文档库 最新最全的文档下载
当前位置:文档库 › Probing Broad Absorption Line Quasar Outflows X-ray Insights

Probing Broad Absorption Line Quasar Outflows X-ray Insights

a r X i v :a s t r o -p h /0212304v 1 12 D e c 2002Probing Broad Absorption Line Quasar Out?ows:X-ray

Insights

S.C.Gallagher,1W.N.Brandt,2G.Chartas,2G.P.Garmire,2and R.M.Sambruna 3

1Center

for Space Research,Massachusetts Institute of Technology,77Massachusetts Avenue,Cambridge,MA 02139,USA 2Department of Astronomy &Astrophysics,The Pennsylvania State University,525Davey Laboratory,University Park,PA 16802,USA 3Department of Physics and Astronomy and School of Computational Sciences,MS 3F3,4400University Drive,George Mason University,Fairfax,VA 22030ABSTRACT Energetic out?ows appear to occur in conjunction with active mass accretion onto supermassive black holes.These out?ows are most readily observed in the ~10%of quasars with broad absorption lines,where the observer’s line of sight passes through the wind.Until fairly recently,the paucity of X-ray data from these objects was notable,but now sensitive hard-band missions such as Chandra and XMM-Newton are routinely detecting broad absorption line quasars.The X-ray regime o?ers qualitatively new information for the understanding of these objects,and these new results must be taken into account in theoretical modeling of quasar winds.INTRODUCTION Recent studies of the strong correlations between black hole masses and the properties of their hosts’galactic bulges (e.g.,Ferrarese and Merritt,2000;Gebhardt et al.,2000)demonstrate clearly the intimate interrelation between the growth of black holes and their host galaxies.This connection implies a physical mechanism for regulating the coeval development of these structures.During the bulk of their accretion phases,supermassive black holes apparently reveal themselves as luminous quasars (e.g.,Yu and Tremaine,

2002),and thus quasar winds are likely to provide a primary source of feedback.These winds are directly observed in the population of Broad Absorption Line (BAL)quasars,~10%of the quasar population that exhibit deep,broad absorption lines from high-ionization UV resonance transitions.Such blueshifted absorption features are understood to arise along lines of sight which pass through radiatively driven winds with terminal velocities reaching 0.1–0.3c .These energetic out?ows are important components of quasar environments;mass ejection is apparently fundamentally linked to active mass accretion.X-rays,generated in the innermost region surrounding accreting black holes,travel through the nuclear environments to the observer.X-ray studies of BAL quasars thus o?er a privileged view through the wind.

The challenge of accelerating gas to the high velocities observed in BAL quasars led Murray et al.(1995)to a picture similar to that presented in Figure 1.In this model,at radii starting at ~1016cm for a 108M ⊙black hole,radiation and gas pressure lift material from the disk photosphere where it is struck by light from the inner accretion disk and corona.The wind,initially co-rotating with the disk,is accelerated radially by the momentum of UV line photons;the material launched from the innermost radii reaches the highest terminal velocities.In order to launch the wind at such small distances,a thick layer of “shielding”gas,highly ionized plasma with N H =1022–1024cm ?2,was introduced to protect the wind from becoming completely ionized by soft X-rays.In the subsequent hydrodynamical modeling of line-driven quasar winds by Proga,Stone,and Kallman (2000),this layer of shielding gas arose naturally from the simulations.This scenario produces several predictions for X-ray observations of BAL quasars:

Fig. 1.Diagram of a possible geometry for the nucleus of a BAL quasar based on the Murray et al.(1995) model.The patterned gray circle represents the X-ray emitting region surrounding the black hole.The region of highly ionized plasma marked“Shielding Gas”was postulated to shield the BAL wind(with a scale of light days) from soft X-ray and extreme UV radiation.These photons would otherwise strip the gas in the wind of electrons, thus preventing it from being radiatively driven by UV line pressure.

?BAL quasars should have intrinsic X-ray continuum shapes(withΓ=2.0±0.3;e.g.,George et al., 2000)and spectral energy distributions typical of radio-quiet quasars.

