1. Martin Elvis, MIT, 4 March 2008 Reynolds C. with the collaboration of Guido Risaliti, Yair Krongold, Fabrizio Nicastro, Chris Reynolds and many others… X-ray absorption and AGN Structure Eclipsing in AGN: Martin Elvis Harvard-Smithsonian Center for Astrophysics

2. Martin Elvis, MIT, 4 March 2008 The Quasar Puzzle Z=6.4 SDSS spectrum, SED Jiang et al., 2007, ApJL, 666, L9, VanDen Berk et al., 2001,AJ,122, 549 z~6 quasar z~1-3 SDSS template CIV MgII Ly  AGN SEDs, spectra same –Over 6 decades of Luminosity –Over 13 Gyr of cosmic time Unchanging structure –Determined within BH sphere of influence: “The Nucleus” Complex phenomenology –Non-trivial structure Answers affect cosmic feedback

3. Martin Elvis, MIT, 4 March 2008 AGN ‘funnel wind’ Structure Elvis 2000 ApJ 545, 63; 2003 astro-ph/0311436; 2009 in preparation Conical geometry Pressure balance Multi-wavelength approach, uniting all atomic physics features in AGNs Thin Wind

4. Martin Elvis, MIT, 4 March 2008 Eclipses in Astronomy …but useful in astronomy: earth is a sphere binary stars black hole masses 1st test of GR extrasolar planets… “These late eclipses of the sun and moon portend no good to us.” King Lear I:2:98 …and for quasars

5. Martin Elvis, MIT, 4 March 2008 Scales of Quasar/AGN Components RsRsRsRs 10 1000 Warm Absorber 1 molecular torus Optical/UV continuum Black Hole High Ionization CIV HeII OVI Low Ionization MgII, FeII Balmer series NIR cont JHK 10 5 NELR H  [OIII] Coronal lines X-ray Cont. broad Fe-K Compton Wind BELR X-ray Spectral Variability inabsorption in absorption

6. Martin Elvis, MIT, 4 March 2008 100,000 Rg: 100,000 Rg: Cold Obscuring Torus RsRsRsRs 10 1000 Warm Absorber 1 molecular torus Optical/UV continuum Black Hole High Ionization CIV HeII OVI Low Ionization MgII, FeII Balmer series NIR cont JHK 10 5 NELR H  [OIII] Coronal lines X-ray Cont. broad Fe-K Compton Wind BELR

7. Martin Elvis, MIT, 4 March 2008 100,000 R S 100,000 R S The Dusty Molecular Torus Flattened optical & X-ray obscuring region Lawrence & Elvis 1982 type1  type2 AGN = ‘ Unified Scheme’ R sub = 1.3x10 5 L 0.1 0.5 M 8 0.5 T 1500 -2.8 R S N H ~10 24 cm -2 ; H/R~1 Puffed up by IR on dust? Krolik 2007, Murray 2007 Is this the whole story?

8. Martin Elvis, MIT, 4 March 2008 NGC1365: an AGN Structure Laboratory Face-on barred spiral D~25Mpc Bright X-ray source f x ~1 mCrab, logL x =42.2 Ward et al. 1978 MN 223, 788 L Bol ~10 -3 L edd (M/M 8 ) -1 logM BH ~ 7.3 - 7.8 M  –M-  bulge, M-K Obscured: N H >10 22.5 cm -2 Highly variable N H :Highly variable N H : –Key to study of AGN structure EPIC-MOS

9. Martin Elvis, MIT, 4 March 2008 NGC1365: Compton thick/Thin Transitions Constant E<2keV Thermal component Constant flux narrow Fe-K component log N H >24.2 log N H ~23.5 log N H ~23.7 Compton thick  thin transitions are rare: 4/30 Matt et al. 2003 1/11 Guainazzi et al. 2005 NGC1365 transitions reliably often: ~7/13 observations Compton-thin NGC1365 is a special object

10. Martin Elvis, MIT, 4 March 2008 NGC1365: Fairly Rapid Compton thin/thick transitions 2002Dec 24, t= 0 d 2003Jan 16, t=23 d 2003Feb 09, t=47 d 2003Aug13, t=232d 23 days 185 days Risaliti et al., 2005a, ApJL, 623, L93  N H >~10 24 cm -2 in 3 weeks: R<10 4 R s –requires R<10 4 R s <4.5x10 17 cm, –n e >10 9 cm -3 What is changing? –f x (intrinsic)? – c.f. disk cooling time > 30R 20 1/2 d – ionization? –  (ionization ) ~ 10 3 – N H  Hard to reconcile with standard torus – Hard to reconcile with standard torus

