Probing high velocity inflows and outflows in AGNs

Probing high velocity inflows and outflows in AGNs via time-resolved X-ray spectroscopy Massimo Cappi INAF-IASF, Bologna Outline I) accretion (in)flows in AGNs Accretion modes/conditions + relativistic/fundamental physics …. from source-by-source analysis to large samples II) ejection (out)flows from AGNs Acceleration physics + interaction with host galaxies/cosmology …. from source-by-source analysis to large samples Credit: A. Mueller Main collaborators: V. Braito, A. Celotti, M. Dadina, B. De. Marco, M. Giustini, K. Iwasawa, G. Miniutti, G. Palumbo, G. Ponti, J. Reeves, F. Tombesi, T. Yaqoob

I - Accretion/inflows: Redshifted Fe lines/components Evidence of complex, narrow, redshifted Fe. K lines… Chandra – NGC 3516 XMM – NGC 3516 Bianchi et al. , 2004 XMM – ESO 198 -G 024 Turner et al. '02 Origin in innermost regions of accretion disk+ blob-like structure (or inflowing blobs? ) Dovciak et al. , 2004 Guainazzi et al. , 2003

I - Accretion/inflows: …and variable Fe lines/components Redshifted (component of) Fe lines DO show fast time variations, although not always with a clear interpretation! XMM - NGC 3516 Can fit line maxima by three Keplerian orbits with same inclination & central mass !! XMM – Mrk 766 Iwasawa, Miniutti & Fabian, 2004 Turner et al. 2005

I - Accretion/inflows: Source-by-source analysis NGC 3783 Tombesi et al. 2007 IC 4329 a De. Marco et al. 2010 b Consistent with origin from hot spots, or spiral waves, in inner regions of accretion disk?

I - Accretion/inflows: From source-by-source to representative samples $ Variability in redshifted Fe line on time-scales ~10 ks demonstrate its origin within few tens Rg (hot spots, spiral waves, or else? ? Not really clear) But see Barbara De Marco’s talk, this afternoon

II - Ejection/outflows: Warm absorbers…probe highest-v gas 50% of all Sey 1 s exhibit WAs Many details from Chandra/XMM gratings NGC 3783 Exp=900 ks ASCA Fabian, et al. ’ 94 Otani, ’ 95, Ph. D Reynolds et al. '97 Georges et al. '97 Kaspi et al. '01; Netzer et al. '02; Georges et al. '03; Krongold et al. ‘ 03 Clear now that often multiple ionization & kinetic components: outflows with v~100 -1000 km/s Blustin et al. 2004

II - Ejection/outflows: Blue-shifted absorption lines/edges – High-v New and unexpected results from Chandra and XMM-Newton observations PG 1211+143 (z=0. 08) v~0. 1 c 2 Energy (ke. V) 5 7 10 Pounds et al. 2003 a, b PDS 456 (z=0. 18) v~0. 1 c (If) interpreted as Kα resonant absorption by Fe XXV (6. 70 ke. V) or Fe. XXVI (6. 96 ke. V) 2 Energy (ke. V) 5 7 10 Reeves et al. 2003 massive, high velocity and highly ionized outflows in several RQ AGNs/QSOs Mass outflow rate: comparable to Edd. Acc. rate (~M◉/yr); velocity ~0. 1 -0. 2 c

II - Ejection/outflows: Blue-shifted absorption lines/edges - Variability Absorbers variability on timescales 1000 -10000 s NGC 1365 Mrk 509 (long-look, 200 ks) Obs 1 Obs 2 Obs 3 Risaliti et al. 2005 (See also Krongold et al. 2007 on NGC 4051) MC et al. , 2009 Dadina et al. ‘ 05 Variability allows to place limits on location, mass, etc.

