On the X-ray low- and high-velocity outflows in AGNs
aa r X i v : . [ a s t r o - ph . C O ] O c t Mon. Not. R. Astron. Soc. , 1– ?? (2011) Printed 21 September 2018 (MN L A TEX style file v2.2)
On the X-ray low- and high-velocity outflows in AGNs
J.M. Ram´ırez ⋆ and F. Tombesi , Leibniz-Institut f¨ur Astrophysik Potsdam, An der Sternwarte 16 D-14482 Potsdam, Germany Department of Astronomy and CRESST, University of Maryland, College Park, MD 20742, USA Laboratory for High Energy Astrophysics, NASA/Goddard Space Flight Center, Greenbelt, MD 20771, USA
Accepted . Received ; in original form
ABSTRACT
An exploration of the relationship between bolometric luminosity and outflow velocity,for two classes of X-ray outflows in a large sample of active galactic nuclei has beenperformed. We find that line radiation pressure could be one physical mechanism thatmight accelerate the gas we observe in warm absorber, v ∼ − − , and oncomparable but less stringent grounds the ultra-fast outflows (UFOs), v ∼ . − . c .If comparable with the escape velocity of the system; the first is naturally located atdistances of the dusty torus, ≈ ≈ .
01 pc,in accordance with large set of observational evidence existing in the literature. Thepresentation of this relationship might give us key clues for our understanding of thedifferent physical mechanisms acting in the center of galaxies, the feedback processand its impact on the evolution of the host galaxy.
Key words: black hole physics - X-ray: galaxies - galaxies: active
Mildly relativistic and non-relativistic absorption troughsare observed in the X-ray spectra of active galactic nu-clei (AGNs). A qualitative separation is usually done be-tween classical v ∼ − − warm absorbers,we refer to here (throughout paper) as low-velocity outflows(Blustin et al. 2005; McKernan et al. 2007), and ultra-fastoutflows (UFOs) v > ,
000 km s − , we refer to here as high-velocity outflows (Tombesi et al. 2010a,b).On one hand, the extracted velocity from the classical v ∼ − − warm absorber is mainly obtainedusing the Fe M-shell 2 p − d unresolved transition array(UTA) (Behar et al. 2001; Ram´ırez et al. 2008), and O vii or O viii resonance lines (e.g., George et al. 1998, the hall-mark of the classical warm absorber). The spatial locationof this absorbing material is uncertain. Models of X-ray ab-sorbers in AGN place them at a wide range of distancesfrom the central source. Specifically they are suggested tobe winds originating from the accretion disk (Murray et al.1995; Elvis 2000), located at the dusty ( ∼ E > ⋆ E-mail:[email protected] xxv-xxvi associated with a zone of cir-cumnuclear gas photoionized by the central X-ray source,with ionization parameter log ξ ∼ − N H ∼ − cm − . The energies ofthese absorption lines are systematically blueshifted and thecorresponding velocities can reach up to mildly relativis-tic values of ∼ . c − . c . These findings are importantbecause they suggest the presence of previously unknownmassive and highly ionized absorbing material outflowingfrom their nuclei, possibly connected with accretion diskwinds/ejecta (e.g., King & Pounds 2003; Proga & Kallman2004; Sim et al. 2008; Ohsuga et al. 2009; King 2010b;Sim et al. 2010).Several acceleration mechanisms have been pro-posed to explain these outflows: (1) thermally acceler-ated winds (Krolik & Kriss 2001); (2) radiation pressurethrough Thomson scattering and magnetic forces (MHD,Ohsuga et al. 2009); (3) and radiation pressure due to theabsorption of spectral lines (e.g., Proga & Kallman 2004;Ram´ırez 2008; Schurch et al. 2009; Sim et al. 2010; Ram´ırez2011). Although the first one can explain the velocities weobserve in the low-velocity outflows, it can be excluded be-cause it can not explain the ∼ . − . c we observe inUFOs (Tombesi et al. 2010a,b). Ohsuga et al. (2009) seemto reproduce the velocities observed in low- and high-velocity outflows. On the other hand, Arav et al. (1994); Ram´ırez(2008); Saez et al. (2009); Chartas et al. (2009), invoke ra-diation pressure due to lines to explain the ∼ . c outflow c (cid:13) Ram´ırez and Tombesi detected in a good S/N X-ray spectrum of a high-z quasar(the broad absorption line [BAL], APM 08279+5255), andthey reproduce, as part of the procedure, the Fe xxv-xxvi lines detected at
E > low- and high-velocity outflows,using an anisotropic radiative pressure framework (e.g.,Proga et al. 2000; Proga & Kallman 2004; Liu & Zhang2011), beginning with line radiation pressure.We present the observables from which we build themodel in §
2. The details of the proposed model are presentedin §
3. The results and the discussion are in §
4. We summarizein § In this section we present the observables of the two typesof outflows, since they are the initial motivation for the pro-posed model.
