Evidence for Obscured broad \oiii Components in Type-2 AGN
aa r X i v : . [ a s t r o - ph . GA ] J a n MNRAS , 1–5 (2020) Preprint 21 January 2021 Compiled using MNRAS L A TEX style file v3.0
Evidence for Obscured broad [O iii ] Components in Type-2 AGN
Xue-Guang Zhang ★ School of Physics and technology, Nanjing Normal University, No. 1, Wenyuan Road, Nanjing, 210023, P. R. China
ABSTRACT
In the manuscr ipt, we report evidence on broad [O iii ] components apparently obscured in Type-2 AGN under the frameworkof the Unified model, after checking properties of broad [O iii ] emissions in large samples of Type-1 and Type-2 AGN in SDSSDR12. We can well confirm the statistically lower flux ratios of the broad to the core [O iii ] components in Type-2 AGN than inType-1 AGN, which can be naturally explained by stronger obscured broad [O iii ] components by central dust torus in Type-2AGN, unless the Unified model for AGN was not appropriate to the narrow emission lines. The results provide further evidenceto support broad [O iii ] components coming from emission regions nearer to central BHs, and also indicate the core [O iii ]component as the better indicator for central activities in Type-2 AGN, due to few effects of obscuration on the core [O iii ]component. Considering the broad [O iii ] components as signs of central outflows, the results provide evidence for strong centraloutflows being preferentially obscured in Type-2 AGN. Furthermore, the obscured broad [O iii ] component can be applied toexplain the different flux ratios of [O iii ] 𝜆 Å / H 𝛽 between Type-1 and Type-2 AGN in the BPT diagram. Key words: galaxies:active - galaxies:nuclei - quasars:emission lines - galaxies:Seyfert
Type-1 AGN (broad line Active Galactic Nuclei) and Type-2 AGN(narrow line AGN) having different observational phenomena canbe well explained by orientation effects of central dust torus, inthe framework of the well-known constantly being revised UnifiedModel (Antouncci 1993; Netzer 2015; Audibert et al. 2017). Cen-tral broad line regions (BLRs) with tens to hundreds of light-days(Kaspi et al. 2000; Bentz et al. 2013; Fausnaugh et al. 2017) to cen-tral black holes (BHs) are totally obscured by central dust torus inType-2 AGN. However, narrow emission line regions (NLRs) withhundreds to thousands of pcs (parsecs) to central BHs (Fischer et al.2013; Hainline et al. 2014; Sun et al. 2017) lead to expected simi-lar properties of narrow emission lines in both Type-1 and Type-2AGN. Therefore, properties of narrow emission lines can be wellapplied to estimate central activities in Type-2 AGN, such as thereported strong linear correlation between AGN power-law con-tinuum luminosity and [O iii ] luminosity (Zakamska et al. 2003;Heckman & Best 2014).Recently, Zhang et al. (2017) have reported the broad [O iii ] emis-sion regions nearer to central BLRs, based on the tighter correlationbetween AGN continuum luminosity and luminosity of broad [O iii ]components, through a larger sample of SDSS (Sloan Digital Sky Sur-vey) blue quasars (Type-1 AGN). Actually, broad [O iii ] componentsin AGN have been studied for more than three decades. Greene & Ho(2005) have shown further effects of central BH potential on broad[O iii ] components. Similar blue broad [O iii ] components can alsobe found in Komossa et al. (2008); Shen et al. (2011); Bennert et al.(2018); Schmidt et al. (2018). More recently, DiPompeo et al. (2018)have shown interesting results on broader and more blue-shiftedbroad [O iii ] emissions in the obscured AGN indicating more power-ful AGN-driven outflows, a probable challenge to the Unified Modelof AGN. Furthermore, besides kinematic study on broad [O iii ] com- ponents, geometric properties of broad [O iii ] components have beenalso well studied in the literature. Sun et al. (2017) have shown thatthe extended narrow emission regions related to broad [O iii ] com-ponents have typically smaller sizes than the sizes of normal [O iii ]emission