1. Mike Crenshaw (Georgia State University) Steve Kraemer (Catholic University of America) Mass Outflows from AGN in Emission and Absorption NGC 4151

2. IUE: black pluses HUT: red diamonds FOS: green triangles STIS: blue x’s Six HST/STIS echelle observations (0.2'' x 0.2''): 1999 July - 2002 May Simultaneous HST, FUSE, and CXO observations in 2002 May NGC 4151: UV Light Curve

3. Absorption Components in STIS and FUSE Spectra A, C, D+E, E are intrinsic; B is Galactic; F, F are host galaxy. D+E (v r = -500 km s -1 ) responsible for bulk of UV and X-ray absorption.

4. So what are the intrinsic absorbers? What is their origin? –Accretion-disk winds, evaporation from torus? What are their dynamics? –Radiatively-driven, thermal wind, hydromagnetic flows? (see Crenshaw, Kraemer, & George, 2003, ARA&A, 41, 117 ) What observational constraints are needed? –Physical conditions: U (ionization parameter), N H (column density), n H (number density), abundances, etc. –Kinematics: radial velocity (v r ), FWHM, transverse velocity (v T ) –Geometry: Global covering factor (C g ), LOS covering factor (C los ), distribution with respect to accretion disk axis (polar angle  )? –Radial location (r), mass outflow rate Are the absorbers seen in emission? Yes: Emission lines from the high-column absorber in NGC 4151 provide tight constraints on dynamical models of the mass outflow.

5. (Kraemer et al. 2006, ApJ, in press, astro-ph/0608383) D+E varies strongly in response to ionizing continuum changes. D+E in 2002: a large amount of gas moved out of the LOS. Absorption Variability in C IV Region

6. Absorption Variability in X-rays X-ray absorption primarily due to D+E D+E decreased in N H between 2000 and 2002 Evidence for a more highly ionized component: X-high (Kraemer et al. 2005, ApJ, 633, 693)

7. Density (n H ) from metastable C III  radial distance of D+E is ~0.1 pc D+Ed change in los covering factor  v T ≈ 2100 km s -1 Other constraints? Photoionization Models of High-Column Absorbers Yes! D+Ea is seen in emission.

8. Emission-Line Profiles at Low Flux Levels He II profile has two components (broad component not detected): narrow: 250 km s -1 FWHM, intermediate: 1170 km s -1 FWHM Evidence for an intermediate line region (ILR)

9. Emission-Line Profiles at Low Flux Levels D+E absorbs ILR and has same velocity extent  self absorption? Are we seeing the absorption in emission?  D+Ea should dominate D+Ea absorber models should match the observed ILR line ratios C IV blue - narrow red - intermediate green - broad

10. blue - narrow red - intermediate green - broad Intermediate Components in Other Lines (Crenshaw & Kraemer, 2006, ApJ, submitted)

11. ILR Line Ratios and D+Ea Photoionization Models Reasonably good match, considering no fine-tuning of absorber models - N V underpredicted (similar to most of our NLR models) Which value of N H is more appropriate globally? - look at the variability of C IV

12. Variability of C IV Emission Components Both BLR and ILR respond positively to continuum changes Size of ILR ≤ 140 light days (0.12 pc)

13. ILR C IV vs. Continuum Flux High-N model is a better match globally Scale factor for High-N model gives C g = 0.4 (global covering factor) + Observed --- High-N Model … Low-N Model

14. Can we constrain the geometry of the ILR? Kinematic studies show the NLR of NGC 4151 is roughly biconical with a half-opening angle of ~33  and an inclination of ~45  (Das et al. 2005). Previous photoionization studies showed the NLR is shielded by an absorber with U, N H similar to D+Ea/ILR (Alexander et al. 1999, Kraemer et al. 2000). Thus, the ILR is concentrated in the polar direction and extends to  ≥ 45  (  = 53  gives C g = 0.4) NLR and host galaxy

15. Simple Geometric Model r = 0.1 pc,  = 45 , v r = v los = - 490 km s -1 Assume v  = 0, then v  = v T = 2100 km s -1 (v T = 10,000 km s -1 also shown) Emission-line v r ≤ 1550 km s -1, close to observed HWZI (1400 km s -1 )

16. Consider the high-column absorbers D+E and X-high: Radiation pressure: To be efficient FM > (L bol /L edd ) -1 = 70 for NGC 4151 From Cloudy models: FM (X-high) < 2, FM (D+Ea) < 40 X-high is not radiatively driven and D+E is marginally susceptible Thermal wind: Radial distance at which gas can escape: r esc ≥ 7 pc (X-high), r esc ≥ 400 pc (D+Ea) Neither are thermally driven. Magnetocentrifugal acceleration: Likely important, at least by comparison to other alternatives. Gives large transverse velocities and large line widths (Bottorff et al. 2000) Dynamical Considerations

17. Conclusions There is an intermediate-line region (ILR) in NGC 4151, characterized by FWHM = 1170 km s -1. The ILR is the same component of outflowing gas responsible for the high- column UV and X-ray absorption (D+Ea) at ~0.1 pc from the nucleus. The ILR has C g  0.4 and it shields the NLR, indicating outflow over a large solid angle centered on the accretion-disk axis. The kinematics at this distance are likely dominated by rotation, but there is a significant outflow component (v T  2100 km s -1 and v r = - 490 km s -1 ). A simple geometric model yields maximum emission-line velocities close to the observed HWZI of the ILR (1400 km s -1 ) and significantly less than v T. The mass outflow rate is ~ 0.16 M  yr -1, about 10x the accretion rate. Dynamical considerations indicate that magnetocentrifugal acceleration is favored over pure radiation driving or thermal expansion. Future work: compare these constraints with predictions from dynamical models (e.g., Proga 2003; Chelouche & Netzer 2005; Everett 2005).

18. THE END