VLT-FORS2 optical imaging and spectroscopy of 9 luminous type 2 AGN at 0.3<z<0.6: I. Ionized gas nebulae
A. Humphrey, M. Villar-Martín, C. Ramos Almeida, C.N. Tadhunter, S. Arribas, P.S. Bessiere, A. Cabrera-Lavers
aa r X i v : . [ a s t r o - ph . GA ] S e p Mon. Not. R. Astron. Soc. , 1– ?? (2013) Printed 20 October 2018 (MN L A TEX style file v2.2)
VLT-FORS2 optical imaging and spectroscopy of 9luminous type 2 AGN at 0.3 < z < ⋆ A. Humphrey , M. Villar-Mart´ın , , C. Ramos Almeida , C. N. Tadhunter ,S. Arribas , , P. S. Bessiere , A. Cabrera-Lavers Instituto de Astrof´ısica e Ciˆencias do Espa¸co, Universidade do Porto, CAUP, Rua das Estrelas, PT4150-762 Porto, Portugal Centro de Astrobiolog´ıa (INTA-CSIC), Carretera de Ajalvir, km 4, 28850 Torrej´on de Ardoz, Madrid, Spain Astro-UAM, UAM, Unidad Asociada CSIC, Facultad de Ciencias, Campus de Cantoblanco, E-28049, Madrid, Spain Instituto de Astrof´ısica de Canarias, (IAC) V´ıa L´actea s/n, E-38205, La Laguna, Tenerife, Spain Department of Physics and Astronomy, University of Sheffield, Sheffield S3 7RH, UK Universidad de Concepci´on, Departamento de Astronom´ıa, Casilla 160-C, Concepci´on, Chile
Accepted 2015 September 17. Received 2015 September 17; in original form 2015 July 30.
ABSTRACT
We present optical imaging and long slit spectroscopic observations of 9 luminous type2 AGNs within the redshift range 0.3 < z < z , suggesting that the merger rate is independent of AGNpower at the high end of the AGN luminosity function.We find the emission line flux spatial profiles are often dominated by the oftenspatially unresolved central source. In addition, all but one of our sample is associatedwith much fainter, extended line emission. We find these extended emission line struc-tures have a variety of origins and ionization mechanisms: star forming companions,tidal features, or extended ionized nebulae. AGN related processes dominate the ex-citation of the nuclear gas. Stellar photoionization sometimes plays a role in extendedstructures often related to mergers/interactions. Key words:
The coevolution and interplay of galaxies with their nuclearblack hole is a topic of key importance for understandinggalaxy evolution. Negative feedback from an active galac-tic nucleus (AGN hereinafter) may impact on the evolutionof the stellar and gaseous components of the host galaxy,by heating or expelling cold gas that may otherwise haveformed stars or fed the AGN, potentially hindering massassembly activity. Feedback such as this has been invokedto explain the deviation of the observed galaxy stellar massfunction from the theoretical function (White & Frenk 1991;Puchwein & Springel 2013), to account the observed correla-tion between the black hole mass and the stellar mass of the ⋆ Based on observations carried out at the European SouthernObservatory (Paranal, Chile) with FORS2 on VLT-UT1 (pro-gramme 087.B-0034) stellar spheroidal component of the host galaxy (Magorrianet al. 1998), and to facilitate the transformation of dusty, ob-scured galaxies to unobscured, optical galaxies (e.g. Sanderset al. 1988; Bessiere et al. 2014). Although some negativefeedback clearly does take place in galaxies as they un-dergo a phase of AGN activity, the dominant mechanismand strength of this feedback remains a matter of debatein the literature. Positive feedback, where star formation istriggered or enhanced due to AGN-related processes, hasalso been suggested to take place under some circumstances(Rees 1989; Silk 2013), with a few tentative detections inthe literature (Croft et al. 2006; Stroe et al. 2014).Triggering is another key issue for understanding AGNactivity and its relationship with the evolution of the hostgalaxy. Gas-rich major mergers are thought able to triggerpowerful nuclear activity, insofar as they provide a mecha-nism to displace large quantities of cold gas into the centralfew kiloparsecs of one (or more) of the merging galaxies (e.g. c (cid:13) Humphrey et al.
Heckman et al. 1986), and there is a growing body of evi-dence to support this idea (e.g. Tadhunter et al. 2011; RamosAlmeida et al. 2011; 2012; Bessiere et al. 2012; 2014).Radio quiet, type 2 quasars are a potential goldmineof information to improve our understanding of several im-portant aspects of the evolution of massive galaxies, includ-ing the triggering and impact of AGN activity therein. Oneof their main advantages is that the fortuitous obscurationby an optically thick structure of the highly luminous cen-tral engine, which can otherwise outshine the entire stel-lar content of the host galaxy, affords a cleaner picture ofthe galaxy. Moreover, compared to their radio-loud cousins(powerful radio galaxies), the space density of radio-quiettype 2 quasars is roughly an order of magnitude higher(Reyes et al. 2008), making them far more representative ofpowerful active galaxies. In addition, the absence of powerfulradio jets removes substantial ambiguity about whether ornot observed properties of the host galaxy are induced by theradio jets, which often impacts studies of radio loud quasarsand radio galaxies. However, their lack of powerful radioemission or highly luminous optical emission meant that,even after their existence was hypothesized during efforts tounify the seemingly disparate varieties of active galaxy radioquiet, type 2 quasars remained relatively elusive for decades(see e.g. Halpern et al. 1999).It was not until the advent of the Sloan Digital SkySurvey (SDSS) that type 2 radio quiet quasars were identi-fied in significant numbers. Using selection criteria designedto find galaxies that contain gas photoionized by a power-ful but obscured AGN, Zakamska et al. (2003) were ableto identify ∼
300 candidate type 2 quasars within the range0.3 < z < α ) to select type 2 quasarsat z & &
10 kpc)regions of warm (T ∼ H =71 km s − Mpc − , Ω Λ =0.73 andΩ m =0.27. At the redshifts of our sample, this gives an arc-sec to kpc conversion ranging from 4.47 kpc/arcsec to 6.52kpc/arcsec. The sample consists of 9 luminous type 2 AGN selected fromthe SDSS sample of high luminosity type 2 AGN selected byReyes et al. (2008). They are objects with narrow ( < − ) emission lines without underlying broad compo-nents for the recombination lines suggestive of a BLR, andwith line ratios characteristic of non-stellar ionizing radia-tion. Some basic information is presented in Table 1. All ob-jects have redshift z ∼ λ λ l O =log L [ OIII ] L ⊙ =7.8-8.8 (luminosities taken fromReyes et al. (2008) Vizier online catalogue). We selected ourtargets to have large emission line equivalent widths. c (cid:13) , 1– ?? Table 1.