?All BAL quasars should be X-ray weak,i.e.,have values ofαox<~?2.0,1as a result of large column densities of absorbing gas.Furthermore,with data of su?cient quality,the spectra will show the signatures of absorption by highly ionized gas.

?A correlation between the minimum absorption velocity andαox is likely to arise due to e?ects of orientation.For example,a line of sight close to the accretion disk will pass through more X-ray absorbing gas,and the minimum observed UV absorption velocity will be close to zero as the gas out?ow is primarily transverse to the line of sight.

The sample of BAL quasars observed by Chandra and XMM-Newton is steadily growing,and the current X-ray data are now of su?cient quality to investigate these predictions directly.In this paper,we sum-marize the insights o?ered from both exploratory and spectroscopic X-ray surveys of BAL quasars to the understanding of quasar winds.

RESULTS FROM EXPLORATORY SUR VEYS

Given the extreme faintness of BAL quasars in the soft X-ray regime(e.g.,Kopko et al.,1994;Green and Mathur,1996),obtaining detailed spectroscopic information requires a signi?cant commitment of observa-tory time.In this situation,only a small fraction of the population will have data of spectroscopic quality, and there is a legitimate danger that the conclusions drawn from such limited,potentially biased samples will not be representative of the population as a whole.Pursuing an alternate strategy of exploratory obser-vations enables much larger numbers of BAL quasars to be observed with the goal of measuring X-ray?ux, hardness ratio,andαox for each object.The low background and excellent spatial resolution of Chandra allow short,exploratory observations of5–7ks to reach sensitive?ux limits of<~10?14erg cm?2s?1in the

0.5–8.0keV band(Green et al.,2001;Gallagher et

al.,2002a).

We are in the process of compiling the largest exploratory survey to date,with a well-de?ned sample of BAL quasars drawn from the Large Bright Quasar Survey(LBQS;Foltz et al.,1989;Hewett et al.,1995). This sample of z>1.352BAL quasars is comprised of22objects with publicly available rest-frame UV spectroscopy from Weymann et al.(1991).To date,15of the20objects observed have been detected,and the remainder have tight upper limits on both X-ray?ux andαox(Gallagher et al.,2002a,in prep.). Though the BAL quasars in our sample are indeed weak in X-rays with a median value ofαox=?2.2 (see Figure2),they are generally detectable with the current generation of X-ray observatories.For those

Fig.2.Histogram of theαox values from the

LBQS BAL quasar exploratory survey.The

solid line is the entire sample,while the dot-

ted line is the sample of LoBAL quasars.The

arrows indicate upper limits.For compari-

son,the dashed gray line represents the low-

redshift Palomar-Green sample of unabsorbed

radio-quiet quasars from Brandt et al.(2000). objects that are detected,a measurement of the hardness ratio3indicates what fraction of the full-band(0.5–8.0keV)counts are coming from hard band(2–8keV)X-rays.Absorbed quasar spectra will have larger values of the hardness ratio as the cross-section for photoelectric absorption in the X-ray band decreases rapidly with increasing energy.The trend of increasing hardness ratio with decreasingαox illustrated in Figure3is therefore consistent with the understanding that X-ray faintness arises from intrinsic absorption. While this result is not new,it supports the idea thatαox can be used in a rough sense to indicate the extent of absorption.Following this premise,αox might be expected to correlate with UV absorption properties such as C iv absorption-line equivalent width,as demonstrated by Brandt et al.(2000)for their sample of Palomar-Green quasars.To complement Brandt et al.(2000)we are probing the extreme end of this parameter space with the BAL quasars.In this regime,there are no apparent correlations betweenαox and several absorption-line properties for the BAL quasars,most signi?cantly the minimum velocity of C iv absorption shown in Figure3.

Furthermore,the large column densities of absorbing gas implied by values ofαox<~?2.0coupled with the blue UV continua of the majority of our sample imply that the X-ray absorbing gas contains little dust, at least as we know it.This is consistent with the X-ray absorber being located within the dust sublimation radius.Alternatively,the bulk of the X-ray absorption may lie interior to the UV emission region,though this implies even smaller radii for the absorption.