11. Martin Elvis, MIT, 4 March 2008 NGC1365: Really Rapid Compton thin/thick transitions Risaliti et al., 2007, ApJL, 659, L111 Thin  Thick  Thin in4 daysChandra: Compton Thin  Thick  Thin in 4 days –  N H >~10 24 cm -2 in 2 days –n e >10 9 cm -3 R(NH) < few 1000 R s Inconsistent with standard torus 2 days

12. Martin Elvis, MIT, 4 March 2008 3000R s 3000R s A More Compact Torus Need another ‘torus’ location near BELR Accretion disk windAccretion disk wind? –Also toroidal –Large H/R natural –No support problem Dust radiation pressure driven? –dust survives in disk to R>~2000 R s (T disk <1000 K) Magnetic support? Kartje, Königl & Elitzur, 1999 ApJ 513, 180  Should see fluorescent Fe-K from other parts of torus Elvis et al., 2004, ApJL, 615, L25 ; Risaliti et al., 2005a, ApJL, 623, L93

13. Martin Elvis, MIT, 4 March 2008 3000R s 3000R s Where is the narrow Fe-K? Suzaku Cycle 2: Elvis (PI) et al. Suzaku/NGC5548: Fe-K Suzaku 2-month monitoring of NGC5548 High S/N 240 ksec spectrum Fe-K narrow core: FWHM = 5500 km/s  = 50+/-15eV R~3000R g = Obscuring material? I.e. R s ~23 l-d M BH = 6.7x10 7 M  But factor ~4 in ~30days –Fe-K core = constant –Larger than 3000 R s ? –Non-Keplerian motions? –Light travel time smearing? –Not a worry yet. ct 3ct 2ct

14. Martin Elvis, MIT, 4 March 2008 Moving inwards: Warm Absorbers RsRsRsRs 10 1000 Warm Absorber 1 molecular torus Optical/UV continuum Black Hole High Ionization CIV HeII OVI Low Ionization MgII, FeII Balmer series NIR cont JHK 10 5 NELR H  [OIII] Coronal lines X-ray Cont. broad Fe-K Compton Wind BELR Dusty Wind

15. Martin Elvis, MIT, 4 March 2008 3000R s 3000R s Where are the Warm Absorbers? Highly ionized, blueshifted (~1000km/s): Wind Common: ~50% of all AGNs and quasars Location? Factor 10 6 in radius proposed –NELR to HiBELR Mass loss rate uncertain by factor 10 6 Is AGN feedback important to galaxy evolution or not? Kaspi et al. 2002, ApJ, 574, 643 Chandra HETG: 900ksec NGC3783

16. Martin Elvis, MIT, 4 March 2008 How to Measure the Warm Absorber Location Step function change Gradual WA response Nicastro et al., 1999, ApJ, 512, 184 WA response to continuum is not instantaneous: ‘Ionization time’ and ‘Recombination time’  measure n e U X ~ 4  L X /R 2 WA measures U x radius, R hence radius, R

17. Martin Elvis, MIT, 4 March 2008 Small R WA in NGC4051: XMM/RGS ~30 ksec High State HS Low State LS Low State High State 4X flux increase in ~30 ksec WA changes! Not continuous U(r) WA is dense & compact Not NELR Krongold et al., 2007, ApJ, 659, 1022

18. Martin Elvis, MIT, 4 March 2008 Small R WA in NGC4051: XMM/EPIC XMM EPIC Light Curve log U x (t), predicted from photoionization equilibrium log U x (t), measured Krongold et al., 2007, ApJ, 659, 1022 Two Warm Absorber components stay close to photo- ionization equilibrium

19. Martin Elvis, MIT, 4 March 2008 Small R WA in NGC4051: XMM/EPIC Krongold et al., 2007, ApJ, 659, 1022 Low Ionization (LIP) R<3.5 l-d R<3.5 l-d High Ionization (HIP) R=0.5-1.0 l-d