When describing the physical conditions of warm absorbers,it is common to use the definition of ionization parame-ter ξ = πF ion n H (Tarter et al. 1969), where F ion is the totalionizing flux ( F ion = L ion / πr ), and n H is the gas den-sity. The source spectrum is described by the spectral (spe-cific, energy dependent ǫ ) luminosity L ǫ = L ion f ǫ , where L ion is the integrated luminosity from 1 to 1000 Ryd, and R f ǫ dǫ = 1.So we describe the ∼ − warm absorber out-flows as absorbing material around a supermassive blackhole (SMBH) with mass M BH ∼ × (Peterson et al.2004; Blustin et al. 2005), column density of the absorb-ing material N H ∼ − cm − , flowing outwards atvelocities v ∼ − − (e.g., Kaspi et al. 2002;Krongold et al. 2003, 2005; Ram´ırez et al. 2005), at mediumionization states log ξ ∼ − − cm (Blustin et al.2005).When computing the energetics they find mass lossrates of ˙ M out ≈ . ⊙ yr − ratios of ˙ M out to ac-cretion rates , ˙ M out / ˙ M acc ≈ L EK = ˙ M out v , of the orders of 10 − erg/s, represent-ing less than 1 % of the bolometric luminosity (Blustin et al.2005). The main conclusion from these estimations is thatthese outflows contributes little to the energy injected in tothe host galaxy. But the amount of matter processed overthe AGN lifetime can be significant (also in accordance withKrongold et al. 2007, for instance). Mean of the ˙ M out reported by Blustin et al. (2005) in theirTable 4, excluding Ark 564 (outlier ˙ M out = 23 M ⊙ yr − ). Mean of the ˙ M out / ˙ M acc reported by Blustin et al. (2005) intheir Table 4, excluding Ark 564 (outlier ˙ M out / ˙ M acc = 550). Excluding PG 0844+349 and PG 1211+143.
The characteristics of the ultra-fast outflows with v > − ( > xxv-xxvi absorption lines detected by Tombesi et al.(2010b) in a complete sample of local Seyfert galaxies. Suchfeatures are detected in ∼ C ∼ ∼ ∼ ∼ ξ ∼ − cm,and have large column densities, in the range N H ∼ –10 cm − .The SMBH masses of the Seyferts in the Tombesi et al.(2010b) sample have a mean value of M BH = 5 . × M ⊙ (Marchesini et al. 2004; Peterson et al. 2004).When computing the energetics of these outflows, wesee that these are more massive than the low-velocity ones,˙ M out ∼ . − M ⊙ yr − ∼ ˙ M acc , and also much more pow-erful, with a mechanical power of ∼ –10 erg/s (e.g.,Pounds et al. 2003; Markowitz et al. 2006; Braito et al. 2007;Cappi et al. 2009; Tombesi et al. 2010b). The latter value is ∼ −
10% of the bolometric luminosity. Therefore, the high-velocity outflows may potentially play an important role onthe expected cosmological feedback from AGNs (e.g., King2010a,b).