We also con?rm the result of Green et al.(2001)that the BAL quasars with broad Mg ii absorption,the LoBAL quasars,are notably X-ray weaker than those with only high-ionization BALs.In fact,only one of the LoBAL quasars in our sample was actually detected(see Figure2).Given that LoBAL quasars generally have reddened optical and UV continua,it is unclear whether the extreme X-ray absorption that they su?er is occuring on the small scales inferred for the high-ionization BAL quasars.However,X-ray variability of the LoBAL quasar Mrk231suggests that a substantial X-ray absorber is located within~1015cm of the X-ray continuum source(Gallagher et al.,2002c).

RESULTS FROM X-RAY SPECTROSCOPY

Though exploratory surveys are essential for determining the general X-ray properties of the population as a whole,spectroscopic investigations are also required to obtain more detailed information on the X-ray continuum shape and absorption properties.Gallagher et al.(2002b)gathered together8spectroscopic-quality observations to investigate general trends from these data.In general,they found that the underlying X-ray continua are consistent with normal radio-quiet quasars,with signi?cant,complex absorption.Possible sources of the observed complexity(illustrated in Figure4)include partial covering of the X-ray continuum, ionized gas,and velocity structure similar to that seen in the UV regime.Furthermore,normalizing the power-law continua above5keV to correct for absorption resulted in typical radio-quiet quasar values of αox.All of these results are consistent with the Murray et al.(1995)picture.

However,some unexpected results have also arisen.Among the most interesting is the observation of

Fig. 3.Left panel:Hardness ratio versusαox.The trend of increasing hardness ratio with decreasingαox is consistent with absorption as the primary cause of X-ray weakness in BAL quasars.Right panel:Minimum velocity of C iv absorption versusαox.Contrary to expectations from the shielding gas models(§5,Murray et al., 1995),there is no apparent correlation between these parameters.

Fig. 4.Observed-frame ACIS-S3spectrum

of PG1115+080.This spectrum has been

?t with a power-law model above rest-frame

5keV which has then been extrapolated back

to lower energies.The signi?cant negative

residuals below~1.2keV indicate strong ab-

sorption,and the structure rules out simple

absorption by neutral gas.While the na-

ture of this complex absorber is not well-

constrained,it is consistent with a high ion-

ization parameter.

two absorption lines in the spectrum of APM08279+5255,apparently from highly ionized Fe out?owing with speeds of~0.2and~0.4c(Chartas et al.,2002).In radiatively driven?ows,such high terminal velocities occur in gas launched at much smaller radii than the UV BAL wind,which has terminal velocities of~12,000km s?1.In addition,when compared with an XMM-Newton observation of this object taken ~12rest-frame days later(Hasinger et al.,2002),there is a clear indication of variability in the absorber properties.APM08279+5255is not unique;similar absorption lines have also been seen in PG1115+080 (Chartas et al.,in prep.).These results are challenging for models in which the X-ray and UV absorption arise in the same medium and are predicted to have the same velocity structure(e.g.,Elvis,2000).Signi?cant iron emission has also been observed in some objects including H1413+117(Oshima et al.,2001;Gallagher et al.,2002b).

In another unanticipated result,the formerly X-ray brightest BAL quasar PG2112+059(Gallagher et al., 2001)exhibited signi?cant variability between the ASCA observation of Oct1999and the Chandra obser-vation of Sep2002.In the most recent Chandra data,the power-law continuum normalization dropped by a factor of a few while the measured column density of the absorber increased signi?cantly.Notably,con-current HST STIS data of the C iv absorption line showed no large changes from earlier UV observations. For PG2112+059at least,evidence for distinct X-ray and UV absorbers is mounting.

IMPLICATIONS FOR THEORETICAL MODELS OF QUASAR OUTFLOWS

In general,X-ray studies are providing support for the family of radiatively driven wind models.BAL quasars are generally X-ray weak with values ofαox indicating a signi?cant reduction in observed X-ray?ux relative to the UV continuum.When su?cient data are available,the intrinsic X-ray power,corrected for absorption, is typical of normal radio-quiet quasars.The complexity in the X-ray absorbers is consistent with highly ionized gas as predicted for the shielding gas;however the ionization state cannot be constrained given the current data quality.

