AGN-driven galactic outflows: comparing models to observations
MMon. Not. R. Astron. Soc. , 1–9 (2012) Printed 29 January 2021 (MN L A TEX style file v2.2)
AGN-driven galactic outflows: comparing models toobservations
W. Ishibashi (cid:63) , A. C. Fabian and N. Arakawa Physik-Institut, Universitat Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland Institute of Astronomy, Madingley Road, Cambridge CB3 0HA
Accepted ? Received ?; in original form ?
ABSTRACT
The actual mechanism(s) powering galactic outflows in active galactic nuclei (AGN)is still a matter of debate. At least two physical models have been considered in theliterature: wind shocks and radiation pressure on dust. Here we provide a first quanti-tative comparison of the AGN radiative feedback scenario with observations of galacticoutflows. We directly compare our radiation pressure-driven shell models with the ob-servational data from the most recent compilation of molecular outflows on galacticscales. We show that the observed dynamics and energetics of galactic outflows canbe reproduced by AGN radiative feedback, with the inclusion of radiation trappingand/or luminosity evolution. The predicted scalings of the outflow energetics withAGN luminosity can also quantitatively account for the observational scaling rela-tions. Furthermore, sources with both ultra-fast and molecular outflow detections arefound to be located in the ‘forbidden’ region of the N H − λ plane. Overall, an encour-aging agreement is obtained over a wide range of AGN and host galaxy parameters.We discuss our results in the context of recent observational findings and numeri-cal simulations. In conclusion, AGN radiative feedback is a promising mechanism fordriving galactic outflows that should be considered, alongside wind feedback, in theinterpretation of future observational data. Key words: black hole physics - galaxies: active - galaxies: evolution
Outflows on galactic scales are now commonly observed inactive galactic nuclei (AGN) host galaxies. The outflows aredetected in ionised, neutral, and molecular gas phases, in-dicative of a multi-phase nature (e.g. Sturm et al. 2011;Cicone et al. 2014; Fiore et al. 2017; Fluetsch et al. 2019;Veilleux et al. 2020, and references therein). Among the dif-ferent gas phases, the cool molecular gas is the dominantcomponent carrying the bulk of the outflowing mass. In fact,the molecular phase may account for the majority of the to-tal mass outflow rate in galaxies, with a greater fraction(up to ∼ ∼ L/c ) and high kinetic powers (a few percent of the (cid:63)
E-mail: [email protected]
AGN luminosity) (e.g. Fluetsch et al. 2019). Despite the re-markable observational progress in quantifying the dynamicsand energetics of galactic outflows, their physical origin re-mains unclear. There is ongoing debate trying to determinewhether outflows are powered by jets, winds, or radiation. Inthis context, two main physical models have been proposedand discussed in the literature: wind shocks and radiationpressure on dust. Radio jets can also drive galactic outflows,but this feedback mode may be mostly limited to radio-loudsources, which only form a minority of the total AGN pop-ulation. In addition, cosmic rays may help the launchingof cool dense outflows with a smooth gas distribution, asobserved in magneto-hydrodynamical simulations includingcosmic ray pressure (e.g. Girichidis et al. 2018). The poten-tial role of cosmic ray-driven outflows in galaxy evolution isnow gaining interest (see Veilleux et al. 2020, and referencestherein) but will not be further pursued here.In the ‘wind feedback’ model, a fast nuclear wind (orultra-fast outflow UFO) shocks against the cold interstel-lar medium of the host galaxy, leading to wind energy-driven outflows propagating on large scales (Zubovas & King2012; King & Pounds 2015, and references therein). The © a r X i v : . [ a s t r o - ph . GA ] J a n wind energy-driving mode predicts large momentum boosts( ˙ p ∼ L/c ) and high kinetic powers ( ˙ E k ∼ . L ) for fidu-cial parameters. These numbers are consistent with obser-vational measurements, and the wind energy-driven mecha-nism has been the favoured interpretation for the high ener-getics observed in galactic outflows. A recent work extendsthe wind shock model to a two-dimensional configuration ina disc geometry and shows that the shock expansion tendsto follow the path of least resistance (Menci et al. 2019).Based on a detailed comparison between model outputs andoutflow data, an overall good agreement is reported.In the ‘radiation feedback’ scenario, galactic outflowsare driven by radiation pressure on dusty gas, due to theenhanced radiation-matter coupling (Fabian 1999; Murrayet al. 2005; Thompson et al. 2015). We note that radiationpressure-driving has often been disregarded on the groundsthat it is unable to yield large momentum boosts. This maybe the case in the single scattering limit, but the outcomeis different when one takes into account the trapping of re-processed radiation. In fact, in contrast to common belief,radiation pressure-driven outflows can also attain large val-ues of the momentum rates ( ˙ p (cid:38) L/c ) and kinetic powers( ˙ E k ∼ few% L), provided that radiation trapping is properlyincluded (Thompson et al. 2015; Ishibashi & Fabian 2015;Ishibashi et al. 2018; Costa et al. 2018). Therefore radiationfeedback should be considered as a viable mechanism fordriving galactic outflows, alongside wind feedback.Up