The sample of type 2 quasars (QSO2, l O =log( L [ OIII ] L ⊙ ) > . l O < L [ OIII ] L ⊙ ) SDSS FIRST S . GHz
NVSS S . GHz
Class(g mag) (mJy) (mJy)SDSS J090307.83+021152.2 SDSS J0903+02 0.329 8.79 19.5 22.5 ± ± ± ± ± ± ± ± ± ± ± Table 2.
Summary of the VLT FORS2 observations. Columns: (1) source name; (2) source redshift; (3) date of the observation (run inApril 2011); (4) type of observation, where NB indicates narrow band imaging, IB indicates intermediate band imaging, BB indicates broadband imaging, LSS indicates long slit spectroscopy, and HST indicates HST WFPC2 imaging; (5) filter, or grism and slit combination;(6) the position angle of the long slit, anti-clockwise from north (north through east); (7) the exposure time on source of the observation;(8) the FWHM of the seeing disc as measured from stars in the broad band images; (9) the average FWHM seeing conditions calculatedover the exposure time as measured by the Differential Image Motion Monitor (DIMM) station; (10) the kiloparsec to arcsec conversion.Name z Night Obs. Filter / Grism PA ( ◦ ) Exp. (s) Seeing ( ′′ ) im Seeing( ′′ ) dimm kpc/ ′′ (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)SDSS J0903+02 0.329 28 BB v HIGH 730 0.81 ± ± ± ± ′′ slit -65.5 (PA1) 4200 1.00 ± ′′ slit 63.4 (PA2) 1400 1.05 ± ′′ slit -65.5 (PA1) 2800 0.93 ± ± ± ′′ slit 40.9 3578 0.85 ± ± ± ± ± ′′ slit 9.9 4200 1.05 ± ± ± ± ± ′′ slit -5.9 (PA1) 2142 0.58 ± ′′ slit 42.1 (PA2) 1400 0.76 ± ± ± ′′ slit 37.7 2100 0.71 ± ± ± ± ± ′′ slit 60.7 1443 1.50 ± ± ± ± ± ′′ slit 57.4 2800 0.62 ± ± ± ± ± ′′ slit 90.0 2800 0.84 ± ± ± ′′ slit -64.6 2100 0.59 ± (cid:13) , 1– ?? Humphrey et al.
Figure 1.
Images, the [OIII] λλ For completeness, in Table 1 we also give the flux den-sity of each target at 1.4 GHz, as measured by the FaintImages of the Radio Sky at Twenty-Centimeters survey(FIRST: Becker et al. 1995) and, where available, as mea-sured by the NRAO VLA Sky Survey (NVSS: Condon et al.1998). Issues such as the radio-loudness of the targets willbe explored in Paper II.The sample is similar to that studied by VM11a andVM11b. The main difference is that here we focus on some-what less luminous objects. Our previous sample had a rangeof [OIII] luminosities with median value corresponding to l O =8.84, while the current sample has 8.27. Zakamska et al.(2003) and Reyes et al. (2008) used slightly different crite-ria to select SDSS QSO2 based on their spectroscopic prop-erties, imposing l O > > l O =8.3). The classificationdoes not change if we use criteria Zakamska et al. (2003) in-stead. As a result, while the sample of VM11a and VM11b consisted of 12 QSO2 and 2 HSy2, the current sample con-tains 3 QSO2 and 6 HSy2. The VLT observations were made during the dark nightsof 25-28 April 2011 at the Antu unit of the Very LargeTelescope, as part of the programme 087.B-0034. The Fo-cal Reducer and Low Dispersion Spectrograph (FORS2: Ap-penzeller et al. 1998) was used in long-slit spectroscopy andimaging modes. The seeing FWHM varied between ∼ ∼ ′′ during the run (Table 2).For each quasar the observing procedure involved firstobtaining an image through the v HIGH broad band filter,and in most cases an image through an intermediate or nar-row band filter chosen to contain the redshifted [OII] λ λ c (cid:13) , 1– ?? tivated our choice of slit position angle during the subse-quent long-slit spectroscopic exposures.The spectroscopic observations used one of two grisms.The 600RI grism gave a useful wavelength range of ∼ ± ′′ slit, or 7.2 ± ′′ slit. The 300I grism provided a useful range of ∼ ± ′′ slit.The spatial pixel scale is 0.25 arcsec pixel − .The data wereprocessed and calibrated using IRAF to apply standard datareduction techniques (see Villar-Mart´ın et al. 2012 for de-tails).A total of ten luminous type 2 AGN at 0.3 < z < Seeing and slit effects
One of the main objectives of this observational programmeis to examine the spatial properties of the narrow emissionline gas of our sample (this paper), and to determine whetherextended ionized outflows are present (Paper II). For this,a careful characterization of the seeing size (FWHM) andshape and its uncertainties is crucial.The seeing was very variable during the VLT observa-tions (FWHM in the range ∼ ′′ -1.3 ′′ ). To account for theuncertainties, we quote in Table 2 the seeing size measuredfrom the broad band and narrow/intermediate band imagesusing several stars in the field (column 8) and the aver-age FWHM seeing conditions calculated over the exposuretime as measured by the Differential Image Motion Monitor(DIMM) station (column 9). The dispersion in these valueswill be carefully taken into account, especially when it couldhave an impact on our conclusions