Though the wind models have had some success,the lack of any apparent correlation between the UV absorption-line and X-ray properties is contrary to the model predictions.While certainly X-ray and UV absorption are found together in the same objects,the connection between the two is not straightforward. This is perhaps not unexpected given the complexity of the UV BALs and potential quasar-to-quasar variations in wind properties.The assertion of Elvis(2000)that the UV and X-ray absorption arise from the same medium is not easily reconciled with the lack of connection between X-ray and UV variability seen in PG2112+059and the mismatch in X-ray and UV absorption properties seen in APM08279+5255. In addition,the large-velocity out?ows seen in APM08279+5255,which imply a much smaller launching radius that for the BAL gas,do not?t within the Elvis(2000)picture with a narrow range of launching radii for the out?ow.

One of the problems for theoretical wind models with small launching radii is the lack of variability in velocity structure in the UV absorption lines on observed-frame timescales of years(e.g.,de Kool,1997). However,the most comprehensive absorption-line variability investigation to date,the3yr study of Barlow (1993),was not long enough for a conclusive test;the hydrodynamic wind models of Proga et al.(2000) generated instabilities on rest-frame timescales of~3yr.With longer time baselines,velocity increases have been seen for at least one object(Vilkoviskij and Irwin,2001),and an extended study is certainly warranted to investigate this issue further.If the bulk of the X-ray absorbing gas is on smaller scales than the BAL gas,it is also possible that X-ray absorption variability on shorter timescales may regularly occur,as seen for PG2112+059and APM08279+5255.To make accurate predictions for such events,the hydrodynamic models may need to be extended to smaller radii(<1015cm)with higher resolution,and also more explicitly include the e?ects of Compton pressure.For highly ionized gas,this will dominate the radiation force(D. Chelouche,https://www.wendangku.net/doc/6e10772010.html,m.).Additional items for an X-ray observer’s wish list from theoretical modeling are speci?c predictions for the strength and pro?les of Fe emission and absorption lines.

An outstanding issue for future investigations concerns the velocity of the X-ray absorbing gas in most BAL quasars;is it stalled,out?owing,or in?owing?Proga et al.(2000)claimed that highly ionized gas should be in?owing or stalled,but the evidence from APM08279+5255and PG1115+080suggests otherwise.An X-ray gratings observation of a BAL quasar o?ers the potential for a qualitative advance in this?eld.The improved resolution of gratings data,from?v~20,000km s?1for CCD data to<1300km s?1,would enable a more penetrating and subtle investigation into the connection between UV and X-ray absorption. In particular,signi?cant constraints could be placed on the velocity,column density,and ionization state of the X-ray absorbing gas.These values can then provide information on the location of the X-ray absorber and the mass-out?ow rate,parameters of fundamental physical importance.

ACKNOWLEDGEMENTS

We thank Sera Markov,Bev Wills,and Doron Chelouche for helpful discussions.This work was made possible by Chandra X-ray Center grants GO1-2105X and GO2-3129A as well as NASA grant NAS8-38252 for the ACIS instrument team.

References

Barlow,T.A.Time variability of broad absorption line QSOs.Ph.D.thesis,University of California,San Diego,1993.

Brandt,W.N.,M.Guainazzi,S.Kaspi,et al.An XMM-Newton detection of the z=5.80X-ray-weak quasar SDSSP J104433.04-012502.2.Astron.J.,121,591,2001.

Brandt,W.N.,https://www.wendangku.net/doc/6e10772010.html,or,and B.J.Wills.On the nature of soft X-ray weak quasi-stellar objects.Astrophys. J.,528,637,2000.

Chartas,G.,W.N.Brandt,S.C.Gallagher,et al.Chandra Detects Relativistic Broad Absorption Lines

from APM08279+5255.Astrophys.J.,579,169–175,2002.

de Kool,M.An overview of dynamical models for out?ows in BAL QSOs and Seyferts.In ASP Conf.Ser. 128:Mass Ejection from Active Galactic Nuclei,edited by I.Shlosman and R.J.Weymann,p.233,1997. Elvis,M.A structure for quasars.Astrophys.J.,545,63,2000.