to now, galactic outflow observations have been pref-erentially interpreted in the framework of the wind energy-driven outflow model. However, some recent observationalworks start to question wind feedback as the only viable andmost favoured mechanism. In a number of sources hostingboth a small-scale UFO and a large-scale molecular outflow,the measured momentum boosts are found to be significantlylower than the values predicted by a purely wind energy-conserving scenario (e.g. Bischetti et al. 2019; Sirressi et al.2019; Reeves & Braito 2019). This suggests the possibilitythat other mechanisms, such as radiation pressure on dust,may play a role in driving the large-scale galactic outflows.The relative importance of the hot wind and radiation pres-sure can also be empirically constrained by emission linediagnosis. The line ratios observed in quasar outflows sug-gest that radiation pressure can dominate hot gas pressureover a wide range of radii (Stern et al. 2016; Somalwar et al.2020).It is thus timely to reconsider AGN radiative feedbackas an alternative mechanism for powering galactic outflows.Here we provide a first quantitative comparison of the ra-diation pressure-driven outflow model with observations ofgalactic molecular outflows. The paper is structured as fol-lows. We start by recalling the basics of AGN feedbackdriven by radiation pressure on dust (Section 2). In Section3, we first provide a detailed individual comparison with aprototype source (Sect. 3.1), and then compare the radiationpressure-driven shell models with a large sample of molec-ular outflows compiled from the literature (Sect. 3.2). Weconsider the resulting outflow scaling relations in Section 4,and analyse the location of the outflowing sources in the N H − λ plane (Section 5). We discuss the global results inrelation to other physical models and recent observationalfindings (Section 6) and conclude in Section 7. We briefly recall the basics of AGN radiative feedback due toradiation pressure on dust (see our previous papers for moredetails, e.g. Ishibashi & Fabian 2015; Ishibashi et al. 2018).The equation of motion of the radiation pressure-driven shellis given by ddt [ M sh v ] = Lc (1 + τ IR − e − τ UV ) − GM ( r ) M sh r , (1)where L is the central luminosity, M sh is the shell mass,and M ( r ) = σ rG is the total mass distribution assuming anisothermal potential with velocity dispersion σ . The infrared(IR) and ultraviolet (UV) optical depths are respectivelygiven by τ IR = κ IR M sh πr , (2) τ UV = κ UV M sh πr , (3)where κ IR =5 cm g − f dg , MW and κ UV =10 cm g − f dg , MW are the IR and UV opacities, with the dust-to-gas ratio nor-malised to the Milky Way value.In order to launch an outflow, a certain minimal lumi-nosity needs to be exceeded. The critical luminosity is ob-tained by equating the radiative force to the gravitationalforce: L c = 2 cσ M sh r (1 + τ IR − e − τ UV ) − , (4)which may be considered as a generalised form of the effec-tive Eddington luminosity.Integrating the equation of motion (Eq. 1), and neglect-ing the UV exponential term, we obtain the analytic solutionfor the velocity of the outflowing shell v ∼ = (cid:114) LcM sh ( r − R ) + κ IR L πc ( 1 R − r ) , (5)where R is the initial radius and the logarithmic factor isignored.By analogy with the observational works on galacticoutflows, we define three model parameters quantifying theoutflow energetics: the mass outflow rate ( ˙ M ), the momen-tum rate ( ˙ p ), and the kinetic power ( ˙ E k ):˙ M = M sh t flow = M sh vr , (6)˙ p = ˙ Mv = M sh v r , (7)˙ E k = 12 ˙ Mv = M sh v r . (8)The above outflow energetics are computed based on theso-called thin-shell approximation adopted in many obser-vational studies (e.g. Gonz´alez-Alfonso et al. 2017).From the approximate form of the radial velocity pro-file (Eq. 5), we can derive the analytic limits for the massoutflow rate˙ M = (cid:18) LM sh cr ( r − R ) + κ IR LM πcr ( 1 R − r ) (cid:19) / , (9)and kinetic power˙ E k = M sh r (cid:18) LcM sh ( r − R ) + κ IR L πc ( 1 R − r ) (cid:19) / . (10) © , 1–9 GN outflows: models vs. observations As a result, we obtain that the mass outflow rate scaleswith luminosity as˙ M ∝ L / , (11)while the kinetic power scales with luminosity as˙ E k ∝ L / , (12)implying a sub-linear scaling for the mass outflow rate and asuper-linear scaling for the kinetic power (see also Ishibashiet al. 2018).Moreover, two derived quantities are generally used incharacterising the outflow energetics: the momentum ratio( ζ ) and energy ratio ( (cid:15) k ), defined by ζ = ˙ pL/c = M sh v r cL , (13) (cid:15) k = ˙ E k L = 12 M sh v r L . (14)In the radiative feedback scenario, large values of theoutflow energetics are reached at small radii, where the out-flowing shell is optically thick to the reprocessed IR radia-tion. In our picture, the maximal values of both the momen-tum ratio and energy ratio are mainly determined by the ini-tial IR optical depth: ζ IR ∼ √ τ IR , and (cid:15) k , IR ∼ √ τ IR , v IR c (where v IR is the velocity near the IR transparency ra-dius). We have previously discussed how AGN radiativefeedback may qualitatively explain the observed outflow en-ergetics, and analysed the effects of luminosity evolution andshell mass configurations (Ishibashi et al. 2018; Ishibashi &Fabian 2018). Here we provide a quantitative comparison of the AGN ra-diative feedback model (outlined in Section 2) with obser-vational data of galactic molecular outflows.