regions expected by AGN luminosity and/or [O iii ] lumi-nosity (Liu et al. 2013; Hainline et al. 2013, 2014). Besides plentyof research results on properties of broad [O iii ] emissions in AGNin the literature, we here will focus on one another interesting pointon broad [O iii ] emissions in AGN.If broad [O iii ] components were nearer to central BHs, broad[O iii ] components could be more likely to be obscured by centraldust torus in Type-2 AGN under the framework of the Unified model.The results could provide further information on linkage betweennarrow and broad emission lines, and furthermore, could providefurther clues on AGN selection criterion through applications ofnarrow emission line ratios in BPT diagrams (Baldwin et al. 1981;Kauffmann et al. 2003; Kewley et al. 2006, 2013, 2019; Zhang et al.2020) with line ratios on [O iii ] lines. Meanwhile, further consider-ations should be given on applications of [O iii ] properties to tracecentral AGN activities in Type-2 AGN. The manuscript is organizedas follows. In Section 2, we show our main data samples of bothType-1 and Type-2 AGN. In Section 3, we show our main results andnecessary discussions. In Section 4, we give our final conclusions.And in the manuscript, we have adopted the cosmological parametersof 𝐻 = · s − Mpc − , Ω Λ = . and Ω m = . . In the manuscript, Type-1 AGN are collected from the SDSSpipeline classified quasars. Type-2 AGN are collected fromthe SDSS sub-classified AGN in SDSS pipeline classi-fied main galaxies based on the MPA-JHU measurements © Zhang X. G.
Figure 1.
Left panel shows an example on the SSP method determined stellar components in the Type-2 AGN SDSS 2152-53874-0372. Solid lines in black,red and blue show the observed spectrum, the best determined stellar components and the pure line spectrum, respectively. Middle and right panels show twoexamples on the best fitted results to the emission lines around H 𝛽 in the Type-1 AGN SDSS 0294-51986-0528 and in the Type-2 AGN SDSS 2152-53874-0372.In middle and right panels, solid lines in black and red show the line spectrum and the best fitted results, solid pink line shows the determined narrow H 𝛽 , solidlines in green and blue show the determined core and broad [O iii ] components, respectively. In middle panel, dashed red line shows the determined power lawcontinuum emissions, solid purple line shows the determined broad H 𝛽 component, dashed blue line shows the determined optical Fe ii emissions, respectively. Figure 2.
BPT diagram for the 5437 Type-1 AGN (contour in blue) and forthe 6587 Type-2 AGN (contour in red) with reliable narrow emission lines.Solid and dot-dashed purple lines represent the dividing lines discussed inKauffmann et al. (2003); Kewley et al. (2006). Contour in green shows theresults based on more than 240000 narrow-emission-line galaxies in SDSSDR12. ( ).Accepted the criterion of redshift less than 0.3, there are 12342Type-1 AGN from SDSS quasars and 16269 Type-2 AGN fromSDSS main galaxies collected from SDSS DR12. Then, emissionline parameters are measured as follows.For Type-2 AGN and a small number of Type-1 AGN, of whichspectra have clear host galaxy contaminations, the widely acceptedSSP (Simple Stellar Population) method has been firstly appliedto subtract the stellar lights, in order to find more accurate emis-sion line properties. Detailed descriptions of the SSP method canbe found in Bruzual & Charlot (2003); Kauffmann et al. (2003);Cid Fernandes et al. (2005) etc.. The same procedure has been ap-plied in our previous studies in Zhang (2014); Zhang et al. (2016);Rakshit et al. (2017); Zhang et al. (2019), etc.. We do not discuss theSSP method any more, but the left panel of Fig. 1 shows an exam-ple on the SSP method determined stellar components in the Type-2AGN with PLATE-MJD-FIBERID = 2152-53874-0372.After subtractions of the stellar lights, line parameters can bewell measured. The emission lines are mainly considered aroundH 𝛽 (rest wavelength