We build our model based on observational evidence, basi-cally from two sources: Blustin et al. (2005) for the classi-cal v ∼ − − warm absorbers, referred hereas low-velocity outflows, and Tombesi et al. (2010b) for theultra-fast outflows (UFOs) v > ,
000 km s − , referred hereas high-velocity outflows.As we see, together the two classes of outflows covera wide range in velocity and black hole masses. We seek aphysical model which provides the context to explain (tofirst approximation), part of the observational evidence wehave up to now.In order to do that, and to gain some insight into therelationship between outflow velocity ( v out ) and luminosity,we invoke the velocity profile as a function of radius, given byhydrodynamical calculations (Proga et al. 2000). The math-ematical shape of the relation is given by assuming an out-flow accelerated by radiation pressure from a central sourcewith a bolometric luminosity L bol , and a mass of M BH : v out [ i ] = (cid:20) GM BH [ i ] (cid:18) Γ f L bol [ i ] L Edd [ i ] − (cid:19) (cid:18) R in − R (cid:19)(cid:21) / , (1)where L Edd is the Eddington luminosity, Γ f is the force mul-tiplier, where the acceleration due to the absorption of dis-crete lines is encapsulated (Laor & Brandt 2002), R in is theradius at which the wind is launched from the disk, and R isthe distance of the accelerated portion of the outflow fromthe central source. The index i , runs simultaneously overthe two distributions: velocity/BH mass. Assuming that weobserve the gas when it has reached the terminal velocity ofthe wind, i.e. v out at R = ∞ , we can write: c (cid:13) , 1– ?? ow- and high-velocity outflows L bol [ i ] = L Edd [ i ]Γ f (cid:18) v out [ i ] R in GM BH [ i ] + 1 (cid:19) . (2)Now we have the basic ingredient of the model, and weare ready to build our two classes of outflows. set 1 : High-velocity . This synthetic sample is composedfrom 1000 black holes with masses normally distributedwith mean, µ m = 2 . × M ⊙ and standard deviation, σ m = 1 . × M ⊙ . Also we use a normal distributionfor the outflow velocity of the absorbing gas around the BHwith µ v = 57 ,
000 km s − (this is the best-fit value for set 1 ),and σ v = 15 ,
000 km s − . set 2 : Low-velocity . In this case we use 1000 black holeswith masses normally distributed with the same mean, µ m =2 . × M ⊙ and standard deviation, σ m = 1 . × M ⊙ as before but we use a normal distribution for the outflowvelocity of the absorbing gas around the BH with µ v = 1800km s − (this is the best-fit value for set 2 ), and σ v = 600km s − .The best-fit values of µ v (1) and µ v (2) (as well Γ f (1) =250 and Γ f (2) = 40, see next section), are estimated by com-paring a grid of models computed using 10 , µ v (1) ,
000 (km s − ) and 100 µ v (2) − ), withthe corresponding set of observation: set 3 vs Tombesi et al.(2011) and set 4 vs Blustin et al. (2005). We performedthe comparison using a Kolmogorov-Smirnov test. We takethe µ v (1) and µ v (2) (as well Γ f (1) = 250 and Γ f (2) = 40,see next section) which give maximum p-values of the test.The final p-value for the comparison set 3 vs Tombesi et al.(2011) gives ≈ . D = 0 .
16) and for the set 4 vs Blustin et al. (2005) ≈ .
07 ( D = 0 . R in , for each set of simulations isbased on the results of Blustin et al. (2005) for set 2 ; i.e., R in ( set
2) = 1 pc (orders of magnitude value). For set 1 , weset R in ( set
1) = 0 . pc , based on Tombesi et al. (2010a,b). In this section we discuss the results of our simulated sam-ples and summarize their physical context and implications.Figure 1 shows the theoretical relationship between out-flow velocity and bolometric luminosity.The curvature we see in v out of set 2 (and also set 1 ,filled circles, our line of sight is along the flow, see below), isdue to the quadratic dependence of the luminosity with v out given by equation 2. It is worth noticing that if we assumea force multiplier of Γ f = 40 (this is the best-fit value for set 2 and set 4 , see section 3) ; we are able to reproducethe range of luminosities seen in Figure 4 of Blustin et al.(2005).On the other hand this model is not reproducing wellthe dispersion we observe in Figure (4) of Blustin et al.(2005), which could be explained by the fact that equation2 will give radially accelerated flows, or in other words, that We take the middle point ( µ m = 2 . × M ⊙ ) between themean extracted from Tombesi et al. (2011) ( µ m = 5 . × M ⊙ ),excluding the mass of Mrk 205 (outlier M BH = 44 × M ⊙ ), andthe mean extracted from Table 5 of Blustin et al. (2005) ( µ m =2 . × M ⊙ ), excluding the mass of IRAS 13349+2438 (outlier M BH = 80 × M ⊙ ). Also we take the larger value of theobserved dispersion σ m ( obs ) = 1 . × M ⊙ . + + + Set 1 L bol (erg/s) v ou t ( k m / s ) Radial flowsRandom l−o−s
Set 2 L bol (erg/s) v ou t ( k m / s ) Radial flowsRandom l−o−s
Figure 1.