We first compare the radiation pressure-driven outflowmodel with observations of a particular individual source:IRAS F08572+3915. This system is a key example of a lo-cal dust-obscured ultra-luminous infrared galaxy (ULIRG),with AGN luminosity of L ∼ × erg/s. A molecularoutflow was previously detected in CO observations, with anaverage velocity of ∼
800 km/s (Cicone et al. 2014), and thecorresponding outflow energetics are reported in Fluetschet al. (2019). New high angular resolution CO observationsspatially resolve the molecular outflow, which has an out-flowing mass of M ∼ × M (cid:12) and reaches a velocityof v ∼ ∼ M ∼
350 M (cid:12) / yr,while the outflow momentum rate and kinetic power are˙ p ∼ × dyn and ˙ E k ∼ × erg/s, respectively. Thecorresponding momentum ratio is of order ζ = ˙ p/ ( L/c ) ∼ (cid:15) k = ˙ E k /L ∼ .
04. The high valuesof the outflow energetics measured in this source have beenexclusively interpreted in the framework of the wind energy-driving scenario (Herrera-Camus et al. 2020).
Figure 1.
Comparison of the radial velocity profile of ra-diation pressure-driven outflows with observations of IRASF08572+3915. Constant luminosity case: L ∼ × erg/s, R = 20 pc, f dg = 5 × f dg , MW (blue dotted). Luminosity decaycase: L ( t ) = L (1+ t/t d ) − δ with L = 4 × erg/s, t d = 10 yr, δ = 1, R = 50 pc, f dg = 3 . × f dg , MW (green dashed). The ob-servational data point (black square) is from the latest CO obser-vations (Herrera-Camus et al. 2020), and the vertical/horizontalbars indicate previous measurements (Cicone et al. 2014; Fluetschet al. 2019). Here we examine whether the outflow energetics ob-served in IRAS F08572+3915 can be explained in terms ofradiation pressure on dust. The available observational val-ues of the source (Herrera-Camus et al. 2020) are adoptedas input parameters of the radiation-driven outflow model.We follow the evolution of an outflowing mass of M sh =3 × M (cid:12) , launched from initial radii of R = (20 −
50) pc,propagating in an isothermal potential with velocity disper-sion σ = 150 km/s. Assuming a Milky Way-like dust-to-gasratio, we find that a large-scale massive outflow can not bepowered by the present-day luminosity. We therefore con-sider two possibilities that may help enhance the outflowpowering: IR radiation trapping and AGN luminosity evo-lution (Ishibashi & Fabian 2018).In the nuclear regions of dense starbursts and obscuredAGNs, large gas masses ( ∼ M (cid:12) ) are likely concentratedin the inner tens of parsec scales. Millimetre observationsreveal the existence of compact ( r < ©000
50) pc,propagating in an isothermal potential with velocity disper-sion σ = 150 km/s. Assuming a Milky Way-like dust-to-gasratio, we find that a large-scale massive outflow can not bepowered by the present-day luminosity. We therefore con-sider two possibilities that may help enhance the outflowpowering: IR radiation trapping and AGN luminosity evo-lution (Ishibashi & Fabian 2018).In the nuclear regions of dense starbursts and obscuredAGNs, large gas masses ( ∼ M (cid:12) ) are likely concentratedin the inner tens of parsec scales. Millimetre observationsreveal the existence of compact ( r < ©000 , 1–9 Figure 2.
Comparison of the outflow energetics of radiation pressure-driven outflows with observations of IRAS F08572+3915. Radialprofiles of the mass outflow rate ˙ M (left panel), momentum ratio ζ (middle panel), and energy ratio (cid:15) k (right panel). The representedmodel cases and observational data are the same as in Figure 1. the whole time span required for the outflow to reach thecurrent location (a typical crossing time may be of order t ∼ r/v ∼ yr for r ∼ v ∼ L ∼ × erg/s), and enhanceddust-to-gas ratio ( f dg = 5 × f dg , MW ), a high-velocity out-flow can propagate on galactic scales (blue dotted line). Butthis requires quite extreme initial conditions (with a columndensity of N ∼ × cm − and initial IR optical depthof τ IR , ∼ L ( t ) = L (1 + t/t d ) − (power-law decay)with an initial luminosity of L = 4 × erg/s and char-acteristic timescale t d = 10 yr, the resulting outflow canbe accelerated to velocities of v ∼ N ∼ cm − and τ IR , (cid:46) M (left panel), the momen-tum ratio ζ = ˙ p/ ( L/c ) (middle panel) and the energy ratio (cid:15) k = ˙ E k /L (right panel) as a function of radius. We ob-serve that large values of the momentum ratio ( ζ (cid:38)
10) andenergy ratio ( (cid:15) k ∼ a few%) can be accounted for, providedthat the outflow is initially optically thick to the reprocessedIR radiation. In particular, the IRAS F08572+3915 datacan be adequately reproduced by considering a radiatively-driven outflow coupled with AGN luminosity decay. Sucha luminosity evolution may be expected in the time spanbetween the outflow launch and its current location, while large IR optical depths are likely reached in the buried nucleiof ULIRG-like systems. We next compare the radiation pressure-driven outflowmodel with the most recent compilation of CO molecu-lar outflows in the local Universe (Fluetsch et al. 2019).We consider all objects identified as AGNs (Seyferts andLINERs), but exclude sources classified as ‘fossil’ outflowcandidates. We also include two local molecular outflows re-cently reported in the literature, MCG-03-58-007 (Sirressiet al. 2019) and PDS 456 (Bischetti et al. 2019). We fur-ther complement the local sample with the addition of fourmolecular outflows detected at higher redshifts ( z > . L ∼ − erg/s) and outflowing gas mass ( M out ∼ − M (cid:12) ). Theoutflow energetics are all computed assuming the thin-shellapproximation (cf. Section 2).In Figure 3, we plot the measurements of the momen-tum ratio and energy ratio of the molecular outflows at theobserved radial location (denoted by different black sym-bols). In the same plots, we show the corresponding radialprofiles of the energetics of radiation pressure-driven out-flows (coloured curves). The different model curves representdifferent physical conditions in the sources, ranging from IR-optically thin cases to heavily obscured systems (see the cap-tion of Fig. 3 for numerical values). Compared to previousobservational works, the latest outflow compilation coversa broader range in the measured outflow properties with alarger scatter, likely due to less biased samples (as noted inFluetsch et al. 2019).From the left panel of Figure 3, we see that the molec-ular outflows span a range of momentum boosts, with themajority having values in the range ζ ∼ −
20. In the ra-diative feedback scenario, low momentum ratios ( ζ ∼
1) areeasily obtained for low IR optical depth (without requiring © , 1–9 GN outflows: models vs. observations Figure 3.