from 4400Å to 5600Å) and around H 𝛼 (rest wavelength from 6250Å to 6850Å), which are fitted simultaneouslyby the following model functions through the Levenberg-Marquardtleast-squares minimization technique (the MPFIT package), similaras what we have done in Zhang et al. (2016, 2017). There are two (ormore if necessary, after checking the fitted results) broad Gaussianfunctions applied to describe each broad Balmer line (especially inType-1 AGN), one narrow Gaussian function applied to each narrowemission line including narrow Balmer lines, [O iii ] , [N ii ] , [O i ] and[S ii ] doublets, and two additional Gaussian components applied todescribe probably broad components of [O iii ] doublet (broad [O iii ]components), one broad Gaussian function applied to describe weakHe ii line, two power law functions applied to describe probable AGNcontinuum emissions underneath the broad H 𝛽 and underneath thebroad H 𝛼 , and the Fe ii template discussed in Kovacevic et al. (2010)applied to describe optical Fe ii lines (especially for Type-1 AGN).When the model functions are applied, the following restrictions havebeen accepted, (1) narrow emission lines have the same redshift, (2)corresponding broad components in broad Balmer lines have thesame redshift, (3) flux ratio of the [O iii ] ([N ii ] ) doublet is fixed tothe theoretical value 3, (4) there are the same line widths of narrowBalmer lines ([O iii ] or [O i ] or [N ii ] or [S ii ] doublets), but differentline widths for different narrow lines. Middle and right panels ofFig. 1 show two examples on the best fitted results to the emissionlines around H 𝛽 .Based on the measured parameters, two criteria have been acceptedto collect Type-2 AGN with reliable narrow emission lines but nobroad Balmer lines. First, measured stellar velocity dispersions andline parameters of the narrow emission lines ([O iii ] 𝜆 Å(at leastcore [O iii ] components), narrow Balmer lines and [N ii ] 𝜆 Å)are at least 5 times larger than their corresponding uncertainties.Second, measured line fluxes of broad Balmer lines are less than5 times of the corresponding uncertainties. Then, there are 6587Type-2 AGN collected. Moreover, two criteria have been accepted tocollect Type-1 AGN with reliable narrow and broad emission lines.First, the measured continuum luminosity and line parameters ofbroad Balmer components are at least 5 times larger than their cor-responding uncertainties. Second, the measured line parameters of[O iii ] 𝜆 Å(at least core [O iii ] components), narrow Balmer linesand [N ii ] 𝜆 Å are at least 5 times larger than their correspondinguncertainties. Then, there are 5437 Type-1 AGN collected. Fig. 2shows the BPT diagram of flux ratio of [O iii ] 𝜆 Å to narrow H 𝛽 (O3HB) versus flux ratio of [N ii ] 𝜆 Å to narrow H 𝛼 (N2HA) forthe collected 5437 Type-1 AGN and 6587 Type-2 AGN. Here, the MNRAS , 1–5 (2020) bscured broad [O iii ] in Type-2 AGN L3 Figure 3.
Distributions of 𝑅 bc of the Type-1 AGN (in blue) and the Type-2AGN (in red) in the main samples. Vertical dashed lines in blue and in redshow the mean value positions of the Type-1 and Type-2 AGN, respectively.Mean value of each distribution is marked in the top-left corner. [O iii ] 𝜆 Å flux means the total [O iii ] line flux. The collected ob-jects can be safely classified as AGN, based on the dividing lines welldiscussed in Kauffmann et al. (2003); Kewley et al. (2006, 2013).Finally, among the 5437 Type-1 AGN and the 6587 Type-2 AGN,the following criteria are applied to create our main samples ofType-1 and Type-2 AGN with reliable broad [O iii ] components: themeasured line flux and line width (second moment) of both the coreand the broad [O iii ] components are at least 5 times larger than theircorresponding uncertainties. Here, the determined [O iii ] componentwith larger second moment is the broad [O iii ] component. Then, infinal main samples, there are 2621 Type-1 AGN and 1987 Type-2AGN, with reliable broad [O iii ] components.