Velocity of the outflow vs the bolometric luminos-ity needed to accelerate the wind. Filled circles are the velocitiesobserved if we see the wind along the flow. Open squares areincluding the effects of random line-of-sights (see text for discus-sion). we are observing all the objects radially along the flow. Toinclude the effect of observing random line-of-sights trans-versely through different sections of the flow we connect theobserved velocity v obs with the intrinsic radial velocity, i.e. v obs = v out cos θ , where θ is the angle between the outflowdirection and the line of sight. Then using the same luminosi-ties produced by set 2 , but plotting (open squares) against1000 v obs (s) randomly computed using a random generatorof numbers (with π/ θ π/ , and equation 2 with v obs = v out cos θ , we produce the sub-sample set 4 . It is easyto see that set 4 , resembles better the plot shown in Figure4 of Blustin et al. (2005), taking into account the possibleincompleteness of the sample.Doing the same for the UFOs, we use the luminositiesproduced by set 1 (Γ f = 250, this is the best-fit value for set 1 and set 3 , see section 3) but plotting (open squares)against 1000 v obs (s) randomly computed using a randomgenerator of numbers (with π/ θ π/ v obs = v out cos θ , to produce the sub-sample set 3 .In this case set 3 , better resembles the relationship between v out and luminosity (see Figure 2).In Figure 2 we place both classes of outflows togetheralong with observational points taken from two samplesof objects: 14 points out of the 23 objects reported byBlustin et al. (2005) (the others either did not report out-flow velocity or were unknown), in the right panel; and 19points from the sample of 19 objects were UFOs have beendetected by Tombesi et al. (2011), in the left panel.There are several interesting facts about the plot: (1) the observational points cover ≈ . ≈ ∼ − erg/s); and (3) thatour proposed model is able to reproduce most of the low-velocity points (11/14) (two of the points might fall in thehigh-velocity set instead) and, less stringently, half of thepoints for the high-velocity set (11/19), using one physi-cal acceleration mechanism and three free parameters: mass The lower limit assumes that the torus cover ∼
30 degrees, sowe are able to observe the outflow only from θ > π/ (cid:13) , 1– ?? Ram´ırez and Tombesi + + + L bol (erg/s) v ob s ( k m / s ) Tombesi et al. 2011Set 3 L bol (erg/s) v ob s ( k m / s ) Blustin et al. 2005Set 4
Figure 2.
Theoretical vs observational outflow velocity againstluminosity. Ellipses (1,2 and 3 σ contours of the model) are the-oretical calculations coming from two samples of outflows; Leftpanel: high-velocity ( set 3 , i.e., including the effects of randomline-of-sights), and
Right panel: low-velocity ( set 4 , i.e., includ-ing the effects of random line-of-sights). Open squares are ob-servational points from the XMM-Newton radio-quite sample ofTombesi et al. (2011), for the UFOs. The open diamonds arepoints compiled by Blustin et al. (2005). of the BH ( M BH ), outflow velocity ( v out ) and force multi-plier (Γ f ). On the other hand, there are two sectors wherethe deviations between model and theory are large: (1) thehigh-velocity/low luminosity; and (2) the low-velocity/highluminosity, both requiring a closer inspection to the com-pleteness of the samples, and/or the addition of other accel-eration mechanisms, like magnetic thrust for instance. But,this may be the topic of future work. We propose here that the low- and high-velocity outflowscan be accelerated by the same physical mechanism (as