regarding the spatial ex-tension of a given object.In addition, we have reconstructed the spatial profile ofthe seeing disk along the slit, using a non-saturated star inimages taken immediately before or after the spectroscopicobservation of the science target. Because our study of thespatial extension of the ionized gas will be based primarilyon [OIII] λ ′′ )or 5 pixels (1.3 ′′ ) wide depending on the slit used for thescience target’s spectroscopic observation. Finally, the skybackground was removed from the stellar spatial profiles.For some of our science targets, the [OII] λ ∼ ′′ and 1.3 ′′ toreach a compromise between optimizing the observing time,obtaining an adequate spectral resolution and avoiding sig-nificant flux loses. The slit was often wider than the seeingdisk as a consequence (see Table 2). This introduces addi-tional uncertainties for spatially unresolved sources on thekinematic measurements. On one hand, if the objects areclumpy, the image of a spatially unresolved clump will besmaller than the slit width and the instrumental profile atthat particular spatial position will be narrower than theprofile measured using the sky or arc lines. This would leadto an underestimation of the intrinsic FWHM of the emis-sion lines, which is the result of subtracting the instrumentalprofile in quadrature. The resulting uncertainties have beencarefully taken into account and will be mentioned whenrelevant. HST data
SDSS J0903+02 and SDSS J0923+01 have also been ob-served using the Wide Field and Planetary Camera 2 on theHubble Space Telescope (
HST ). These data were retrievedfrom the Hubble Legacy Archive (HLA). The images wereobtained for the HST program with identification 10880 andprincipal investigator Henrique Schmitt. In both cases thebroad band F814W filter was used.
Figs. 1-10 show our FORS2 images along with the corre-sponding
HST image where available. Sections of the two-dimensional spectra are also shown to highlight extendednebular features, if detected.In addition, in Fig. 12 we show emission line spatialprofiles (usually [OIII] λ The VLT and
HST broad band images of this QSO2 show atadpole morphology comprising a spatially asymmetric lightprofile, with several bright knots along PA ∼ ◦ in the cen-tral ∼ ′′ (14 kpc; Fig. 1). A low surface brightness tail alongthis same PA extends ∼ ′′ (75 kpc) from the position ofthe quasar. Very faint, diffuse emission is also seen to extend ∼ ′′ (75 kpc) to the north east. In addition, a faint tonguealso extends ∼ ′′ (14 kpc) to the south east. The narrow c (cid:13) , 1– ?? Humphrey et al. band image, containing [OIII] λ α , [NII], [OIII],H β and [OII], located 1.8 ′′ (8 kpc) South East from thequasar. Showing line ratios [OIII] λ β = 1.5 ± λ α = 0.47 ± .
200 km s − ,taking into account possible slit effects (see § ◦ (Fig. 1). Very faint (not plotted in this figuredue to its weakness) [OIII] reaches a maximum extensionof 4.8 ′′ (22 kpc) from the quasar towards the West. The[OIII] λ β ratio is 1.8 ± . It shows narrowlines with FWHM
180 km s − . The broad band morphology of the host galaxy of this QSO2was described in detail by Bessiere et al. (2012); based ondeep Gemini-GMOS broad band optical images, they founda strong morphological disturbance. Our FORS2 optical im-age shows a galaxy morphology that is elongated along aPA ∼ -45 ◦ , with low surface brightness features that resem-ble tails and broad fans (Fig. 7). In addition, the central fewarcsec shows a lopsided flux distribution.Modeling of the stellar populations in the host galaxy ofthis QSO2 reveals a large contribution from a young stellarpopulation with an age of between 50 and 100 Myr (Bessiereet al. 2015, in prep.).The spatial profile of the emission lines along the slitPA=40.9 ◦ is dominated by a spatially unresolved centralcomponent (Fig. 12) with FWHM ∼ ′′ , consistent within Throughout this paper, we define an EELR as a spatially ex-tended nebula of line emitting gas, which shows a clear, contin-uous physical connection to the line emitting gas in the galaxy’snucleus.
Figure 4.
Images, the [OIII] λλ the errors with the seeing values derived from broad bandimage (0.66 ± ′′ ; Table 2). In addition, both [OIII] λ α show a faint and compact (along the slit) featureof narrow emission lines (FWHM <
200 km s − , taking intoaccount possible slit effects). It is located at 2 ′′ (10 kpc)north east of the AGN, blueshifted by 550 km s − (ignoringslit effects) relative to the nuclear emission. Continuum isnot detected at the location of the knot. The object is clearlyseen in the two-dimensional spectrum (Fig. 1) and also as afaint excess above the seeing profile in Fig. 12.This knot shows [OIII] λ β & α . c (cid:13) , 1– ?? Figure 2.