Based on the determined core and broad [O iii ] components, proper-ties of the parameter 𝑅 bc = log ( 𝐿 b ) − log ( 𝐿 c ) have been checked andshown in Fig. 3, where 𝐿 b and 𝐿 c represent luminosities of the broadand the core [O iii ] components, respectively. The mean 𝑅 bc areabout -0.07 and -0.21 in the Type-1 and Type-2 AGN, respectively.And the Student’s T-statistic technique shows that the different meanvalues are significant with levels higher than 𝜎 . Therefore, therecould be intrinsic different properties of [O iii ] components betweenType-1 and Type-2 AGN.In order to show more accurate results with contaminations as lessas possible, The following effects have been mainly considered. Asthe results shown in the top panels of Fig. 4, the Type-1 and Type-2 AGN have much different distributions of redshift, O3HB andN2HA. The different redshift distribution will lead to much differentluminosity properties of [O iii ] components. And, the different dis-tributions of O3HB and N2HA will indicate much different centralactivities between the Type-1 and Type-2 AGN in the main samples.In order to totally ignore effects of different distributions of redshiftand emission line ratios between the Type-1 and Type-2 AGN, a sim-ple method has been considered, by comparing Type-1 and Type-2AGN in two subsamples which have the same distributions of redshiftand emission line ratios (called BPT/redshift-matched samples).Based on the distributions of redshift, O3HB c (flux ratio of thecore [O iii ] component to narrow H 𝛽 ) and N2HA of the 2621 Type-1AGN and the 1987 Type-2 AGN in the main samples, 667 Type-1 AGN and 667 Type-2 AGN are randomly collected from the mainsamples to create the two BPT/redshift-matched samples which havethe same distributions of redshift, O3HB c and N2HA, as the resultsshown in the bottom panels of Fig. 4. Here, not O3HB but O3HB c is applied, because of the broad [O iii ] components in Type-2 AGNprobably obscured. Based on the strong correlation between core[O iii ] luminosity and AGN continuum luminosity in (Zhang et al.2017), central activities can also be well traced by applications ofcore [O iii ] components. The two-sided Kolmogorov-Smirnov statis-tic technique has been applied to confirm the same distributionsof redshift, O3HB c and N2HA with significance levels higher than92% between the 667 Type-1 AGN and the 667 Type-2 AGN in theBPT/redshift-matched samples.Then, the parameter 𝑅 bc has been re-checked in the top panel ofFig. 5 for the AGN in the BPT/redshift-matched samples, with themean 𝑅 bc of about -0.02 and -0.26 in the 667 Type-1 AGN and inthe 667 Type-2 AGN, respectively. The difference is more apparentthan the results for the AGN in the main samples shown in Fig. 3.And the different mean values have Student’s T-statistic determinedsignificance levels higher than 𝜎 . Furthermore, the bottom panelof Fig. 5 shows the luminosity distribution of 𝐿 c for the AGN inthe BPT/redshift-matched samples with the same mean values. Andthrough the two-sided Kolmogorov-Smirnov statistic technique, theType-1 and Type-2 AGN in the BPT/redshift-matched samples havethe same 𝐿 c distributions with significance levels higher than 60%.Therefore, intrinsic different broad [O iii ] components lead to thedifferent 𝑅 bc between the Type-1 and Type-2 AGN, and the directand natural explanation to the lower 𝐿 b in Type-2 AGN is that thebroad [O iii ] emissions have been obscured, under the framework ofthe Unified model for AGN.Before proceeding further, there is one point we should note. Inthe manuscript, we do not discuss the effects of beam smearing onour results. The beam smearing effects have been discussed for morethan five decades (Begeman 1989; Wright et al. 2009; Green et al.2010; Stott et al. 2016; Zheng et al. 2017), especially on kinematicproperties through integral-filed spectra. The more recent discussionson the effects of beam smearing can be found in Husemann et al.(2016, 2020). For the SDSS optical fiber spectra discussed in themanuscript, it is hard to clearly determine and remove the beamsmearing effects on emission lines, due to loss of physical informationof spatially resolved velocity field. In order to roughly check thebeam smearing effects, line width (second moment) difference Δ 𝑏𝑐 = 𝜎 𝑏 − 𝜎 𝑐 between the broad and core [O iii ] components can beroughly applied to show properties of central velocity gradient. Then,the dependence of Δ 𝑏𝑐 on the parameter 𝑅 bc have been checked inthe Type-1 and Type-2 AGN in the BPT/redshift-matched samples.The Spearman rank correlation coefficients are about -0.12 and 0.04for the Type-1 and Type-2 AGN, respectively, strongly indicatingno dependence of the parameter of 𝑅 bc on central velocity gradientin Type-1 AGN nor in Type-2 AGN. Therefore, the beam smearingeffects have few effects on our final results, even there are differentbeam smearing effects between Type-1 and Type-2 AGN due todifferent orientation effects.Meanwhile, we provide further discussions on the obscured broad[O iii ] components in Type-2 AGN. First, similar as results discussedin Zhang et al. (2017), the core rather than the broad [O iii ] compo-nents (or the total [O iii ] ) could be the better indicator to centralactivities in Type-2 AGN. Through the parameter of 𝑅 bc different inType-1 and Type-2 AGN, we can roughly estimate that about 50%of broad [O iii ] emissions are obscured in Type-2 AGN. Therefore,the classification by narrow emission line ratios in the BPT diagramshould lead to lower O3HB, if total [O iii ] lines were considered. MNRAS , 1–5 (2020) Zhang X. G.