Comparison of the outflow energetics observed in a sample of molecular outflows (black symbols) with radiation pressure-driven outflow models (coloured curves). Observational samples: local molecular outflows (dots), upper limits (triangles), additionalhigh-z sources (stars). Model cases: L = 10 erg/s, τ IR , ∼ . L = 5 × erg/s, τ IR , ∼ L = 10 erg/s, τ IR , ∼
50 (green dashed); L ( t ) = L (1 + t/t d ) − δ , with L = 10 erg/s, δ = 1, t d = 10 yr (red dotted). radiation trapping); whereas higher values of the momen-tum ratio ( ζ ∼
10) can be achieved by considering large IRoptical depths. More precisely, the maximal values of themomentum ratio are reached at small radii, where the out-flow is optically thick to the reprocessed IR radiation. As ζ IR ∼ √ τ IR , , a large momentum boost requires a large IRoptical depth at launch. On the other hand, a decay in thecentral luminosity output may help explain the large valuesof the momentum ratio observed at large radii. For instance,a power-law decay in luminosity may account for the highobserved values ( ζ ∼
10) out to large radii ( r ∼
10 kpc).A similar picture is found in the case of the outflow en-ergy ratios, which also span a broad range, with typical val-ues lying between ∼ − and ∼ − (Fig. 3, right panel).The global distribution may be reproduced by consideringdifferent physical conditions in the sources, and in particularthe IR optical depth. Since (cid:15) k , IR ∼ √ τ IR , v IR c , large valuesof the energy ratio require large initial IR optical depths,and the maximal values are again obtained at inner radii.But in contrast to the case of momentum boosts, the large (cid:15) k -values observed at large radii can be mostly accounted forby just assuming large IR optical depth, without the need toinvoke AGN luminosity decay. This could be related to thefact that the energy ratio gets an even greater boost thanthe momentum ratio for enhanced dust-to-gas ratios (due toa steeper dependence on f dg ).Overall, the broad range of momentum ratios and en-ergy ratios observed in galactic molecular outflows can bereproduced by considering different physical conditions inthe radiation pressure-driven outflows. Lower values of theoutflow energetics are naturally obtained in the radiativefeedback scenario, while higher outflow energetics may beaccounted for by including radiation trapping, possibly cou-pled with AGN luminosity decay (Ishibashi & Fabian 2018,see also Zubovas (2018)). We note that objects with ex-ceptionally high values of the outflow energetics ( ζ (cid:29) (cid:15) k (cid:29) .
05) are prime candidates for ‘fossil’ outflows (as identified in Fluetsch et al. (2019)). Such fossil outflowsare observed in sources with (current) low Eddington ratios,suggesting that they were likely powered by a past AGNoutburst that has since faded.
Correlations between the global outflow properties and thecentral AGN have been observed in different samples (Sturmet al. 2011; Veilleux et al. 2013; Cicone et al. 2014; Fioreet al. 2017; Fluetsch et al. 2019; Veilleux et al. 2020, andreferences therein). In general, the outflow rates and kineticpowers are observed to scale with AGN luminosity; in fact,luminosity correlations are to be expected in AGN feedback-driven models.Figure 4 shows the mass outflow rate and the kineticpower as a function of AGN luminosity for the observationalsample described in Section 3.2 (black symbols). In the samefigure, we also plot the predicted analytic scalings derivedin Equations 9-10 (coloured lines). The model scaling rela-tions are computed for a given radius ( r = 1 kpc), whilevarying the outflowing shell mass and dust-to-gas ratio (seethe figure caption for numerical values). We recall that inour radiative feedback scenario, the mass outflow rate scaleswith luminosity as ˙ M ∝ L / , while the kinetic power scalesas ˙ E k ∝ L / (Section 2).Different observational scaling relations for molecularoutflows have been considered in the literature. A sub-linearscaling of the form ˙ M ∝ L . ± . is reported in Fiore et al.(2017), while Fluetsch et al. (2019) obtain a shallower cor-relation of the form ˙ M ∝ L . ± . , based on an enlargedand less biased sample. A flatter slope of ∼ . ± .
