The role of failed accretion disk winds in active galactic nuclei
aa r X i v : . [ a s t r o - ph . H E ] F e b Nuclear Activity in Galaxies Across Cosmic TimeProceedings IAU Symposium No. 356, 2019M. Povic et al., eds. c (cid:13) The Role of Failed Accretion Disk Windsin Active Galactic Nuclei
Margherita Giustini and Daniel Proga Centro de Astrobiolog´ıa (CSIC-INTA), Departamento de Astrof´ısica;Camino Bajo del Castillo s/n, Villanueva de la Ca˜nada, E-28692 Madrid, Spainemail: [email protected] Department of Physics & Astronomy, University of Nevada, Las Vegas;4505 South Maryland Parkway, NV 89154-4002 Las Vegas, USAemail: [email protected]
Abstract.
Both observational and theoretical evidence point at outflows originating from ac-cretion disks as fundamental ingredients of active galactic nuclei (AGN). These outflows canhave more than one component, for example an unbound supersonic wind and a failed wind(FW). The latter is a prediction of the simulations of radiation-driven disk outflows which showthat the former is accompanied by an inner failed component, where the flow struggles to escapefrom the strong gravitational pull of the supermassive black hole. This FW component couldprovide a physical framework to interpret various phenomenological components of AGN. Herewe briefly discuss a few of them: the broad line region, the X-ray obscurer, and the X-ray corona.
Keywords. accretion, accretion disks; black hole physics; galaxies: active
1. Introduction
The inner structure of luminous active galactic nuclei (AGN) is shaped by the presenceof winds launched on accretion disk scales, as indicated by their large terminal velocities( υ out ≫ − and up to several 0 . Here “successful” or “failed”means the ability of inability of the gas in reaching the escape velocity υ esc . In LD diskwind models, the FW is a fundamental ingredient of the mass flow of luminous AGNas the wind itself. In fact, its formation is more robust than the formation of the wind.Therefore if LD disk wind models hold, observable quantities related to the FW shouldbe identifiable in AGN as well. In the following section, we briefly discuss the possiblerole of the FW in the inner accretion flow of luminous AGN.
2. Failed line-driven accretion disk winds in the AGN inner structure
Line driving can deposit much more momentum in the gas than pure electron scatter-ing, allowing the launching of material with velocity greater than υ esc in sub-Eddingtonregimes (Castor, Abbott & Klein 1975). For LD to be effective, the presence of spectraltransitions is therefore fundamental. The spectral energy distribution of luminous AGNallows for LD winds to be launched from accretion disk scales pushing on the many UVtransitions available (e.g., Murray et al. 1995; Proga, Stone & Kallman 2000; Proga &Kallman 2004). A large X-ray flux is also characteristic of luminous AGN, and while the1 Margherita Giustini & Daniel ProgaX-ray photons will push on relatively few available X-ray lines, they will mostly concurto strip the electrons off the UV-absorbing atoms, thus destroying the many UV spectraltransitions available and effectively “overionizing” the wind material (e.g., Dannen et al.2019, and references therein). The term “overionization” is referred to a level of ioniza-tion that is too large to produce the observed UV lines and to sustain LD winds abovelocal υ esc . Therefore the ratio between the UV and X-ray radiation flux is crucial for thesuccessful launch and acceleration of LD winds in AGN.In the first models of LD accretion disk winds in AGN, a layer of dense gas (a shield ),absorbing the strong ionizing X-ray flux, was assumed to exist between the X-ray contin-uum source and the UV-absorbing wind. This is the “hitchhicking gas” of Murray et al.1995, which postulated the presence of this gas just in front of the flow that is effectivelyaccelerated out of the system. It was speculated that a gradient in pressure would thencause the hitchhicking gas to accelerate together with the farther out wind (hence thenickname).Hydrodynamical simulations performed by Proga and collaborators (Proga, Stone &Kallman 2000; Proga & Kallman 2004) showed that for massive (black hole mass M BH > M ⊙ ), luminous AGN (Eddington ratio ˙ m > . υ esc before getting overionized, and itfalls back toward the disk. The failure or success of the wind is measured in terms ofovercoming or not the local υ esc . The inner FW effectively protects (shields) the materiallocated farther out from the strong ionizing continuum radiation, therefore allowing forthe successful launch of the wind at radii larger than where the FW dominates, andwhere the radiation pressure is large enough to overcome BH gravity (Proga & Kallman2004; Risaliti & Elvis 2010).In the scenario proposed by Giustini & Proga (2019), for M BH & M ⊙ and ˙ m & . The FW as BLR
The broad line region (BLR) is a fundamental ingredient of luminous AGN: it consists ofgas photoionized by the AGN continuum and whose motion responds to the gravitationalpotential of the central SMBH (e.g., Peterson et al. 2004). The BLR is phenomenologicallydivided into a low-ionization (e.g., Mg II, H β ) and a high-ionization (e.g., C IV, Ly α )component which show distinct kinematics (e.g., Marziani et al. 1996). In particular, adifference in peak position between low-ionization and high-ionization emission lines isindicative of strong radial motions of the gas producing the latter (Gaskell 1982).The high-ionization emission lines in luminous AGN can in fact be blueshifted byseveral hundreds (up to thousands) km s − with respect to the low-ionization emission he Failed Wind in Active Galactic Nuclei † are also observedin a large number of luminous AGN (up to 40%, Allen et al. 2011). These are the so-calledbroad absorption line quasars (BAL QSOs), and display the most direct evidence for thepresence along the line of sight of strong winds, which can reach velocities of several0 . m & .