Figure 4.
Distributions of redshift, N2HA and
O3HB c of the Type-1 AGN (in blue) and the Type-2 AGN (in red) in the main samples (top panels) and in theBPT/redshift-matched samples (bottom panels). In each top panel, mean values of the distributions are marked in the top-left corner. In each bottom panel, thecalculated Kolmogorov-Smirnov statistical significance level is marked in the top-left corner. Figure 5.
Top panel shows the results similar as those in Fig. 3, but for theAGN in the BPT/redshift-matched samples. Bottom panel shows luminositydistributions of log ( 𝐿 c ) of the AGN in the BPT/redshift-matched samples. Second, based on similar intrinsic properties of [O iii ] emissioncomponents expected by the framework of the Unified model forAGN, there are few selection effects on the results shown in Fig. 5.Third, based on the results shown in Fig. 2, the Type-2 and Type-1 AGN have statistical different O3HB in the BPT diagram, with mean log ( O3HB ) about 0.91 and 0.71 in the Type-1 and Type-2 AGN, re-spectively. Once, we simply accepted that about 50% of broad [O iii ]emissions are obscured in Type-2 AGN by properties of 𝑅 bc , wecould expect the intrinsic flux ratio of log ( O3HB ) in Type-2 AGNabout . + log ( . ) ∼ . similar as the mean value of 0.91 of theType-1 AGN.As the direct and natural explanation on weaker broad [O iii ]emissions in Type-2 AGN by obscuration, it will be interesting toconsider sources of the obscurations. Central dust torus in AGNcould be preferred, rather than randomly moving dust clouds incentral regions, otherwise, there should be similar obscurations onbroad [O iii ] emissions in Type-1 AGN. As discussed properties ofcentral dust torus in Burtscher et al. (2015); Gandhi et al. (2015);Zhuang et al. (2017), opening angle could be around 𝜃 ∼ − ◦ in AGN, and the dust sublimation radius 𝑅 sub ∝ . × 𝐿 . could be accepted as the radius of dust torus. In order to pro-vide appropriate obscurations on broad [O iii ] emissions in Type-2AGN, scales of the distance 𝑅 B3 of broad [O iii ] emission regionsto central BHs could be simply around 𝑅 B3 ∼ 𝑅 sub × tan ( 𝜃 / ) .If a global mean value of 𝜃 ∼ ◦ was accepted, we will have 𝑅 B3 ∼ . ( 𝐿 UV / erg / s ) . . Meanwhile, the BLRs size canbe well estimated as 𝑅 BLRs ∝ ( 𝐿 opt / erg / s ) . light − days(Kaspi et al. 2000; Bentz et al. 2013). Then, an oversimplified resultcan be expected 𝑅 B3 ∼ × 𝑅 BLRs , about 3 magnitudes smallerthan the distance of common [O iii ] emission regions to central BHs(Hainline et al. 2013), providing interesting clues on very broader[O iii ] components in AGN. Certainly, we should note that the ex-pression 𝑅 B3 ∼ × 𝑅 BLRs is estimated through tremendously over-simplified structures of central dust torus, however, the results canprovide structure information on the potential obscured broad [O iii ]emission regions much nearer to central BHs between BLRs andnormal NLRs in AGN, as expected by properties of kinematically-disturbed broad [O iii ] regions related to outflows discussed inSun et al. (2017).