20. Martin Elvis, MIT, 4 March 2008 NGC4051: An Accretion Disk Scale Warm Absorber Rules out Narrow Emission Line Region Rules out Narrow Emission Line Region (kpc scale) Rules out Obscuring molecular torus Rules out Obscuring molecular torus (Krolik & Kriss, 2001) Minimum dust radius, r subl (NGC4051) ~ 12-170 light-days Rules out Low Ionization (H  broad emission line region Rules out Low Ionization (H  broad emission line region (BELR); R(H  ) = 5.9 light-days (Peterson et al. 2000) = disk? = wind?Co-located with High Ionization BELR?: HeII = wind? Driven by UV absorption lines, as in O-star winds?Driven by UV absorption lines, as in O-star winds? Murray et al.1995 light-days 5 10 15 HIP ~HeII BELR H  BELR Dusty molecular torus LIP Dust sublimation radius Krongold et al., 2007, ApJ, 659, 1022 Warm Absorber

21. Martin Elvis, MIT, 4 March 2008 NGC5548: R WA in a normal Type 1 AGN Choose a normal type 1 –Well characterized WA –Known M BH –Known R(BELR), R(dust) –Reliable variable –Bright NGC5548 one of very few 7 x 35 ksec in 1 2-month Suzaku window Behaved perfectly! Suzaku Cycle 2: Elvis (PI) et al. 2 month Suzaku light curve of NGC5548 x4 #1 #3

22. Martin Elvis, MIT, 4 March 2008 NGC5548: R WA in a normal Type 1 AGN Suzaku Cycle 2: Elvis (PI) et al. Suzaku/NGC5548: obs1,3 ratioed to 2-5 keV Power law log U x (t), measured log U x (t), predicted from photoionization equilibrium EARLY RESULTS! Subject to change without notice 2 velocity systems Chandra –LoVel:  changes, but not in equilibrium Dense, compact –HiVel: no change in  Low density, large –WA acceleration? Need a sample: proposed NGC3227 in Cycle 3

23. Martin Elvis, MIT, 4 March 2008 50R s 50R s Narrow Fe-K Absorption RsRsRsRs 10 1000 Warm Absorber 1 molecular torus Optical/UV continuum Black Hole High Ionization CIV HeII OVI Low Ionization MgII, FeII Balmer series NIR cont JHK 10 5 NELR H  [OIII] Coronal lines X-ray Cont. broad Fe-K Compton Wind BELR Dusty Wind

24. Martin Elvis, MIT, 4 March 2008 50R s 50R s NGC1365: Compton Wind Risaliti et al., 2005, ApJL, 630, L129 EPIC-MOS K  /K  observed observed Curves of growth Log N H 2324 High V t EW Log N H pn mos 4 strong absorption lines: FeXXV, FeXXVI K- , K-  EW~100eV –Broad: v turb ~500-1000 km/s K  /K  small - saturated: – High N H >10 23.5 Compton thick Super-High Ionization WA: –logU X ~0, R ~ 50 (M/M 8 ) -1 R s Blueshifted 2500  5000 km/s in 18mo. –acceleration?

25. Martin Elvis, MIT, 4 March 2008 50R s 50R s NGC1365: Failed Compton Wind V(Fe-K) << V esc ~50,000 (M/M 8 ) -1 km/s –high U –L/L Edd ~ 10 -3 –‘failed wind’ –‘failed wind’? or transverse = ‘hitchhiking gas’ ? = ‘hitchhiking gas’ ? Murray & Chiang –Filters X-ray continuum from outer gas… –Lowers ionization state… –Enabling UV line driven acceleration (as in O-star winds) –Disk geometry essential Risaliti et al., 2005, ApJL, 630, L129 X-ray continuum source

26. Martin Elvis, MIT, 4 March 2008 4. <30 R S 4. <30 R S X-ray Continuum Eclipses RsRsRsRs 10 1000 Warm Absorber 1 Optical/UV continuum Black Hole High Ionization CIV HeII OVI Low Ionization MgII, FeII Balmer series NIR cont JHK 10 5 NELR H  [OIII] Coronal lines X-ray Cont. broad Fe-K Compton Wind BELR Dusty Wind

27. Martin Elvis, MIT, 4 March 2008 NGC1365: Rapid X-ray Continuum Eclipses Risaliti et al., 2007, ApJL, 659, L111 Compton Thin  Thick  Thin in 4 days Now consider constraints on X-ray Continuum source 2 days