Images, the [OIII] λλ Figure 5.
FIRST 20 cm (1.4 GHz) postage-stamp image of theradio-loud type 2 quasar SDSS J124749.79+015212.6. See also Lal& Ho (2010). (e.g., Kewley et al. 2001). The compact appearance alongthe slit, the narrow lines and the line ratios suggest thatthis is a star forming object.Also within the slit is the unrelated irregular galaxy SDSS J092317.57+010136.6 (see Fig. 7) at ∼ ′′ to the SW.The detected [OIII] λλ α lines place it atz=0.191. This HLSy2 shows a strongly asymmetric light profile inits broad band image (Fig. 2). A low surface brightness tailextends ∼ ′′ (22 kpc) North from the quasar, along a similarposition angle to that of the high surface brightness regionsof the galaxy, and a faint tail (or possibly a poorly resolvedcompanion nucleus) also extends ∼ ′′ (27 kpc) southward.The narrow band image, which samples [OII] λ ∼ ′′ . Seeing broadening at decreasing λ (ex-pected to be § ′′ (24kpc) and maximum extent from the continuum centroid of2.3 ′′ (12 kpc; Table 3). The line is very narrow in the ex-tended gas, with a total doublet FWHM .
235 km s − and c (cid:13) , 1– ?? Humphrey et al.
Figure 3.
Images, the [OIII] λλ ±
45 km s − respectively at both sides of the continuumcentroid, taking possible slit effects into account. In the broad band image, the host galaxy of this HLSy2shows a cometary, asymmetrical light profile in its highersurface brightness region, with an arc (or tail) of low surfacebrightness emission extending North East and then NorthWest, terminating in a knot of emission ∼ ′′ (21 kpc)North of the galaxy (Fig. 3).Our spectroscopic data shows that the [OIII] spatialprofiles along both position angles are dominated by thecentral spatially unresolved source. Along PA1=-5.9, thisappears somewhat narrower than the seeing derived fromthe narrow band image (FWHM ∼ ± ′′ ) (Fig. 12).The slightly broader central [OII] spatial profile along PA1 isconsistent with seeing broadening at shorter λ . Along PA2,[OII] appears slightly extended.In addition, very low surface brightness (too faint tobe plotted in Fig. 12), diffuse [OII] is detected along PA2towards the South-West up to ∼ ′′ or 18 kpc from the con-tinuum centroid. The lines are very narrow in this EELR.The doublet has FWHM .
140 km s − in the EELR – in the most extreme case of maximum possible slit effects, the linescould have at most FWHM=211 ±
37 km s − .The galaxy 43 ′′ North East from the AGN along PA2detected in extended [OII] emission with z=0.571, confirm-ing its association with the quasar. The galaxies 8 ′′ southand 19 ′′ north along PA1 unrelated star forming galaxiesat substantially different redshifts. The broad band image of this HLSy2 shows an extendedlow surface brightness amorphous halo and a compact knotlocated 3 ′′ (17 kpc) SW of the nucleus (Fig. 4).The [OIII] λ ∼ ′′ from the spatial centroidtowards the NE. In addition, faint spatially extended emis-sion is detected toward the SW (too faint to be discerned inFig. 12). It reaches a distance of 3.5 ′′ (20 kpc) from the AGN(Fig. 4) where it seems to connect with the position of theknot seen in the broad band image. This knot also shows lineemission. The [OIII] line is very narrow with FWHM . − (309 ±
30 accounting for the most extreme possibleslit effects) along its full extension and it is redshifted by up c (cid:13) , 1– ?? to ∼
530 km s − relative to the [OIII] emission at the posi-tion of peak flux. For the [OIII] λ β ratio we obtaina 3 σ lower limit of > The broad band image shows several other galaxies within10 ′′ (56 kpc) of the HLSy2, including a pair of galaxieslocated 4 ′′ (22 kpc) south west of the HLSy2 (Fig. 6). Onemember of this galaxy pair fell within the slit, and showsweak [OII] emission and a continuum break consistent withthe 4000 ˚A break at z=0.42, similar to that of the AGNhost.We detect a faint tidal tail extending ∼ ′′ west fromthe western-most of the companion galaxy pair. In ad-dition, after processing the broad band image using the’smoothed galaxy subtraction’ technique of Ramos Almeidaet al. (2011), we detect a tidal tail (or bridge) connecting theHLSy2 to its nearest companion galaxy (see lower panel inFig. 6). We conclude that this is a triple merger/interactingsystem.Our spectroscopic data shows that the [OII] and [OIII]spatial profiles along PA=60.7 are both spatially extended.The [OII] and [OIII] profiles are shown in Fig. 12 togetherwith a seeing profile of FWHM ∼ ′′ . Both [OII] and [OIII]show a clear excess above the seeing disk at both sides ofthe spatial centroid. This is also clear in the 2-dim spectrum(Fig. 6). A visual inspection shows that extended faint [OII]is detected along the slit up to ∼ ′′ NE from the continuumcentroid ( ∼
17 kpc), with well differentiated kinematics. Thelines appear spectrally unresolved with FWHM .