Mrk 509: Among brightest F~2 -5 x 10 -11 cgs and most luminous L~1 -3 x 1044 ergs/s Type 1 Sey known EW(Fe. XXVI)~-20 -60 e. V V~0. 14 -0. 2 c ∆t~100 ks Log ~5 erg cm/s, Nh~2 -4 x 1023 cm-2, v~0. 14 -0. 2 c From absorber parameters and variability, we estimated: R<500 Rs MC et al. 2009 MRK 509 contour plots (MC et al. 2009)

X-ray Observations: Variability Again, absorber variability on timescales ~20000 s MCG-5 -23 -16 (XMM+Chandra) Braito et al. , 2007 Data require large Nw, high , and vout=0. 1 c

Interpretation: (Three main) Wind dynamical models i) Thermally driven winds from BLR or torus ii) Radiative-driven wind from accretion disk Murray et al. ‘ 95, Proga et al. ‘ 00 …and/or… iii) Magnetically driven winds from accretion disk Balsara & Krolik, 93; Woods et al. ‘ 96 i) Large R, low v ii) and iii) Low R and large v Emmering, Blandford & Shlosman, ’ 92; Kato et al. ‘ 03

Theoretical Interpretation: (Three main) Wind dynamical models MHD+LD model by Proga et al. ‘ 00, ‘ 03 Overall numbers (Nh, , vout, etc. ) are consistent with observations…

Data Interpretation: Yes indeed…one expects (mostly/only) strong Fe line absorptions when accounting for proper wind geometries and physics Sim et al. , 2008

Critical Issues (i/ii): Observations ü Lines statistical significance? (transient features, number of trials in time and energy, etc…) ü Identifications of edges/lines energies? (Kallman et al. 2005, Kaspi et al. for PG 1211) ü Publication bias? Only positive detection, low signif. , Vaughan & Uttley ‘ 08) Outflow v (km/s) ü Local “contamination”? (PDS 456 at risk? Mc. Kernan et al. ’ 04, ‘ 05) cz (km/s)

Critical Issues (ii/ii): Interpretation/Theory Fundamental to: ü Nw (cm-2) i) PHYSICS of accelerated and accreted flows (winds? , blobs? , etc. ), i. e. understand how BHs accelerate earth-like quantities of gas to relativistic velocities ii) COSMOLOGY: i. e. estimate the mass outflow rate, thus the impact of AGN outflows on ISM and IGM enrichment and heating! ü Location (R, Delta. R) ü Ionization state ( ) ü Velocity ü Covering factor ü Frequency in AGNs Elvis et al. ‘ 00, Creenshaw et al. ’ 03, King et al. ‘ 03, Chartas et al. ‘ 03, Yaqoob et al. ‘ 05, Blustin et al. ‘ 05, Risaliti et al. ’ 05, Krongold et al. ‘ 07 Current estimates have order of magnitude uncertainties, they go from: d. M/dt ( Lkin) few % to several times d. Macc/dt ( Ledd) This is a fundamental (open) issue

I - Ejection/outflows: From source-by-sources to representative samples… Again, we analysed in a systematic and uniform way, a (almost) complete sample of nearby, X-ray bright, radio-quiet AGNs (Tombesi et al. 2010 a) z distribution of sources 4 -10 ke. V fluxes • Selection of all NLSy 1, Sy 1 and Sy 2 in RXTE All-Sky Slew Survey Catalog (XSS; Revnivtsev et al. 2004) • Cross-correlation with XMM-Newton Accepted Targets Catalog • 44 objects for 104 pointed XMM-Newton observations • Local (z<0. 1) • X-ray bright (F 4 -10 ke. V=10 -12 -10 -10 erg s-1 cm-2)

Absorption lines search Uniform spectral analysis: • Reduction and analysis of all EPIC pn spectra in the 4 -10 ke. V • Baseline model: absorbed power-law + Gaussian Fe K emission lines Absorption lines search: • Addition of narrow line to baseline model stepping energy in 410 ke. V and recording Δχ2 deviations • Visualization on energy-intensity contour plot (significance 68% red, 90% green, 99% blue) (e. g. Miniutti et al. 2007, MC et al. 2009) Absorbed power-law • Selection of narrow lines with F-test confidence levels ≥ 99% • Line parameters determined by direct fitting to the data Example of PG 1211+143 (Tombesi et al. 2010 a) Absorbed power-law + emission line

Absorption lines significance F-test can overestimate the detection significance for a blind search of emission/absorption lines over a range of energies (e. g. Protassov et al. 2002). Extensive Monte Carlo simulations (e. g. MC et al. 2009) • Additional significance test for lines at energies ≥ 7. 1 ke. V • Null hypothesis that spectra are fitted by model without absorption lines • 103 simulated spectra for each case • Simulated Δχ2 distribution for random generated lines • Selection of lines with MC confidence levels ≥ 95% Global probability for the lines to be generated by random fluctuations is very low (≤ 10 -8 from Binomial distribution). • Checked no contamination from pn background and calibration • Independent confirmation of blue-shifted lines detection from MOS data (without relying on any statistical method)