MNRAS , 1–5 (2020) bscured broad [O iii ] in Type-2 AGN L5 Before the end of the section, line intensity properties rather thankinematic properties of broad and core [O iii ] components are mainlyconsidered in the manuscript. More detailed discussions can be foundon kinematic properties of gas outflows through properties of [O iii ]emissions in Type-2 AGN in Woo et al. (2016, 2017) and in Type-1AGN in DiPompeo et al. (2018). Here, we simply check the corre-lation between stellar velocity dispersions ( 𝜎 ★ ) and line widths ofbroad ( 𝜎 𝑏 ) and core [O iii ] ( 𝜎 𝑐 ) components in the Type-2 AGN,and find that the mean ratios of 𝜎 ★ to 𝜎 𝑐 and of 𝜎 ★ to 𝜎 𝑏 are about1.05 and 0.27, respectively. The results well consistent with the re-ported results in Woo et al. (2016) strongly indicate broad [O iii ]components tightly related to outflowing gases in Type-2 AGN. Andthe wider broad [O iii ] components can be well expected due to thebroad [O iii ] emission regions with deeper gravitational potentialnearer to central BHs. Moreover, we check the expected positive cor-relation between line width of the broad [O iii ] components and the[O iii ] luminosity in Type-1 and Type-2 AGN, such as the positivecorrelations shown in Figure 6 in DiPompeo et al. (2018). Here, thebroad [O iii ] luminosity is applied to trace the central AGN luminos-ity as discussed in Zhang et al. (2017) in Type-1 AGN, but the core[O iii ] luminosity is applied in Type-2 AGN. Then, positive correla-tions can be found with Spearman rank correlation coefficients about0.40 with 𝑃 null ∼ − and about 0.32 with 𝑃 null ∼ − in theType-1 AGN and in the Type-2 AGN, respectively. It is clear that thekinematic properties of the collected Type-1 and Type-2 AGN areconsistent with the reported results in the literature. Finally, we give our main conclusions as follows. Based on the SDSShigh quality spectra of large samples of Type-1 and Type-2 AGNwith reliable broad [O iii ] components, different properties of broad[O iii ] components can be confirmed between the Type-1 and Type-2AGN: statistically lower broad [O iii ] luminosities and statisticallylower 𝑅 bc (flux ratio of the broad to the core [O iii ] component)in the Type-2 AGN, after considering necessary effects. The resultsindicate stronger obscuration on the broad [O iii ] components inthe Type-2 AGN due to broad [O iii ] emission regions nearer tocentral BHs, under the framework of the Unified model for AGN.Considering the broad [O iii ] components as robust signs of centraloutflows, the results provide evidence for obscured central outflowsin Type-2 AGN. Moreover, rather than total [O iii ] lines, the core[O iii ] components can be treated as the better indicator of centralactivities in Type-2 AGN, due to few effects of obscuration on thecore [O iii ] components. Furthermore, the obscured broad [O iii ]components can be well applied to explain the different flux ratios ofO3HB in the BPT diagram between Type-1 and Type-2 AGN. ACKNOWLEDGEMENTS
DATA AVAILABILITY
The data underlying this article will be shared on reasonable requestto the corresponding author ([email protected]).