04 isfound in X-ray observations of a sample of local galaxieshosting molecular outflows (Laha et al. 2018). Concerningthe kinetic power of molecular outflows, a super-linear rela-tion of the form ˙ E k ∝ L . ± . is instead reported (Fiore ©000
04 isfound in X-ray observations of a sample of local galaxieshosting molecular outflows (Laha et al. 2018). Concerningthe kinetic power of molecular outflows, a super-linear rela-tion of the form ˙ E k ∝ L . ± . is instead reported (Fiore ©000 , 1–9 Figure 4.
Luminosity scaling relations of radiation pressure-driven outflow models ( r = 1 kpc and R = 50 pc) compared to observationalmeasurements. Model cases: M sh = 10 M (cid:12) , f dg = 1 × f dg , MW (red dashed); M sh = 5 × M (cid:12) , f dg = 3 × f dg , MW (blue dash-dot); M sh = 10 M (cid:12) , f dg = 5 × f dg , MW (green dotted). The orange fine-dotted line marks the standard prediction ( ˙ E k /L = 0 .
05) of the windenergy-driven model for fiducial parameters. Observational samples: local molecular outflows (dots), upper limits (triangles), additionalhigh-z sources (stars). et al. 2017). Indeed, the ˙ E k /L ratio increases with increasingAGN luminosity, and the observed steeper-than-linear rela-tion is found to be consistent in terms of slope with the pre-diction of the radiation pressure-driven scenario (Fluetschet al. 2019). On the other hand, as noted in the latter pa-per, most of the sources tend to fall below the canonical 5%line predicted by the fiducial wind energy-driven model (butalso see Section 6).In Figure 4, we note that there are a couple of objectsat the lowest luminosities, which seem to be offset from themain scaling relations. The low values of the present-day lu-minosity ( L ∼ erg/s) in these two sources are likely notsufficient to power the currently observed outflows. In fact,a minimal critical luminosity needs to be exceeded to launcha galactic-scale outflow (Section 2). It is possible that theoutflows we observe today were in fact launched by a pow-erful AGN episode in the past. If the luminosities were actu-ally higher, the source positions in Figure 4 could be shiftedtowards the right, possibly bringing them somewhat closerto the scaling relations for the kinetic power. In principle,we should then also consider the past higher luminosities inthe estimates of the momentum and energy ratios (as thepresent ζ and (cid:15) k values could be overestimated).Most recently, outflowing molecular gas has been ob-served for the first time in the Milky Way galaxy (DiTeodoro et al. 2020). The detection of this fast cold molecu-lar gas presents a challenge to outflow driving models, sincethe current level of nuclear activity of SgrA* is clearly inad-equate. In the radiative feedback scenario, we estimate thatan outflowing mass of ∼ M (cid:12) , launched from the innerfew pc region, would require a central luminosity of the orderof ∼ − erg/s (for a Milky Way dust-to-gas ratio).These values are several orders of magnitude larger than thecurrent luminosity output of the Galactic Centre, and wouldimply accretion at a (cid:46) few percent of the Eddington limit.Such a possibility is actually supported by observational ev- idence, indicating that the Galactic Centre underwent muchmore active episodes (with bright flares) in the past. Forinstance, iron line fluorescence measurements indicate thatthe luminosity of SgrA* was L X (cid:38) erg/s in the pastfew hundred years (Koyama 2018, and references therein).Even greater Seyfert-like luminosities ( ∼ − erg/s)may be expected a few million years ago, e.g. to explain theFermi bubbles (Veilleux et al. 2020, and references therein). N H − λ PLANE
In a limited number of sources (currently a dozen objects),both a small-scale UFO and a galactic-scale molecular out-flow have been detected. UFOs are detected as blueshiftediron absorption lines in the X-ray spectra, while molecularoutflows are observed at sub-mm/far-IR wavelengths. It isoften assumed that a fraction of the kinetic energy of the in-ner UFO is transferred to the large-scale molecular outflowon galactic scales. The associated energy transfer rate orcoupling efficiency factor f can vary between 0 and 1, corre-sponding to the momentum-conserving and energy-drivinglimits respectively (Mizumoto et al. 2019; Smith et al. 2019).Comparison of the momentum rate vs. outflow velocityindicate that the measured momentum boosts can be signifi-cantly lower than the values expected from the wind energy-conserving mechanism. For instance, Reeves & Braito (2019)note that only Mrk 231 and IRAS 17020+4544 (which isa radio-loud source) are consistent with a purely energy-conserving shock. In contrast, in the cases of I Zw 