Results: Yes, we confirm there are indeed UFOs! (Ultra-Fast Outflows…) Blue-shift velocity distribution Cumulative velocity distribution • 36 absorption lines detected in all 104 XMM observations • Identified with Fe. XXV and Fe. XXVI K-shell resonant absorption • 19/44 objects with absorption lines (≈43%) • 17/44 objects with blue-shifted absorption lines (lower limit ≈39%, can reach a maximum of ≈60%) • 11/44 objects with outflow velocity >0. 1 c (≈25%) • Blue-shift velocity distribution ~0 -0. 3 c, peak ~0. 1 c • Average outflow velocity 0. 110± 0. 004 c Tombesi et al. 2010 a (The UFO’s hunters commander in chief)

Results Lines EW distribution • Most frequent detected line is Fe. XXVI Lyα • EW is in the range ≈10 -100 e. V, with mean ≈40 -50 e. V • Estimated global covering factor from fraction of sources with lines (C=Ω⁄4Π)≈0. 4 -0. 6 • Geometry not very collimated, large opening angles favored (similar to WA)

Last but not least…for those still skepticals on UFOs Publication bias solved : • Uniform analysis on complete sample of sources Not due to local contamination: • Lines detection assessed by MC simulations No correlation between cosmological redshift and lines blue-shift, no local (z≈0) absorption. • Global chance probability very low (≤ 10 -8) • Detection independently confirmed by MOS data

Physical photo-ionization modelling using XSTAR Mean phenomenological SED from the radioquiet sample: • SEDs 34 Type 1 s from NED database • scattering due to source variability and different instruments • normalized SEDs to near-IR inflection point ~1. 25μm (like Elvis et al. 1994) • average flux for each energy point Simple mean SED with three intervals: Radio to mm (Γ~2), mm to IR (Γ~0. 1), IR to X-rays (Γ~2) Xstar grid for direct spectral fitting: • input enenergy band 0. 1 e. V, 106 e. V • mean SED for radio-quiet sources • power-law Γ=2 for radio-loud sources • turbulent velocity v=500 km/s NH~1022 -1024 cm-2, logξ~3 -6 erg s-1 cm-2 (vout, NH lower limits, unknown inclination angle)

Ejection/outflows: • estimated distances r<0. 01 -0. 1 pc (<102 -105 rs) (accretion disk winds? e. g. Elvis 2000; King & Pounds 2003) • Often vout > vesc, but not always, material shall fall back sometimes? (“aborted jet”? e. g. Ghisellini et al. 2004, Dadina et al. 2005) • variability time scales t~1 day – 1 year • Lbol/LEdd~0. 1 -1 • Mout/Macc~0. 1 -1 • Ek~1044 -1045 erg s-1 ~0. 1 Lbol (last two estimates depend on covering fraction C) • Acceleration mechanism? Line, magnetically or momentum driven?

WAs in RQ from Mc. Kernan et al. (2007) (filled black circles), WA in RL from Torresi et al. (2009) and Reeves et al. (2009) (filled blue triangle), UFOs in RQ (red crosses)

Ejection/outflows: Momentum-driven accretion disk winds/outflows? As in King 2009? “Eddington winds from AGN are likely to have velocities ~0. 1 c and show the presence of helium- or hydrogen-like iron” (King 2009) • Eddington accretion episodes Lbol/LEdd~0. 1 -1 • electron scattering wind τ~1 at infinity • wind momentuum ~ photon momentuum (Moutv~LEdd/c) • typical velocity ~0. 1 c • typical ionization parameter logξ~4 and linear relation vout and ξ (Fe XXV, Fe XXVI + Nickel ions? ) • wind interaction with host galaxy, ram pressure • important for feedback SMBH and host galaxy • does explain the M-σ relation

Future prospects: Astro-H calorimeter Calorimeter on board Astro-H, next Japanese X-ray satellite to be launched in 2013: good effective area (≈250 cm 2 @ 6 ke. V) and high energy resolution (FWHM≈7 e. V) from 0. 1 ke. V up to 12 -13 ke. V. • realistic spectra simulations of UFOs • absorption lines resolved, measured velocity broadening (down to ~100 -200 km/s @ 6 ke. V) • estimates EW, blue-shift, centroid energies ~10 times better XIS-FI (100 ks simulation) • better constrains on NH, logξ, vturbu