issuggested by Figure 2); i.e., radiation pressure, and thatthe differences depend on the values of the 3 fundamentalparameters: mass of the BH ( M BH ), the object luminosity( L bol ) and force multiplier (Γ f ). The values of Γ f we haveused are of the orders of those found in observational (e.g.,Laor & Brandt 2002) and theoretical (e.g., Saez et al. 2011)works . However, detailed photoionization computations arerequired to verify if the opacity of the gas under the physicalconditions presented here can overcome the over-ionizationproblem (Proga et al. 2000), and are left for future work.The anisotropic property of the radiation is basically de-manded by the existence of obscured and un-obscured (Type1 and Type 2) AGNs (Antonucci 1993), and it explains thedecreasing of Type 2 AGNs as a function of the X-ray Lu-minosity (Hasinger 2008). It is also intrinsically linked tothe existence of the dusty torus and it gives the natu-ral frame to locate our two classes of outflows. If compara-ble with the escape velocity of the system (a black hole of M BH = 5 . × M ⊙ ), fast-outflow, v out ∼ . A convenient definition is that of cool ( ∼ ∼ escape radius at r ∼ (20 − r S (Schwarzschild radius, r S = GM BH c ). This is very close to the SMBH, and theorigin is likely the accretion disk.Again, using 500 − − , as escape velocity froma 2 . × M ⊙ BH, locates the gas at ∼ . − For the first time, an exploration on the relationship betweenbolometric luminosity and outflow velocity, for two classes ofX-ray outflows in a large sample of active galactic nuclei hasbeen performed. We find that: (1) Line radiation pressureis an efficient mechanism to accelerate the low-velocity(500 − − ) gas we observe in the classical ∼ − warm absorber. (2) It might also become efficientto accelerate the high-velocity (0 . − . c ) gas we observein the UFOs. (3) They both might be placed in the samecontext of anisotropic radiation pressure (e.g., Proga et al.2000; Proga & Kallman 2004; Liu & Zhang 2011).However, there are many open questions, which will re-quire close investigation and detailed modeling. In the firstplace the fact that we are assigning one type of outflow todifferent portions of the parameter space ( M BH , L bol andΓ f ), does not preclude the existence of both in the sameobject. Also, careful studies of the connection between theseoutflows observed in X-rays and outflows seen in other bandsof the electromagnetic spectrum, UV, infrared or optical(multi-wavelength studies) are important points to be ad-dressed. Finally the close inspection of the main parametersof the system ( M BH , v out and Γ f ) with cosmological param-eters (like redshift) is of high relevance, and might be thetopic of future works. ACKNOWLEDGMENTS
JMR would like to thank T. Kallman for a reading of themanuscript. He also wants to thank the useful and con-structive comments from the referee which helped to im-prove several aspects of the work. FT provided the databased on observations obtained with the XMM-Newtonsatellite, an ESA funded mission with contributions by ESAmember states and USA. FT acknowledge support fromNASA through the ADAP/LTSA program. This researchhas made use of the NASA/IPAC Extragalactic Database(NED) which is operated by the Jet Propulsion Laboratory,California Institute of Technology, under contract with theNational Aeronautics and Space Administration.
REFERENCES