28. Martin Elvis, MIT, 4 March 2008 <30 R S <30 R S X-ray Continuum Eclipses Dimensions of X-ray source:  D s = V c x T ecl D s ~ 10 14 cmIf V c ~ 10 4 km/s  D s ~ 10 14 cm Can’t be much higher (to avoid overionization, from iron line width) ¯If lower, even smaller X-ray source Black hole mass in NGC 1365: –Log M BH /M  = 7.3 (0.3,0.3) M-  (Ferrarese et al. 2005) –Log M BH /M  = 7.86 (0.15,0.3) M-L K (Marconi & Hunt 2003) X-ray source size < 33 R GX-ray source size < 33 R G Risaliti et al., 2007, ApJL, 659, L111 VKVK X-ray source Eclipsing matter DsDs DCDC

29. Martin Elvis, MIT, 4 March 2008 <30 R S <30 R S Future Continuum Eclipses NGC1365 Resolved eclipse will measure size. ~10hours ~7R S‘anti-eclipse’ in XMM long look suggests ingress ~10hours ~7R S Goal of Suzaku Cycle 3 proposal Risaliti et al., 2008, in preparation 

30. Martin Elvis, MIT, 4 March 2008 Future Continuum Eclipses Schwartzschild eclipse profile Kerr eclipse profile Resolving Eclipse profile: Size/shape/1-D emissivity map Kerr/Schwartzschild? C.Reynolds Close to resolving? Suzaku cycle 3 propoal

31. Martin Elvis, MIT, 4 March 2008 <30 R S <30 R S X-ray Relativistic Fe-K Eclipses RsRsRsRs 10 1000 Warm Absorber 1 Optical/UV continuum Black Hole High Ionization CIV HeII OVI Low Ionization MgII, FeII Balmer series NIR cont JHK 10 5 NELR H  [OIII] Coronal lines X-ray Cont. broad Fe-K Compton Wind BELR Dusty Wind

32. Martin Elvis, MIT, 4 March 2008 <30 R S <30 R S Broad Fe-K Eclipses Red wing on Fe-K is 1% over continuum. Ambiguous. NGC1365 has strong red wing & strong Fe-K absorption lines  High Fe abundance? Gone in Compton-thick spectrum.Gone in Compton-thick spectrum. D Fe-K < 10 14 cm, <33 R SD Fe-K < 10 14 cm, <33 R S Risaliti et al., 2007, ApJL, 659, L111

33. Martin Elvis, MIT, 4 March 2008 <30 R S Future: <30 R S Future: Tomography of Fe-K Apply Binary physics Ingress, egress successively cover/uncover red-/blue- shifted Fe-K Clean, definitive experiment EstablishEstablish –Fe-K –Rotation – z(R) Goal of Suzaku Cycle 3 proposal More sensitive mission could probe metric in detail Reynolds, Risaliti, Elvis, … Blue wing occulted Red wing occulted

34. Martin Elvis, MIT, 4 March 2008 Mission for Future Eclipse Mapping of AGNs Weeks-long observations of each source - not an observatory-class mission 10 x S/N of Chandra in ~1hour vs. 4 hours -large area, high count rate Better spectral resolution Only ~1 arcmin spatial resolution (HPD) Extreme Physics Explorer 5m 40 Elvis, Astro-ph/0403554; 2004, Nucl. Phys. B 134, p.78-80

35. Martin Elvis, MIT, 4 March 2008 X-ray Eclipses and AGNs Several techniques probe scales from 10 5 - 10 R s A simpler, physically driven picture of AGN Structure Makes quasar puzzle of uniformity semi-understandable rgrgrgrg 10 1000 100 10000 WA NAL HiBEL Dusty Obscurer Broad Fe-K X-ray Cont. Compton (failed) Wind Narrow Fe-K

36. Martin Elvis, MIT, 4 March 2008 Eclipse results confirm & extend ‘funnel wind’ model Elvis 2000 ApJ 545, 63; 2003 astro-ph/0311436 Becoming a secure basis for physical wind models: will allow extrapolation

37. Martin Elvis, MIT, 4 March 2008 The Restless AGN Accretion Disk Replace observational with physical names Active Disk ‘atmosphere’ at all radii –mechanism must work at all radii Not standard ‘hot corona’ More like Solar Corona rgrgrgrg 10 1000 100 10000 WARM Dust driven Wind Broad Fe-K X-ray Cont. HOT Narrow Fe-K Line driven Wind Compton driven Wind COLD failed Wind L<L Edd failed Wind UV Cont.

38. Martin Elvis, MIT, 4 March 2008 Reynolds C. AGN X-ray Eclipses …give new views of AGN structure