1, PDS456, and MCG-03-58-007, the observed momentum rates lienearly two orders of magnitude below the predicted relationin the energy-conserving scenario. In these sources, the ob-served energetics might be more consistent with momentum-driven outflows (Reeves & Braito 2019; Bischetti et al. 2019; © , 1–9 GN outflows: models vs. observations Sirressi et al. 2019). More in general, some objects may be-long to the momentum-driven regime, while others may becloser to the energy-driven mode, depending on the specificphysical conditions in each particular source (Smith et al.2019; Chartas et al. 2020).This suggests a wide range of coupling efficiencies f ∼ . − . H − λ plane, defined by the column density versus theEddington ratio (Fabian et al. 2008, 2009; Ishibashi et al.2018). The effective Eddington limit for dusty gas delin-eates two distinct areas in this plane: the region of ‘long-lived obscuration’ (to the left of the critical curve) and theso-called ‘forbidden region’ (to the right of the dividing line).In Figure 5, we show the location of ten sources having bothUFO and molecular outflow detections (along with X-raycolumn density estimates) in the N H − λ plane. The ob-jects are collected from the literature and are listed in orderof increasing Eddington ratio (with corresponding labels inbrackets): IC 5063 [ a ] (Mizumoto et al. 2019), NGC 1068 [ b ](Mizumoto et al. 2019), HS 0810+2554 [ c ] (Chartas et al.2020), NGC 6240 [ d ] (Mizumoto et al. 2019), IRAS F05189-2524 [ e ] (Smith et al. 2019), APM 08279+5255 [ f ] (Feruglioet al. 2017; Hagino et al. 2017), PDS 456 [ g ] (Nardini et al.2015; Bischetti et al. 2019), I Zw 1 [ h ] (Reeves & Braito2019), Mrk 231 [ i ] (Mizumoto et al. 2019), F11119+3257 [ j ](Veilleux et al. 2017; Tombesi et al. 2017). We observe thatmost of the sources (9 out of 10) are located in the forbiddenregion (or close to borderline) where outflows are expected;and only one source (which is a radio-loud AGN) lies in theregion of long-lived obscuration. This reinforces the notionthat radiation pressure on dust may play an important rolein driving the observed galactic outflows. Comparison of the AGN radiation pressure-driven modelwith observations of galactic molecular outflows shows thatthe global distribution of momentum ratios and energy ra-tios can be plausibly accounted for: low values of the out-flow energetics are a natural outcome of radiative feedback,whereas higher energetics may be reproduced by invoking IRradiation trapping and/or luminosity decay. Furthermore,the AGN radiative feedback scenario predicts luminosityscalings (a sub-linear scaling for the mass outflow rate anda super-linear scaling for the kinetic power) quantitativelyconsistent with the observational scaling relations. We thusfind a nice overall agreement between model predictions andobservations over a wide range of AGN and host galaxy pa-rameters.
Figure 5.
Location of sources with both UFO and molecular out-flow detections in the N H − λ plane. The observational data pointsare retrieved from different works in the literature and labelled inorder of increasing λ (as listed in the main text). The model curvesrepresent two cases of differing dust opacities (cf. Ishibashi et al.2018): κ IR = 5 cm g − and κ UV = 400 cm g − (green solid), and κ IR = 10 cm g − and κ UV = 800 cm g − (blue dashed). Thehorizontal line at log N = 22 (grey fine-dotted) marks the limitbelow which absorption by outer dust lanes becomes dominant. Menci et al. (2019) consider the wind shock model ina disc geometry and perform a detailed comparison withmolecular outflows in single objects, as well as a broadercomparison with a large sample of (mostly) ionised outflows.Focusing on the radial profiles of the outflow velocity andmass outflow rate, they also analyse the outflow scaling re-lations with AGN luminosity. In all such comparisons, oneshould keep in mind that the model predictions are com-puted at a particular radius r , while the observational valuesare distributed over a range of radial distances. Two radialbins, divided into small-scales ( r < r > r = 1 kpc for the computation of the model energetics inSection 4 should be acceptable. In any case, accurate radialand angular dependences of the outflow properties can notbe achieved by currently available observations. (This is alsowhy Menci et al. (2019) do not make use of the full powerof their two-dimensional analysis). Future observations withhigher resolution and improved sensitivity should allow usto better disentangle between different model predictions.An empirical constraint on the relative importance ofradiation pressure versus hot gas pressure may be obtainedby the analysis of emission line ratios. The observed line ra-tios in the average quasar spectrum are consistent with aradiation pressure-dominated scenario over a wide range ofradial distances ( r ∼ . −