120 km s − (slit effects do not affect this object). For [OIII], the excessis clear towards the east.This galaxy was detected by the FIRST survey (20 cm /1.4 GHz) with a flux density of 7.1 ± ∼ ′′ (400kpc; Fig. 5). The substantially higher NVSS flux densityof 34.1 ± ∼ ′′ to the North East from the AGNhas photometric z =0.456 ± This apparently isolated galaxy shows an elliptical com-pact morphology in our broad and intermediate band images(Fig. 8). No spatial structure is apparent, and there are noother bright galaxies visible within a radius of at least 80kpc. In spite of this unremarkable appearance, the opticalspectrum shows strong Balmer absorption lines and a strongBalmer break (Fig. 11), suggesting a burst of star formationoccurred &
100 Myr ago.Although several strong nebular emission lines are de-tected in our spectrum, we find no evidence for any spatialextension along our slit PA. [OII] is not plotted for this ob-ject because the strong adjacent stellar features prevent anaccurate measurement of the line flux at different spatial
Figure 7.
Images, the [OIII] λλ (cid:13) , 1–, 1–
Images, the [OIII] λλ (cid:13) , 1–, 1– ?? Humphrey et al.
Figure 6.
Images and the [OIII] λλ ′′ × ′′ of broad band image of the field around the HLSy2, after applying the ’smoothed galaxy subtraction’ technique ofRamos Almeida et al. (2011). This analysis has revealed a tidal tail (or bridge) connecting the HLSy2 to its nearest companion galaxy,visible as a dark arc to the south and south-west of the HLSy2. c (cid:13) , 1– ?? Figure 9.
Images, the [OIII] λλ locations (Fig. 11), but there is no evidence of spatial ex-tension for this line either. Indeed, the [OIII] emission israther more centrally-peaked than the 5000-6000 ˚A contin-uum emission. It is also narrower (Fig. 12) than the seeingsize implied by the images and DIMM values ( ∼ ′′ ;Table 2). The broad and narrow band images of this HLSy2 showno clear tidal features or possible interacting neighbours(Fig. 9). However, due to its non-uniform surface brightnessdistribution, with an elongation towards the north west, weconsider it plausible that this is a post-interacting system.The long slit spectrum shows spatially extended emis-sion lines emitted by an EELR (Fig. 9 and (Fig. 12),with the strongest line [OIII] λ ′′ (27 kpc) and maximum extension from the continuumcentroid of ∼ ′′ or 14.5 kpc. The spatial peaks of the emis-sion lines have a small but significant offset of ∼ ′′ (1.3kpc) East of the continuum peak. Across the observed spa-tial extent of the nebulosity, the [OIII] λ β flux ratiois within the range 5-10. The detection of the HeII λ .
200 kms − ), while FWHM .
250 km s − on the East side, taking sliteffects into account. The broad band image of this HSy2 reveals an arm or tail offaint emission extending out to ∼ ′′ (18 kpc) North West(Fig. 10). The association of this extended source is con-firmed by its detection in [OIII] emission at a similar redshiftto the HSy2.The [OIII] spatial profile (Fig. 12) is dominated bya compact unresolved (FWHM ∼ ′′ ) source. In addition,very faint [OIII] extended emission (undetected in the fig-ure) is detected up to ∼ ′′ or 16 kpc from the continuumcentroid (see 2-dim spectrum in Fig. 10). This emission over-laps with the tidal feature discussed above. The detection ofrelatively strong continuum suggests that it is forming stars.Unfortunately, the spectrum of this region is noisy and theline FWHM and ratios cannot be constrained in a usefulway. c (cid:13) , 1– ?? Humphrey et al. !"!!
SDSS J0903+02
PA1 -65.5
Compact feature Seeing. FWHM=0.90” Total flux [OIII] )*+,$ -.+,$ /0+*,$.1234$+15,$678$ : ; $ < * + = . * > $ !"!! SDSS J0903+02
PA2 +63.4 *+,-$ ./,-$
Seeing. FWHM=0.90” Total flux [OIII] : ; < $ = + , > / + ? $ !"!!! SDSS J0923+01
PA +40.9
Seeing. FWHM=0.61” Total flux [OIII] )*+,$ -.+,$ (Compact feature) / $ * + . * $ !"!! SDSS J0950+01
PA +9.9
Seeing. FWHM=1.0” Total flux [OIII] Total flux [OII] )*+,$ -.+,$ / $ * + . * $ !"!!! SDSS J1014+02
PA1 -5.9
Seeing. FWHM=0.68” Total flux [OIII] Total flux [OII] )*+,-$ .*/,-$ + $ / $ :;54,$71*<=$51>,$3?9$ !"!!! SDSS J1014+02
PA2 +42.1 )*+,$ -.+,$
Seeing. FWHM=0.68” Total flux [OIII] Total flux [OII] / $ * + . * $ !"!!! SDSS J1017+03
PA +37.7
Seeing. FWHM=0.79” Total flux [OIII] )*+,$ -.+,$ / $ * + . * $ !"!! SDSS J1247+01
PA +60.7
Seeing. FWHM=1.2” Total flux [OIII] Total flux [OII] )*+,$ -.+,$ / $ * + . * $ !"!!! SDSS J1336-00
PA +57.4
Seeing. FWHM=0.60” Total flux [OIII] ()*+$ ,-*+$ . / $ ) * - / ) $ !"!! SDSS J1416-02
PA +90.0
Seeing. FWHM=1.0” Total flux [OIII] )*+,$ -.+,$ / $ * + . * $ !"!!! SDSS J1452+00
PA -64.6
Seeing. FWHM=0.65” Total flux [OIII] ()*+$ ,-*+$ . / $ ) * - / ) $ Figure 12.