Ferrarese,L.and D.Merritt.A fundamental relation between supermassive black holes and their host galaxies.Astrophys.J.Lett.,539,L9–L12,2000.

Foltz,C.B.,F.H.Cha?ee,P.C.Hewett,et al.The Large,Bright QSO Survey.II.-QSOs in three equatorial ?elds.Astron.J.,98,1959,1989.

Gallagher,S.C.,W.N.Brandt,G.Chartas,et al.An Exploratory Chandra survey of Large Bright Quasar Survey broad absorption line QSOs.In X-ray Spectroscopy of AGN with Chandra and XMM-Newton, edited by T.Boller,S.Komossa,S.Kahn,H.Kunieda,and L.Gallo,p.117,2002a.

Gallagher,S.C.,W.N.Brandt,G.Chartas,et al.X-ray spectroscopy of QSOs with broad ultraviolet absorption lines.Astrophys.J.,567,37,2002b.

Gallagher,S.C.,W.N.Brandt,G.Chartas,et al.X-raying the ultraluminous infrared galaxy and broad absorption line QSO,Markarian231with Chandra.Astrophys.J.,569,655,2002c.

Gallagher,S.C.,W.N.Brandt,https://www.wendangku.net/doc/6e10772010.html,or,et al.Heavy X-ray absorption in soft X-ray-weak active galactic nuclei.Astrophys.J.,546,795,2001.

Gebhardt,K.,R.Bender,G.Bower,et al.A relationship between nuclear black hole mass and galaxy velocity dispersion.Astrophys.J.Lett.,539,L13–L16,2000.

George,I.M.,T.J.Turner,T.Yaqoob,et al.X-ray observations of optically selected,radio-quiet quasars.

I.The ASCA results.Astrophys.J.,531,52,2000.

Green,P.J.,T.L.Aldcroft,S.Mathur,et al.A Chandra survey of broad absorption line quasars.Astrophys. J.,558,109,2001.

Green,P.J.and S.Mathur.Broad absorption line quasars observed by the ROSAT PSPC.Astrophys.J., 462,637,1996.

Hasinger,G.,N.Schartel,and S.Komossa.Discovery of an ionized Fe K edge in the z=3.91broad absorption line quasar APM08279+5255with XMM-Newton.Astrophys.J.Lett.,573,L77–L80,2002. Hewett,P.C.,C.B.Foltz,and F.H.Cha?ee.The Large,Bright Quasar Survey.6:Quasar catalog and survey parameters.Astron.J.,109,1498,1995.

Kopko,M.,D.A.Turnshek,and B.R.Espey.Optical,HST,and ROSAT observations of BAL QSOs.In IAU Symp.159:Multi-Wavelength Continuum Emission of AGN,edited by T.Courvoisier and A.Blecha, volume159,p.450.Kluwer,Dordrecht,1994.

Murray,N.,J.Chiang,S.A.Grossman,et al.Accretion disk winds from active galactic nuclei.Astrophys. J.,451,498,1995.

Oshima,T.,K.Mitsuda,R.Fujimoto,et al.Detection of an Iron Emission Feature from the Lensed Broad Absorption Line QSO H1413+117at z=2.56.Astrophys.J.Lett.,563,L103–L106,2001.

Proga,D.,J.M.Stone,and T.R.Kallman.Dynamics of line-driven disk winds in active galactic nuclei. Astrophys.J.,543,686,2000.

Vilkoviskij,E.Y.and M.J.Irwin.The spectrum of BAL QSO Q1303+308:intrinsic variability and line-locking stability.Mon.Not.R.Astron.Soc.,321,4–10,2001.

Weymann,R.J.,S.L.Morris,C.B.Foltz,et https://www.wendangku.net/doc/6e10772010.html,parisons of the emission-line and continuum properties of broad absorption line and normal quasi-stellar objects.Astrophys.J.,373,23,1991.

Yu,Q.and S.Tremaine.Observational constraints on growth of massive black holes.Mon.Not.R.Astron. Soc.,335,965–976,2002.

Email address of S.C.Gallagher:scg@https://www.wendangku.net/doc/6e10772010.html,.

相关文档
相关文档 最新文档