10 kpc) (Stern et al. 2016).Spatially resolved UV emission spectroscopy of the outflowin the obscured quasar SDSS J1356+1026 indicates that theUV line ratios are very similar in the nuclear region ( r (cid:46) r ∼
10 kpc) (Somalwar et al. ©000
10 kpc) (Somalwar et al. ©000 , 1–9 P hot /P rad (cid:46) . H ∼ cm − ) and Compton-thick obscuration on tens ofpc scales (Aalto et al. 2019, and references therein). Suchheavily obscured and buried nuclei can be optically thick tothe IR and even sub-mm wavelengths. For instance, ALMAobservations of the nearby infrared luminous galaxy IC 860indicate that the mm-continuum emission is dominated bydust, leading to significant dust opacities at mm-wavelengths(Aalto et al. 2019).Another important requirement in the radiative feed-back scenario is the presence of significant amounts of dust,which sustain the overall AGN feedback process. Dust grainscan be produced in large quantities in core-collapse super-novae (Owen & Barlow 2015; Wesson et al. 2015), enrich-ing the local environment in nuclear starbursts. Moreover,grain growth in the interstellar medium can also contributeto the increase of the dust mass at later times (Micha(cid:32)lowski2015), while turbulence may further accelerate the growthof dust grains (Mattsson 2020). The associated dust-to-gasratios can be quite high, with values up to ∼ /
30 in dust-reddened quasars (Banerji et al. 2017), and potentially upto ∼ / − /
10 in heavily dust-obscured ULIRG nuclei(Gowardhan et al. 2018). The combination of high columndensities and substantial dust content should lead to large IRoptical depths, forming particularly favourable conditionsfor AGN radiative feedback.On the theoretical side, recent radiation hydrodynamic(RHD) simulations of radiation-driven shells indicate thatthe boost factor is roughly equal to the IR optical depth( ˙ p ∼ τ IR L/c for moderate optical depths), confirming thesimple analytic picture (Costa et al. 2018). The correspon-dence may break down at the highest IR optical depths( τ IR ∼ ζ IR ∼ √ τ IR , ). The impor-tance of radiation trapping is further corroborated in newRHD simulations of AGN radiative feedback, which reportan excellent agreement between the simulated outflow prop-agation and the expected analytic solutions, when includingIR multi-scattering (Barnes et al. 2020).Interestingly, most of the sources with both UFO andmolecular outflow detections are located in the forbiddenregion of the N H − λ plane (Section 5). Similarly, highly dust-reddened quasars at high redshifts ( z ∼ Despite the huge observational progress in the detection andcharacterisation of galactic outflows, the underlying drivingmechanism is still ill-defined. Here we directly compare theAGN radiative feedback model with the most recent com-pilation of galactic molecular outflows. We show that radi-ation pressure on dust can quantitatively account for thedynamics and energetics of galactic outflows (Section 3), aswell as the observational scaling relations (Section 4). Theagreement we find between model predictions and observa-tions, over a wide range of AGN and host galaxy parameters,is quite encouraging. The simple analytic picture of radia-tion feedback on dusty gas also seems to be broadly sup-ported by the latest RHD numerical simulations. Thereforewe should move towards a consensus acknowledging the roleof radiation pressure on dust in driving galactic outflows,alongside wind-driving. Indeed, both wind feedback and ra-diation feedback need to be considered in the interpretationof future observational data to better constrain the physicalorigin of galactic outflows.
DATA AVAILABILITY
No new data were generated or analysed in support of thisresearch.
REFERENCES
Aalto S., Battersby C., Rigopoulou D., Armus L., FalstadN., Rangwala N., Hickox R., Lanz L., Gallagher J., EvansA., Alatalo K., Glenn J., Wheeler J., Fyhrie A., van derWerf P., Yoast-Hull T., Privon G., Mangum J., 2019,BAAS, 51, 515Aalto S., Muller S., K¨onig S., Falstad N., et al. 2019, A&A,627, A147Banerji M., Carilli C. L., Jones G., Wagg J., McMahonR. G., Hewett P. C., Alaghband-Zadeh S., Feruglio C.,2017, MNRAS, 465, 4390Barnes D. J., Kannan R., Vogelsberger M., Marinacci F.,2020, MNRAS, 494, 1143 © , 1–9 GN outflows: models vs. observations Bischetti M., Piconcelli E., Feruglio C., Fiore F., CarnianiS., Brusa M., Cicone C., Vignali C., Bongiorno A., CresciG., Mainieri V., Maiolino R., Marconi A., Nardini E.,Zappacosta L., 2019, A&A, 628, A118Brusa M., Cresci G., Daddi