01, when the inner accretion and ejection flow consists ofa disk, LD wind, FW, and inner X-ray source; the LD wind + FW then produce thehigh-ionization BLR. In particular, the production of the symmetric portion of the high-ionization BLR is associated to the FW, while its blueshifted and blue-skewed portion,to the wind itself.The strongest (fastest, densest) LD disk winds are those produced in AGN with a lowX-ray/UV flux ratio (
X-ray weak ), either because of high ˙ m and/or a large M BH . Thesehave a vast radial zone of the inner flow dominated by winds, and only a small innerregion where the wind fails. Thus they produce winds with a large range of velocities,including large terminal velocities ≫ ,
000 km s − and up to several 0 .
1c when launchedin the innermost regions of the disk. On the contrary, in the case of AGN with a large X-ray/UV flux ratio (
X-ray bright ), a larger inner region of the accretion flow is dominatedby the FW. In these AGN, successful winds are only launched at larger scales, thusreaching lower terminal velocities.In LD disk winds scenarios, the BLR of X-ray weak AGN is dynamically dominatedby the wind: the emission lines of e.g. C IV are strongly blueshifted and blue-skewed.Their equivalent width is lower than the one of the same emission lines produced in X-raybright AGN: here the dynamics of the BLR is dominated by the FW, that does not reach υ esc . The emission lines have a larger equivalent width, a more symmetric profile, andlittle or no blueshift with respect to the redshift of the host galaxy. In other words, theFW extent regulates the extent of the symmetric, non-shifted BLR at the expense of theskewed, blueshifted BLR produced in the wind. When the disk is observed through thewind, X-ray weak AGN will display deeper, broader, and more blueshifted absorptiontroughs compared to X-ray bright AGN. Broadly speaking, the first type of AGN wouldcorrespond to the population A of quasars along their main sequence, while the secondtype of AGN to their population B (Sulentic et al. 2000; Sulentic & Marziani 2015). † Low-ionization broad absorption lines are observed in a small fraction (about 5-10%) ofbroad absorption line quasars, the low-ionization BAL QSOs; those who do not display themare called high-ionization broad absorption line quasars.
Margherita Giustini & Daniel Proga2.2.