Antonucci R., 1993, ARAA, 31, 473Arav N., Li Z., Begelman M. C., 1994, ApJ, 432, 62Behar E., Sako M., Kahn S. M., 2001, ApJ, 563, 497 Based on observed force multipler (Laor & Brandt 2002).c (cid:13) , 1– ?? ow- and high-velocity outflows Blustin A. J., Page M. J., Fuerst S. V., Branduardi-Raymont G., Ashton C. E., 2005, A&A, 431, 111Braito V., Reeves J. N., Dewangan G. C., George I., Grif-fiths R. E., Markowitz A., Nandra K., Porquet D., PtakA., Turner T. J., Yaqoob T., Weaver K., 2007, ApJ, 670,978Cappi M., Tombesi F., Bianchi S., Dadina M., Giustini M.,Malaguti G., Maraschi L., Palumbo G. G. C., PetrucciP. O., Ponti G., Vignali C., Yaqoob T., 2009, A&A, 504,401Chartas G., Brandt W. N., Gallagher S. C., 2003, ApJ, 595,85Chartas G., Brandt W. N., Gallagher S. C., Garmire G. P.,2002, ApJ, 579, 169Chartas G., Saez C., Brandt W. N., Giustini M., GarmireG. P., 2009, ApJ, 706, 644Dadina M., Cappi M., Malaguti G., Ponti G., de Rosa A.,2005, A&A, 442, 461Elvis M., 2000, ApJ, 545, 63George I. M., Turner T. J., Netzer H., Nandra K.,Mushotzky R. F., Yaqoob T., 1998, ApJS, 114, 73Hasinger G., 2008, A&A, 490, 905Kaspi S., Brandt W. N., George I. M., Netzer H., Cren-shaw D. M., Gabel J. R., Hamann F. W., Kaiser M. E.,Koratkar A., Kraemer S. B., Kriss G. A., Mathur S.,Mushotzky R. F., Nandra K., Peterson B. M., ShieldsJ. C., Turner T. J., Zheng W., 2002, ApJ, 574, 643King A. R., 2010a, MNRAS, 408, L95King A. R., 2010b, MNRAS, 402, 1516King A. R., Pounds K. A., 2003, MNRAS, 345, 657Krolik J. H., Begelman M. C., 1988, ApJ, 329, 702Krolik J. H., Kriss G. A., 2001, ApJ, 561, 684Krongold Y., Nicastro F., Brickhouse N. S., Elvis M.,Liedahl D. A., Mathur S., 2003, ApJ, 597, 832Krongold Y., Nicastro F., Elvis M., Brickhouse N., BinetteL., Mathur S., Jim´enez-Bail´on E., 2007, ApJ, 659, 1022Krongold Y., Nicastro F., Elvis M., Brickhouse N. S.,Mathur S., Zezas A., 2005, ApJ, 620, 165Laor A., Brandt W. N., 2002, ApJ, 569, 641Laor A., Draine B. T., 1993, ApJ, 402, 441Liu Y., Zhang S. N., 2011, ApJ, 728, L44+Marchesini D., Celotti A., Ferrarese L., 2004, MNRAS, 351,733Markowitz A., Reeves J. N., Braito V., 2006, ApJ, 646, 783McKernan B., Yaqoob T., Reynolds C. S., 2007, MNRAS,379, 1359Murray N., Chiang J., Grossman S. A., Voit G. M., 1995,ApJ, 451, 498Ogle P. M., Marshall H. L., Lee J. C., Canizares C. R.,2000, ApJ, 545, L81Ohsuga K., Mineshige S., Mori M., Kato Y., 2009, PASJ,61, L7+Peterson B. M., Ferrarese L., Gilbert K. M., Kaspi S.,Malkan M. A., Maoz D., Merritt D., Netzer H., OnkenC. A., Pogge R. W., Vestergaard M., Wandel A., 2004,ApJ, 613, 682Pounds K. A., King A. R., Page K. L., O’Brien P. T., 2003,MNRAS, 346, 1025Proga D., Kallman T. R., 2004, ApJ, 616, 688Proga D., Stone J. M., Kallman T. R., 2000, ApJ, 543, 686Ram´ırez J. M., 2008, A&A, 489, 57Ram´ırez J. M., 2011, RMxAA, 47, 385 Ram´ırez J. M., Bautista M., Kallman T., 2005, ApJ, 627,166Ram´ırez J. M., Komossa S., Burwitz V., Mathur S., 2008,ApJ, 681, 965Reeves J. N., O’Brien P. T., Braito V., Behar E., MillerL., Turner T. J., Fabian A. C., Kaspi S., Mushotzky R.,Ward M., 2009, ApJ, 701, 493Saez C., Chartas G., Brandt W. N., 2009, ApJ, 697, 194Saez C., & Chartas G., 2011, ApJ, 737, 91Schurch N. J., Done C., Proga D., 2009, ApJ, 694, 1Sim S. A., Long K. S., Miller L., Turner T. J., 2008, MN-RAS, 388, 611Sim S. A., Miller L., Long K. S., Turner T. J., Reeves J. N.,2010, MNRAS, 404, 1369Sim S. A., Proga D., Miller L., Long K. S., Turner T. J.,2010, MNRAS, 408, 1396Tarter C. B., Tucker W. H., Salpeter E. E., 1969, ApJ, 156,943Tombesi F., Cappi M., Reeves J. N., Palumbo G. G. C.,Yaqoob T., Braito V., Dadina M., 2010, A&A, 521, A57+Tombesi F., Sambruna R. M., Reeves J. N., Braito V., BalloL., Gofford J., Cappi M., Mushotzky R. F., 2010, ApJ,719, 700Tombesi F., Cappi M. & Reeves J. N., 2011,arXiv:1109.2882 c (cid:13) , 1–, 1–