E., Paladino R., et al. 2018,A&A, 612, A29Chartas G., Davidson E., Brusa M., Vignali C., Cappi M.,Dadina M., Cresci G., Paladino R., Lanzuisi G., ComastriA., 2020, MNRAS, 496, 598Cicone C., Maiolino R., Sturm E., Graci´a-Carpio J.,Feruglio C., Neri R., Aalto S., Davies R., Fiore F., et al.2014, A&A, 562, A21Costa T., Rosdahl J., Sijacki D., Haehnelt M. G., 2018,MNRAS, 473, 4197Di Teodoro E. M., McClure-Griffiths N. M., Lockman F. J.,Armillotta L., 2020, arXiv e-prints, p. arXiv:2008.09121Fabian A. C., 1999, MNRAS, 308, L39Fabian A. C., Vasudevan R. V., Gandhi P., 2008, MNRAS,385, L43Fabian A. C., Vasudevan R. V., Mushotzky R. F., WinterL. M., Reynolds C. S., 2009, MNRAS, 394, L89Feruglio C., Ferrara A., Bischetti M., Fiore F., Downes D.,Ceccarelli C., et al. 2017, ArXiv e-prints, 1706.05527Fiore F., Feruglio C., Shankar F., Bischetti M., BongiornoA., Brusa M., Carniani S., Cicone C., Duras F., LamastraA., Mainieri V., Marconi A., Menci N., Maiolino R.,Piconcelli E., Vietri G., Zappacosta L., 2017, A&A, 601,A143Fluetsch A., Maiolino R., Carniani S., Arribas S., BelfioreF., Bellocchi E., Cazzoli S., Cicone C., Cresci G., FabianA. C., Gallagher R., Ishibashi W., Mannucci F., MarconiA., Perna M., Sturm E., Venturi G., 2020, arXiv e-prints,p. arXiv:2006.13232Fluetsch A., Maiolino R., Carniani S., Marconi A., CiconeC., Bourne M. A., Costa T., Fabian A. C., Ishibashi W.,Venturi G., 2019, MNRAS, 483, 4586Girichidis P., Naab T., Hanasz M., Walch S., 2018,MNRAS, 479, 3042Gonz´alez-Alfonso E., Fischer J., Spoon H. W. W., StewartK. P., Ashby M. L. N., Veilleux S., Smith H. A., SturmE., et al. 2017, ApJ, 836, 11Gowardhan A., Spoon H., Riechers D. A., Gonz´alez-Alfonso E., Farrah D., Fischer J., Darling J., Fergulio C.,Afonso J., Bizzocchi L., 2018, ArXiv e-printsHagino K., Done C., Odaka H., Watanabe S., TakahashiT., 2017, MNRAS, 468, 1442Herrera-Camus R., Janssen A., Sturm E., Lutz D., VeilleuxS., Davies R., Shimizu T., Gonz´alez-Alfonso E., RupkeD. S. N., Tacconi L., Genzel R., Cicone C., Maiolino R.,Contursi A., Graci´a-Carpio J., 2020, A&A, 635, A47Herrera-Camus R., Tacconi L., Genzel R., et al. 2019, ApJ,871, 37Ishibashi W., Fabian A. C., 2015, MNRAS, 451, 93Ishibashi W., Fabian A. C., 2016, MNRAS, 463, 1291Ishibashi W., Fabian A. C., 2018, MNRAS, 481, 4522Ishibashi W., Fabian A. C., Maiolino R., 2018, MNRAS,476, 512Ishibashi W., Fabian A. C., Ricci C., Celotti A., 2018,MNRAS, 479, 3335King A., Pounds K., 2015, ARA&A, 53, 115King A. R., Pringle J. E., 2007, MNRAS, 377, L25Koyama K., 2018, PASJ, 70, R1 Laha S., Guainazzi M., Piconcelli E., Gand hi P., Ricci C.,Ghosh R., Markowitz A. G., Bagchi J., 2018, ApJ, 868,10Lansbury G. B., Banerji M., Fabian A. C., Temple M. J.,2020, MNRAS, 495, 2652Mattsson L., 2020, MNRAS, 491, 4334Menci N., Fiore F., Feruglio C., Lamastra A., Shankar F.,Piconcelli E., Giallongo E., Grazian A., 2019, ApJ, 877,74Micha(cid:32)lowski M. J., 2015, A&A, 577, A80Mizumoto M., Izumi T., Kohno K., 2019, ApJ, 871, 156Murray N., Quataert E., Thompson T. A., 2005, ApJ, 618,569Nardini E., Reeves J. N., Gofford e. a., 2015, Science, 347,860Owen P. J., Barlow M. J., 2015, ArXiv e-printsReeves J. N., Braito V., 2019, ApJ, 884, 80Sanders D. B., Soifer B. T., Elias J. H., Madore B. F.,Matthews K., Neugebauer G., Scoville N. Z., 1988, ApJ,325, 74Sirressi M., Cicone C., Severgnini P., Braito V., Dotti M.,Della Ceca R., Reeves J. N., Matzeu G. A., Vignali C.,Ballo L., 2019, MNRAS, 489, 1927Smith R. N., Tombesi F., Veilleux S., Lohfink A. M.,Luminari A., 2019, ApJ, 887, 69Somalwar J., Johnson S. D., Stern J., Goulding A. D.,Greene J. E., Zakamska N. L., Alexandroff R. M., ChenH.-W., 2020, ApJ, 890, L28Stern J., Faucher-Gigu`ere C.-A., Zakamska N. L., HennawiJ. F., 2016, ApJ, 819, 130Sturm E., Gonzalez-Alfonso E., Veilleux S., Fischer J.,Gracia-Carpio J., Hailey-Dunsheath S., Contursi A.,Poglitsch A., Sternberg A., Davies R., Genzel R., LutzD., Tacconi L., Verma A., Maiolino R., de Jong J. A.,2011, ApJ, 733, L16Thompson T. A., Fabian A. C., Quataert E., Murray N.,2015, MNRAS, 449, 147Tombesi F., Veilleux S., Mel´endez M., Lohfink A., ReevesJ. N., Piconcelli E., Fiore F., Feruglio C., 2017, ApJ, 850,151Veilleux S., Bolatto A., Tombesi F., Mel´endez M., SturmE., Gonz´alez-Alfonso E., Fischer J., Rupke D. S. N., 2017,ApJ, 843, 18Veilleux S., Maiolino R., Bolatto A. D., Aalto S., 2020,A&A Rev., 28, 2Veilleux S., Mel´endez M., Sturm E., et al. 2013, ApJ, 776,27Wesson R., Barlow M. J., Matsuura M., Ercolano B., 2015,MNRAS, 446, 2089Zou F., Brandt W. N., Vito F., Chen (???) C.-T., GarmireG. P., Stern D., Ayubinia A., 2020, MNRAS, 499, 1823Zubovas K., 2018, MNRAS, 473, 3525Zubovas K., King A., 2012, ApJ, 745, L34 ©000