The FW as the obscurer
The presence of dense, variable layers of X-ray absorbing gas on BLR-scales has beenrecently inferred by high-quality observations of local Seyfert 1 galaxies (Kaastra et al2014; Ebrero et al. 2016; Mehdipour et al. 2017; Kriss et al. 2019). This gas is called“obscurer”, as it absorbs the X-ray continuum flux and obscures the view of the strongionizing X-ray continuum to the material located further out. This further out material,in fact, responds to the changes in X-ray ionizing flux, as strong UV absorption lines areobserved emerging in concomitance with the appearance of strong X-ray absorption.In LD accretion disk wind scenarios, the FW is the material located close to thesource of X-rays, that gets all the ionizing continuum, and thus fails reaching υ esc andfalls back toward the disk. The FW motion is complex: highly dynamical, with locallyvariable motion made of upward and downward components, and dense filaments andknots embedded in a much hotter medium (Proga 2005). The FW absorbs the X-raycontinuum photons, effectively shielding the gas located farther out that can be thenaccelerated by radiation pressure on UV spectral lines. The FW has therefore all thecharacteristics to be identified with the “obscurer” of local Seyfert galaxies.2.3. The FW as the X-ray warm corona(e)
Much closer to the central SMBH than the wind, but maybe partially co-spatial with theinner FW, lies the source of X-ray photons. The X-ray radiation is the clearest signatureof accreting BHs, yet its physical origin is still unclear. We know that some compact andhot region must be responsible for the bulk of the intense and variable X-ray emission ofAGN, and we call it X-ray “hot corona”. The X-ray hot corona has become a synonymfor a low-density (optical depth τ ≪
1) medium full of hot (temperature kT e > kT e ∼ −
300 eV, τ ∼ E &
3. Conclusions
Accretion disk winds have been recognized as fundamental ingredients of the innerregions of luminous AGN. In the case of LD disk winds, the inner FW component mighthelp interpreting in a physical framework phenomenological features of AGN such as thehigh-ionization BLR, the obscurer, and the X-ray coronae. The FW solutions of the inneraccretion and ejection flow of AGN deserve further attention, in order to assess whetherthey can change significantly the physical and geometrical structure of the very inneraccretion flow around highly accreting SMBHs. he Failed Wind in Active Galactic Nuclei
4. Acknowledgements
MG warmly thanks the IAU 356 symposium organizers for a memorable, transforma-tional meeting. We thank G. Richards, G. Miniutti, E. Lusso, and M. Mehdipour forinteresting discussions. MG is supported by the “Programa de Atracci´on de Talento” ofthe Comunidad de Madrid, grant number 2018-T1/TIC-11733, and by the Spanish StateResearch Agency (AEI) Projects number ESP2017-86582-C4-1-R and ESP2015-65597-C4-1-R. This research has been partially funded by the AEI Project number MDM-2017-0737 Unidad de Excelencia “Mar´ıa de Maeztu” - Centro de Astrobiolog´ıa (INTA-CSIC).
References
Allen, J. T., Hewett, P. C., Maddox, N., Richards, G. T., et al. 2011,
MNRAS ,4105, 860Castor, J. I., Abbott, D. C., & Klein, R. I. 1975,
ApJ , 195, 157Czerny, B., & Hryniewicz, K. 2011,
A&A , 525, 8Dannen, R. C., Proga, D., Kallman, T. R., & Waters, T. 2019,
ApJ , 882, 99Done, C., Davis, S. W., Jin, C., Blaes, O., & Ward, M. 2012,
MNRAS , 420, 1848Ebrero, J., Kriss, G. A., Kaastra, J. S., & Ely, J. C. 2016,
Ap&A , 586, 72Gaskell, C. M. 1982,
ApJ , 263, 79Giustini, M., & Proga, D. 2019,
A&A , 630, 94Haardt, F., & Maraschi, L. 1993,
ApJ , 413, 507Kaastra, J. S., Kriss, G. A., Cappi, M., Mehdipour, M., et al. 2014,
Science , 345, 64Kriss, G. A., Mehdipour, M., Kaastra, J. S., Rau, A., et al. 2019,
A&A , 621, 12Marziani, P., Sulentic, J. W., Dultzin-Hacyan, D., Calvani, M., et al. 1996,
ApJ , 104, 37Mehdipour, M., Kaastra, J. S., Kriss, G. A., Cappi, M., et al. 2015,
A&A , 575, 22Mehdipour, M., Kaastra, J. S., Kriss, G. A., Arav, N., et al. 2017,
A&A , 607, 28Murray, N., Chiang, J., Grossman, S. A., & Voit, G. M. 1995,
ApJ , 451, 498Peterson, B. M., Ferrarese, L., Gilbert, K. M., Kaspi, S., et al. 2004,
ApJ , 613, 682Petrucci, P. -O., Ursini, F., De Rosa, A., Bianchi, S., et al. 2018,
A&A , 611, 59Proga, D., Stone, J. M., & Kallman, T. R. 2000,
ApJ , 543, 686Proga, D., & Kallman, T. R. 2004,
ApJ , 616, 688Proga, D. 2005,
ApJ , 630, 9Rankine, A. L., Hewett, P. C., Banerji, M., & Richards, G. T. 2020,
MNRAS in press ,arXiv:1912.08700Richards, G. T., Kruczek, N. E., Gallagher, S. C., Hall, P. B., et al. 2011, AJ , 141, 167Risaliti, G, & Elvis, M. 2010, A&A , 516, 89Sulentic, J. W., Zwitter, T., Marziani, P., & Dultzin-Hacyan, D. 2000,
ApJ , 536, 5Sulentic, J. W., & Marziani, P. 2015,
FrASS , 2, 6Waters, T., Kashi, A., Proga, D., Eracleous, M., et al. 2016,