Spatial profile of the [OIII] λ λ We have presented optical imaging and long slit spectro-scopic observations of 9 luminous type 2 AGNs within theredshift range 0.3 < z < The host galaxies of our sample show a variety of opticalmorphologies, ranging from heavily disturbed morphologies c (cid:13) , 1– ?? Table 3.
Summary of results. We specify for each object whether an EELR has been detected (column 2), and where the observationaldata allow, its nature (column 3). In columns 4 and 5 we give the maximum radial extension of the EELR from the AGN, in arcsec andkpc, respectively ; in the case of physically distinct emission line sources (e.g., star-forming knots), we instead give their offset from theAGN. We also specify whether morphological signatures of galaxy interactions have been detected in our images (column 6).Galaxy Feature Spectrum Max. Ext. or Dist. Max. Ext. or Dist. Interactions(arcsec) (kpc)(1) (2) (3) (4) (5) (6)SDSS J0903+02 PA2 +63.4 EELR ? 4.8 22 YesPA1 -65.5 SF compact object Composite 1.8 8SDSS J0923+01 PA-40.9 SF compact object SF 2 10 YesSDSS J0950+01 PA-9.9 EELR ? 2.3 12 YesSDSS J1014+02 PA -42.1 EELR ? 2.8 18 YesSDSS J1017+03 PA-37.7 EELR ? 3.0 17 YesPA -37.7 Compact object ? 3.5 20SDSS J1247+01 PA-60.7 EELR ? 3 17 YesSDSS J1336+00 PA-57.4 No – – – NoSDSS J1416-02 PA90 EELR AGN 3.25 14.5 MaybeSDSS J1452+00 PA64.6 Tidal tail? ? 3.5 16 Yes suggestive of a recent or ongoing major merger, to morpho-logically regular and unremarkable systems.7 out of 9 objects (78%) show strong morphological ev-idence for interactions or mergers in the form of disturbedmorphologies and/or peculiar features such as tidal features,amorphous halos, compact emission line knots (see Table3). The two remaining objects – SDSS J1336-00 (a QSO2)and SDSS J1416-02 (a HLSy2) – appear as isolated galaxiesand have no unequivocal signs of mergers/interactions at thedepth of our images. However, SDSS J1416-02 is associatedwith extended low surface brightness continuum emission,with a lopsided flux distribution that could be interpretedas a post-merger system (among other possible interpreta-tions).In comparison, the detection rate of interac-tions/mergers found by VM11a was 5/13 objects (38%).However, their fraction is a conservative lower limit (statedby the authors), because they had only shallow continuumimages, and no narrow band emission line images.Indeed, our detection rate of interaction signatures isequal to that found by Bessiere et al. (2012) in a completesample of 20 SDSS selected QSO2 at 3 < z <
Our analysis shows that the spatial distribution of the emis-sion lines is in general dominated by a compact spatially un-resolved (i.e. consistent with the seeing disk) central source which emits very strong emission lines associated with thenarrow line region. In addition, extranuclear line emissionof much lower surface brightness is often detected due tostructures of diverse nature: tidal tails, star forming nu-clei/knots/companions and extended ionized nebulae. Themain results are summarized in Table 3. Specifically, wehave detected extranuclear line emission for 8 out of 9 ofour sample. The only exception is SDSS J1336+00, which,on the other hand, is an isolated galaxy with no obvious signsof mergers/interactions. The non detection of extranuclearline emission might be due to a real absence of such emis-sion. However, unlike for other objects, given the absenceof peculiar/interesting features in the optical images the slitPA was chosen blindly so that extended lines along otherPA might have been missed.Our previous studies (VM11a and Villar Mart´ın et al.2012) revealed extended emission lines associated with 7/15QSO2 at similar z although this is a gross lower limit, giventhat the slit was placed blindly for several objects with noobvious peculiar morphological features. On the other hand,we have shown here that [OII] λ z , although theseauthors studied objects with very high [OIII] luminosities(Liu et al. 2013a). c (cid:13) , 1– ?? Humphrey et al.
Extranuclear emission line features other than EELR(companion nuclei, knots, tidal tails) are often identified(4/9 in our current sample or 10/24 in the total sampleincluding our prior studies). Thus, in ∼
50% of the objectsemission line structures whose nature is linked with merg-ers/interactions are detected.EELR (which we differentiate from knots, tails, com-panion nuclei) are detected for 6/9 objects. The maximumextension from the AGN is in the range 12-22 kpc, with mean(and median) values of ∼
17 kpc. Liu et al. (2013a) detectedextranuclear ionized nebulae via [OIII] λ z to our sample. They measured radii ofthe ionized gas nebulae ranging from r =7.5 to r =20 kpc,as seen from the 5 σ detection values. Their mean value is r =14 kpc.Thus, as shown by our previous studies, luminoustype 2 AGN are associated with extranuclear emissionline regions, which are often a complex mixture oftidal/interaction/companion features and EELR, with amixture of excitation mechanisms (AGN related processesand stellar photoionization) whose relative contributionvaries spatially (see also McElroy et al. 2015). These regionstherefore cannot be interpreted as a single gaseous struc-ture where the morphology, kinematics and excitation aredetermined by a single mechanism. VM11a found evidence for recent star formation in theneighbourhood of 5/14 objects. All show signs of merg-ers/interactions. The star formation is happening in generalin companion galaxies, knots, and/or nuclei. Definite (i.e.,confirmed by the emission line spectrum) evidence for recentstar formation is confirmed in two objects of the sample in-vestigated here: SDSS 0903+02 and SDSS J0923+01, bothwith associated compact star forming knots, and possiblyalso SDSS 1017+03. All three also show evidence of interac-tions. These fractions are lower limits, given the obviouslylimited spatial coverage and the lack of sufficient line ratioinformation for diagnostics of the ionizing mechanisms. It isclear that at least part of the extended SF occurring in lu-minous type 2 AGN is triggered by mergers/interactions. Amore complete study in two spatial dimensions would prob-ably reveal recent star formation at least in some objects(e.g. disk galaxies) not necessarily related to this type ofprocesses (e.g. McElroy et al. 2015). To ascertain the pres-ence of a young stellar population in the nuclear regions itwould be necessary to fit the optical nuclear spectrum withspectral synthesis modeling techniques (e.g. Tadhunter et al.2011); Bessiere et al. 2016, in prep.).