REFERENCES
Antonucci, R., 1993, ARA&A, 31, 473Audibert, A.; Riffel, R.; Sales, D. A.; Pastoriza, M. G.; Ruschel-Dutra, D.,2017, MNRAS, 464, 2139Baldwin, J. A.; Phillips, M.; Terlevich, R., 1981, PASP, 93, 5Begeman, K. G., 1989, A&A, 223, 47Bennert, V. N.; Loveland, D.; Donohue, E.; et al., 2018, MNRAS, 481, 138Bentz, M. C., et al., 2013, ApJ, 767, 149Burtscher, L.; Orban de Xivry, G.; Davies, R. I.; et al., 2015, A&A, 578, 47Bruzual, G.; Charlot, S. 2003, MNRAS, 344, 1000Cid Fernandes, R.; Mateus, A.; Sodre, L.; Stasinska, G.; Gomes, J. M., 2005,MNRAS, 358, 363DiPompeo, M. A.; Hickox, R. C.; Carroll, C. M.; et al., 2018, ApJ, 856, 76Fausnaugh, M. M., et al., 2017, ApJ, 840, 97Fischer, T. C.; Crenshaw, D. M.; Kraemer, S. B., Schmitt, H. R., 2013, ApJS,209, 1Gandhi, P.; Honig, S. F.; Kishimoto, M., 2015, ApJ, 812, 113Green, A. W.; Glazebrook, K.; McGregor, P. J.; et al., 2010, Natur, 467, 684Greene, J. E.; Ho, L. C., 2005, ApJ, 627, 721Hainline, K. N.; Hickox, R.; Greene, J. E.; Myers, Adam D.; Zakamska, N.L., 2013, ApJ, 774, 145Hainline, K. N., et al., 2014, ApJ, 787, 65Heckman, T. M.; Best, P. N., 2014, ARA&A, 52, 589Husemann, B.; Scharwachter, J.; Bennert, V. N.; Mainieri, V.; Woo, J.; D.Kakkad, D., 2016, A&A, 594, 44Husemann, B.; Heidt, J.; De Rosa, A.; et al., 2020, A&A, 639, 117Kaspi, S., et al., 2000, ApJ, 533, 631Kauffmann, G., et al. 2003, MNRAS, 346, 1055Kewley, L. J., Groves, B., Kauffmann, G., Heckman, T., 2006, MNRAS, 372,961Kewley, L. J., et al., 2013, ApJ, 774, 100Kewley, L. J.; Nicholls, D. C.; Sutherland, R. S., 2019, ARA&A, 57, 511Komossa, S.; Xu, D.; Zhou, H.; Storchi-Bergmann, T.; Binette, L., ApJ, 2008,680, 926Kovacevic, J.; Popovic, L. C.; Dimitrijevic, M. S., 2010, ApJS, 189, 15Liu, G.; Zakamska, N. L.; Greene, J. E.; Nesvadba, N. P. H.; Liu, X., 2013,MNRAS, 436, 2576Netzer, H., 2015, ARA&A, 53, 365Rakshit, S.; Stalin, C. S.; Chand, H.; Zhang, X. G., 2017, ApJS, 229, 39Schmidt, E. O.; Oio, G. A.; Ferreiro1, D.; Vega1, L.; Weidmann, W., 2018,A&A, 615, 13Shen, Y.; Richards, G. T.; Strauss, M. A.; et al., 2011, ApJS, 194, 45Stott, J. P.; Swinbank, A. M.; L. Johnson, H. L.; et al., 2016, MNRAS, 457,1888Sun, A. L.; Greene, J. E.; Zakamska, N. L., 2017, ApJ, 835, 222Woo, J.; Bae, H.; Son, D.; Karouzos, M., 2016, ApJ, 817, 108Woo, J.; Son, D.; Bae, H., 2017, ApJ, 839, 120Wright, S. A.; Larkin, J. E.; Law, D. R.; Steidel, C. C.; Shapley, A. E.; Erb,D. K.; 2009, ApJ, 699, 421Zakamska, N. L., et al., 2003, AJ, 126, 2125Zhang, X. G., 2014, MNRAS, 438, 557Zhang, X. G.; Feng, L. L., 2016, MNRAS, 457, 3878Zhang, X. G.; Feng, L. L., 2017, MNRAS, 468, 620Zhang, X. G.; Bao, M.; Yuan, Q. R., 2019, MNRAS Letter, 490, 81Zhang, X. G.; Feng, Y. Q.; Chen, H.; Yuan, Q. R., 2020 ApJ in pressZheng Z.; Wang H.; Ge J.; et al., 2017, MNRAS, 465, 4572Zhuang, M.; Ho, L. C.; Shangguan, J., 2018, ApJ, 862, 118This paper has been typeset from a TEX/L A TEX file prepared by the author.MNRAS000