaastra, J. S., Kriss, G. A., Cappi, M., et al. 2014, Science, 345,64Kanekar, N., & Chengalur, J. N. 2008, MNRAS, 384, L6Kelly, B. C., Becker, A. C., Sobolewska, M., et al. 2014, ApJ, 788,33Khalatyan, A., Cattaneo, A., Schramm, M., et al. 2008, MNRAS,387, 13Knigge, C., Scaringi, S., Goad, M. R., & Cottis, C. E. 2008, MN-RAS, 386, 1426Korista, K. T., Bautista, M. A., Arav, N., et al. 2008, ApJ, 688,108-115Krips, M., Mart´ın, S., Eckart, A., et al. 2011, ApJ, 736, 37Laurenti, M., Vagnetti, F., & Middei, R. 2020, Journal of PhysicsConference Series, 012015Loeb A., 2005, gbha.conf, 18Lucy, A. B., Leighly, K. M., Terndrup, D. M., Dietrich, M., &Gallagher, S. C. 2014, ApJ, 783, 58Miller, T. R., Arav, N., Xu, X., et al. 2018, ApJ, 865, 90.Miller, T. R., Arav, N., Xu, X., et al. 2020a, ApJS, 247, 39Miller, T. R., Arav, N., Xu, X., et al. 2020b, ApJS, 247, 41Miller, T. R., Arav, N., Xu, X., et al. 2020c, ApJS, 245, 15Moe, M., Arav, N., Bautista, M. A., & Korista, K. T. 2009, ApJ,706, 525Morganti, R., & Oosterloo, T. 2018, A&ARv, 26, 4Murray, N., Chiang, J., Grossman, S. A., & Voit, G. M. 1995,ApJ, 451, 498Muzahid, S., Srianand, R., Savage, B. D., et al. 2012, MNRAS,424, L59Nardini, E., Reeves, J. N., Gofford, J., et al. 2015, Science, 347,860Nomura, M., Ohsuga, K., & Done, C. 2020, MNRAS, 494, 3616Ostriker, J. P., Choi, E., Ciotti, L., Novak, G. S., & Proga, D.2010, ApJ, 722, 642Peirani, S., Dubois, Y., Volonteri, M., et al. 2017, MNRAS, 472,2153Proga, D., Stone, J. M., & Kallman, T. R. 2000, ApJ, 543, 686Proga, D., & Kallman, T. R. 2004, ApJ, 616, 688Reichard, T. A., Richards, G. T., Hall, P. B., et al. 2003, AJ, 126,2594Rosas-Guevara, Y. M., Bower, R. G., Schaye, J., et al. 2015, MN-RAS, 454, 1038Rumbaugh, N., Shen, Y., Morganson, E., et al. 2018, ApJ, 854,160Rupke, D. S. N., & Veilleux, S. 2013, ApJ, 768, 75S´anchez-S´aez, P., Lira, P., Cartier, R., et al. 2019, ApJS, 242, 10Savage, B. D., & Sembach, K. R. 1991, ApJ, 379, 245Scannapieco, E., & Oh, S. P. 2004, ApJ, 608, 62Schaye, J., Crain, R. A., Bower, R. G., et al. 2015, MNRAS, 446,521Shen, Y., Richards, G. T., Strauss, M. A., et al. 2011, ApJS, 194,45Silk, J., & Rees, M. J. 1998, A&A, 331, L1Taylor, P., & Kobayashi, C. 2015, MNRAS, 452, L59Thompson, T. A., Fabian, A. C., Quataert, E., et al. 2015, MN-RAS, 449, 147Tombesi, F., Mel´endez, M., Veilleux, S., et al. 2015, Nature, 519,436Tornatore, L., Borgani, S., Viel, M., & Springel, V. 2010, MNRAS,402, 1911Valentini, M., Murante, G., Borgani, S., et al. 2020, MNRAS, 491,2779Volonteri, M., Dubois, Y., Pichon, C., & Devriendt, J. 2016, MN-RAS, 460, 2979Weymann, R. J., Morris, S. L., Foltz, C. B., & Hewett, P. C. 1991,ApJ, 373, 23Xu, X., Arav, N., Miller, T., & Benn, C. 2018, ApJ, 858, 39Xu, X., Arav, N., Miller, T., et al. 2019, ApJ, 876, 105Xu, X., Arav, N., & Miller, T. 2020a, ApJS, 247, 38MNRAS000 , 1–8 (2020)
Miller et al.