Future pospects: IXO calorimeter X-ray Microcalorimeter Spectrometer (XMS): high effective area ~0. 65 m 2 (~6500 cm 2 !!) @ 6 ke. V and high energy resolution (FWHM≈2. 5 e. V) from 0. 1 ke. V up to 12 -13 ke. V. Flux limits (EW=10 e. V) (Tombesi et al. 2009) logξ=3 erg s-1 cm, NH=1023 cm-2, b=1000 km/s (Tombesi et al. 2009) Flux limits Spectra simulations • 2 -10 ke. V flux limits for 5σ detection of narrow absorption lines in the 3 -11 ke. V • Simulations of highly ionized and massive absorbers • Different EWs, exposure times and responses • Line details (profile, energy, broadening) measured with high accuracy (>30 times Astro-H) • Lines of EW=10 e. V (50 e. V) in ≈6 -9 ke. V for ≈10 -12 (10 -13) erg s-1 cm-2 (expo 100 ks) • Fe. XXV/XXVI K lines detectable with high significance • Extend study to less bright sources • Spectral variability on time-scales of 5 (10) ks for • Time variability, dynamics of absorbers ≈10 -11 (10 -12) erg s-1 cm-2

Summary After 10 years of Chandra and XMM-Newton observations, we are able to move from source-by-source analysis to representative samples: I) accretion (in)flows in AGNs Redshifted Fe. K lines/components do vary on short time-scales Probe of accretion modes/conditions + relativistic/fundamental physics II) ejection (out)flows from AGNs Blueshifted Fe absorption lines confirmed, and variable Probe launching regions of winds/jets + Impact for feedback

IXO Science Case for YB in CV- Call for ideas/contributions/suggestions Section on Strong Gravity: (<7 pages) The idea is to start from white papers + RFI 1&2 for Decadal, and UPDATE with new results I- Probing Strong Gravity with X-rays: • Time-resolved spectroscopy to probe the kinematics and dynamics of the inner BLR and wind/jet formation regions • Time-resolved spectroscopy (e. g. waterfalling diagram) to probe the kinematics and dynamics of the inner regions of accretion disk. . the idea could be to present these studies as unique way to probe/go/travel from the outer parts down into the innermost regions of the accretion disk (where winds/blobs/ejecta/jets also form). II- Measuring BH Spins with X-rays, using several independent techniques: • Fe. K+Fe. L fitting, and time-lags • Continuum fitting (in GBHCs and AGNs) • QPOs • time-resolved spcetroscopy (i. e. reverberation) • Polarization measurements. . . the idea would/could be to present the 5 different (independent!) techniques as a way to then "calibrate" the Fe. K measurements, that is the technique we want to be able to apply to several tens of AGNs (maybe up to z about 1) in order to attempt the "spin distributions" Contact me directly (cappi@iasfbo. inaf. it) for any contributions Any other section (i. e. Neutron Stars) or Observatory Science: Call for ideas/proposals from Xavier Barcons Due date is end of April

Thanks for your attention, and by the way, watch-out for UFOs!

After 10 years of Chandra and XMM-Newton observations, we are able to move from source-by-source analysis to representative samples Jet II- X-ray absorption spectroscopy: Probe of launching regions of winds/jets/UFOs I - X-ray emission spectroscopy: Probe of innermost regions of accretion disk Credit: A. Mueller

Accretion/inflows: A parenthesis…. Mrk 509 and 3516, III– Complex Absorption Fe. K lines: Narrow/broad(? ) redshifted absorption lines: VERY NEW! and much debated Mrk 335 MCG-6 -30 -15 in low state (Ponti, private com. ) nandra and dadina (Longinotti et al. , submitted to A&A. ) PG 1211+143 (z=0. 08) Chandra data (Reeves et al. , astro-ph/0509280) E 1821+643 (z=0. 3) HETG data (Yaqoob & Serlemitsos, 2005) i) Fake lines? ii) (very) wrong continuum (WA)? iii) 2 different lines? 1 narrow + 1 Kerr (red, reverberation, tail of a Kerr line ? ) iv) 1 relativistic line with resonant absorption? Direct probe of relativistic bulk inflows! (v~0. 1 -0. 3 c)