We have presented optical imaging and long slit spectro-scopic observations of 9 luminous type 2 AGNs withinthe redshift range 0.3 < z < L [ OIII ] L ⊙ ) < .
3) and three are QSO2 (log( L [ OIII ] L ⊙ ) > . • Signatures of mergers/interaction . 7 out of 9 objects(78%) show clear morphological evidence for interactions ormergers in the form of disturbed morphologies and/or pecu-liar features such as tidal tails, amorphous halos, compactemission line knots, etc. This rate of interaction is consis-tent with other relevant studies of (more luminous) QSO2at similar z , suggesting the merger rate is independent ofthe AGN luminosity at the high end of the AGN luminosityfunction (QSO2 and HLSy2). • Extended line emission . The emission line spatial pro-files are dominated by a bright compact, usually spatiallyunresolved central source. In addition, much fainter, ex-tranuclear emission line features are detected associatedwith 8/9 objects. They are of a diverse nature: EELR (6/9objects) of typical radial sizes 12-22 kpc (consistent withrelated works focused on more luminous objects) as well asfeatures related to mergers/interactions such as star form-ing compact knots and tidal tails (4/9). There is a mixtureof excitation mechanisms (AGN related processes and stellarphotoionization) whose relative contribution varies spatially.While the emission line spectrum of the ionized gas near thecentral engine ( R . few kpc) is clearly excited by AGN re-lated processes, stellar photoionization can also be presentin the extranuclear ionized gaseous structures. Moreover, wehave found evidence, based on the [OII]/[OIII] line flux ra-tio, that the extranuclear ionized gas is often in a lower ion-ization state than the nuclear ionized gas. In addition, the[OII] emission is often more spatially extended than [OIII],suggesting that low ionization lines (such as [OII]) might berelatively more efficient than [OIII] for detecting extranu-clear and extended emissionn line structures in QSO2. ACKNOWLEDGMENTS
We thank the staff at Paranal Observatory for their sup-port during the observations. AH acknowledges Funda¸c˜aopara a Ciˆencia e a Tecnologia (FCT) support throughUID/FIS/04434/2013, and through project FCOMP-01-0124-FEDER-029170 (Reference FCT PTDC/FIS-AST/3214/2012) funded by FCT-MEC (PIDDAC) andFEDER (COMPETE), in addition to FP7 projectPIRSES-GA-2013-612701. AH also acknowledges a MarieCurie Fellowship co-funded by the FP7 and the FCT(DFRH/WIIA/57/2011) and FP7 / FCT Complemen-tary Support grant SFRH/BI/52155/2013. MVM andSA acknowledge support from the Spanish Ministeriode Econom´ıa y Competitividad through the the grantsAYA2012-32295 and AYA-2012-39408-C02-01. CRA is sup-ported by a Marie Curie Intra European Fellowship withinthe 7th European Community Framework Programme(PIEF-GA-2012-327934). RGD acknowledges supportthrough the grant AYA2010-15081.
REFERENCES
Alexandroff R., et al., 2013, MNRAS, 435, 3306Arribas S., Colina L., Bellocchi E., Maiolino R., Villar-Mart´ın M., 2014, A&A, 568, A14 c (cid:13) , 1– ?? Appenzeller I. et al., 1998, The Messenger, 94, 1Baldwin J., Philips M., Televich R., 1981, PASP, 93, 5Becker R. H., White R. L., Helfand D. J., 1995, ApJ, 450,559Bessiere P. S., Tadhunter C. N., Ramos Almeida C., VillarMartin M., 2012, MNRAS, 426, 276Bessiere P. S., Tadhunter C. N., Ramos Almeida C., VillarMart´ın M., 2014, MNRAS, 438, 1839Bian W.H., 2007, in ”The Central Engine of Active Galac-tic Nuclei” ASP Conference Series, Ed. Ho & Wang, Vol.373, page 675Condon J. J., Cotton W. D., Greisen E. W., Yin Q. F.,Perley R. A., Taylor G. B., Broderick J. J., 1998, AJ, 115,1693Hainline K. N., Hickox R. C., Greene J. E., Myers A. D.,Zakamska N. L., Liu G., Liu X., 2014, ApJ, 787, 65Halpern J. P., Turner T. J., George I. M., 1999, MNRAS,307, L47Harrison C. M., Alexander D. M., Mullaney J. R., Swin-bank A. M., 