Xu, X., Arav, N., & Miller, T. 2020b, ApJS, 247, 40Xu, X., Arav, N., & Miller, T. 2020c, ApJS, 247, 42Yuan, F., Yoon, D., Li, Y.-P., et al. 2018, ApJ, 857, 121Zubovas, K., & King, A. R. 2014, MNRAS, 439, 400
APPENDIX
This paper has been typeset from a TEX/L A TEX file prepared bythe author. MNRAS , 1–8 (2020) uasar Absorption Outflows Table 1.
Velocities, Velocity Widths, and Distances of Each Outflow SystemSystem v ( a ) v ( b ) wt ∆ v ( c ) wt Ion ( d ) wt R(km s − ) (km s − ) (km s − ) (pc) EUV500 OutflowsHE 0238-1904 viii + − viii + − PKS J0352-0711 viii + − viii + . − . SDSS J0755+2306 viii + − viii + − SDSS J0936+2005 viii + − viii + − viii + − SDSS J1042+1646 viii + − viii + − viii + − viii + − viii + − viii + − viii + − UM 425 x + − x + − x + − x e < + VV2006 J1329+5405 viii f − . —0.4 + .
7C 1631+3930 viii g > − viii + − viii + − Table 1 continued
MNRAS000
MNRAS000 , 1–8 (2020) Miller et al.
Table 1 ( continued )System v ( a ) v ( b ) wt ∆ v ( c ) wt Ion ( d ) wt R(km s − ) (km s − ) (km s − ) (pc) S iv OutflowsSDSS J0046+0104 iv + − SDSS J0831+0354 iv + − SDSS J0941+1331 iv + − SDSS J1111+1437 iv + − SDSS J1135+1615 iv e < + LBQS J1206+1052 v + − SDSS J1512+1119 iv g > − iv f − . —300 + Note:See Table 2 for references.(a). The velocity centroid of each outflow system.(b). The velocity at the middle of the widest trough. Blended troughsbetween outflows have the same value.(c). Velocity width of the widest trough. Determined by continuousabsorption below a residual intensity of 0.9. Blended troughs betweenoutflows have the same value.(d). Ion of the widest trough.(e). Upper limit with a 1- σ uncertainty.(f). Both a lower and upper limit with a 1- σ uncertainty for each limit.(g). Lower limit with a 1- σ uncertainty. MNRAS , 1–8 (2020) uasar Absorption Outflows Table 2.
Quasar Properties and Total Outflow EnergeticsQuasar L bol log( M BH ) ( a ) L bol L Edd (cid:219) M PP AGN log( (cid:219) E k ) (cid:219) E k L bol (cid:219) E k L Edd ( b ) (10 erg s − ) log( M (cid:12) ) (M (cid:12) yr − ) log(ergs s − ) (%) (%) EUV500 Outflows
HE 0238-1904 ( c ) + − + . − . + . − . + . − . + . − . PKS J0352-0711 ( d ) + − + . − . + . − . + . − . + . − . SDSS J0755+2306 ( e )( m ) > > > > > ( f ) + − + . − . + . − . + . − . + . − . SDSS J1042+1646 ( g ) + − + − + . − . + − . + − ( h ) + − + . − . + . − . + . − . + . − . UM 425 ( f ) + − + . − . + . − . + . − . + . − . VV2006 J1329+5405 ( f )( n ) > − . > − . > − . > − . > − .
7C 1631+3930 ( f ) + − + . − . + . − . + . − . + . − . S iv Outflows
SDSS J0046+0104 ( i ) + − + . − . + . − . + . − . + . − . SDSS J0831+0354 ( j ) + − + . − . + . − . + − + − SDSS J0941+1331 ( i ) + − + . − . + . − . + . − . + . − . SDSS J1111+1437 ( k ) + − + . − . + . − . + . − . + . − . SDSS J1135+1615 ( k )( o ) < + < + . < + . < + . < + . LBQS J1206+1052 ( l ) + . − . + . − . + . − . + . − . + . − . SDSS J1512+1119 ( i )( n ) > − . > − . > − . > − . > − . Note:(a). Accurate to about 0.3 dex.(b). Includes the uncertainties in M BH References: (c). Arav et al. (2013); (d). Miller et al. (2020b); (e). Xu et al. (2020c); (f). Miller et al. (2020c); (g). Xu et al. (2020a); (h). Miller et al. (2020a);(i). Borguet et al. (2012, 2013), Xu et al. (2019); (j). Chamberlain & Arav (2015), Xu et al. (2018); (k). Xu et al. (2018); (l). Miller et al. (2018)(m). Lower limits in columns 5-9 with associated 1- σ uncertainty incorporated within the limit.(n). Lower limits in columns 5-9 with associated 1- σ uncertainties.(o). Upper limits in columns 5-9 with associated 1- σ uncertainties.MNRAS000