2014, MNRAS, 441, 3306Heckman T. M., Smith E. P., Baum S. A., van BreugelW. J. M., Miley G. K., Illingworth G. D., Bothun G. D.,Balick B., 1986, ApJ, 311, 526Humphrey A., Villar-Mart´ın M., S´anchez S. F., Mart´ınez-Sansigre A., Delgado R. G., P´erez E., Tadhunter C.,P´erez-Torres M. A., 2010, MNRAS, 408, L1Kewley L., Dopita M., Sutherland R., Heisler C. TrevenaJ., 2001, AJ, 556, 121Lal D. V., Ho L. C., 2010, AJ, 139, 1089Liu G., Zakamska N. L.,Greene J. E., Nesvadba N., Liu X.,2013a, MNRAS, 430, 2327Liu G., Zakamska N. L.,Greene J. E., Nesvadba N., Liu X.,2013b, MNRAS, 436, 2576Magorrian J., et al., 1998, AJ, 115, 2285Mart´ınez-Sansigre A., Rawlings S., Lacy M., Fadda D.,Marleau F. R., Simpson C., Willott C. J., Jarvis M. J.,2005, Natur, 436, 666Mart´ınez-Sansigre A., Rawlings S., Lacy M., Fadda D.,Jarvis M. J., Marleau F. R., Simpson C., Willott C. J.,2006a, MNRAS, 370, 1479Mart´ınez-Sansigre A., Rawlings S., Garn T., Green D. A.,Alexander P., Kl¨ockner H.-R., Riley J. M., 2006b, MN-RAS, 373, L80McElroy R., Croom S., Pracy M., Sharp R., Ho, I.T.,Medling A., 2015, MNRAS, 446. 2186Norman C., et al., 2002, ApJ, 571, 218Ohta K., Yamada T., Nakanishi K., Ogasaka Y., Kii T.,Hayashida K., 1996, ApJ, 458, L57Ptak A., Zakamska N. L., Strauss M. A., Krolik J. H., Heck-man T. M., Schneider D. P., Brinkmann J., 2006, ApJ,637, 147Puchwein E., Springel V., 2013, MNRAS, 428, 2966Ramos Almeida C., Tadhunter C. N., Inskip K. J., Mor-ganti R., Holt J., Dicken D., 2011, MNRAS, 410, 1550Ramos Almeida C., et al., 2012, MNRAS, 419, 687Ramos Almeida C., Bessiere P. S., Tadhunter C. N., InskipK. J., Morganti R., Dicken D., Gonz´alez-Serrano J. I.,Holt J., 2013, MNRAS, 436, 997Rees M. J., 1989, MNRAS, 239, 1PReyes R., Zakamska N., Strauss M. et al. 2008, AJ, 136,2373Rodr´ıguez M. I., Villar-Mart´ın M., Emonts B., Humphrey A., Drouart G., Garc´ıa Burillo S., P´erez Torres M., 2014,A&A, 565, A19Sanders D. B., Soifer B. T., Elias J. H., Madore B. F.,Matthews K., Neugebauer G., Scoville N. Z., 1988, ApJ,325, 74Silk J., 2013, ApJ, 772, 112Stroe A., Sobral D., R¨ottgering H. J. A., van Weeren R. J.,2014, MNRAS, 438, 1377Tadhunter C., et al., 2011, MNRAS, 412, 960Villar-Mart´ın M., Humphrey A., Mart´ınez-Sansigre A.,P´erez-Torres M., Binette L., Zhang X. G., 2008, MNRAS,390, 218Villar-Mart´ın M., Tadhunter C., P´erez E., Humphrey A.,Mart´ınez-Sansigre A., Delgado R. G., P´erez-Torres M.,2010, MNRAS, 407, L6Villar-Mart´ın M., Tadhunter C., Humphrey A., FragaEncina R., Gonz´alez Delgado R., P´erez Torres M.,Mart´ınez-Sansigre A., 2011a, MNRAS, 416, 262 (VM11a)Villar-Mart´ın M., Humphrey A., Gonz´alez Delgado R., Col-ina L., Arribas S., 2011b, MNRAS, 418, 2032 (VM11b)Villar-Mart´ın M., Cabrera Lavers A., Bessiere P., Tad-hunter C., Rose M., de Breuck C., 2012, MNRAS, 423,80Villar-Mart´ın M., Emonts B., Rodr´ıguez M., Torres M. P.,Drouart G., 2013a, MNRAS, 432, 2104Villar-Mart´ın M., et al., 2013b, MNRAS, 434, 978Villar Mart´ın M., Emonts B., Humphrey A., CabreraLavers A., Binette L., 2014, MNRAS, 440, 3202Villar Mart´ın M., Bellocchi E., Stern J., TadhunterC., Gonz´alez Delgado R., 2015, MNRAS, in press.arXiv:1509.01056White S. D. M., Frenk C. S., 1991, ApJ, 379, 52Zakamska N. L. et al. 2003, AJ, 126, 2125Zakamska N. L., Strauss M. A., Heckman T. M., Ivezi´c ˇZ.,Krolik J. H., 2004, AJ, 128, 1002Zakamska N. L., et al., 2006, AJ, 132, 1496 c (cid:13) , 1– ?? Humphrey et al.
Figure 8.
Images and the [OIII] λλ Figure 10.
Images, the [OIII] λλ (cid:13) , 1– ?? Flux / 10**-16
Wavelength
SDSS-J1336-00
Sky res. [OII]3727 [OIII]4959,5007--- H-beta
Atm.band
Figure 11.
Nuclear spectrum of SDSS J1336-00. Strong Balmerabsorption lines and the Balmer break suggest the presence ofa post-starburst stellar population of &
100 Myr of age. H β isaffected by an atmospheric band. The flux is in units of × − erg s − cm − arcsec − .c (cid:13) , 1–, 1–