INTRODUCTION to AGN “Sometimes you can’t stick your head in the engine, so you have to examine the exhaust” -- D. E. Osterbrock Astro 596G Spring 2007 In 2006, there were 2151 papers in ADS with “quasar” or “AGN” in their abstracts

Key Questions Growth of SMBHs 1.Due to mergers, accretion, stellar captures? SMBHs and Galaxy/LSS formation 1.Were BHs seeds or byproducts of galaxy formation? 2.How important is is the feedback from SMBHs in structure formation? Jets in AGN 1.Are jets matter or Poynting flux dominated? 2.Why are some AGN radio-loud, some radio-quiet? Accretion Physics 1.Is our basic cartoon of the central engine correct? 2.Is the unified model viable? Useful? 3.Have we included the essential physics?

AGN & Related Objects Energy output not related to “ordinary” stellar processes Continuum bright in virtually every wave band Mass accretion onto supermassive black holes Radio Galaxies Radio-Loud Quasars BL Lac’s, Blazars, OVV’s (Optically Violent Variables) Radio-Quiet QSOs (Quasi-Stellar Objects) Seyfert 1 & 2 Galaxies LINERS (Low ionization nuclear emission line regions) Starburst galaxies; ULIRGS (Ultraluminous IR Galaxies) Refn: An Introduction to Active Galactic Nuclei, by Brad Peterson

Why study quasar emission lines ? Probe inner few pc’s of the AGN Learn about the unobservable UV continuum c.f. Zanstra temperatures UV originates in the inner region of the accretion disk UV photons from quasars an important contributor to the metagalactic UV background Interesting line-formation phenomena because conditions are extreme Want to understand the source of the gas which fuels the central engine Unique probe of abundances at z=5-6 Possible uses as cosmological probes (e.g. Baldwin effect) Use to measure central black hole masses in AGN

Broad red wing of Fe Kalpha emission  reflection of X-ray continuum off inner accretion disk Nandra et al. ASCA composite for Sy1’s

Radio: Synchrotron (relativistic electrons in B-field) Mm-1 micron: Dust emission 1 micron -.2 keV: Thermal emission from optically thick accretion disk X-rays: Synchrotron, Inverse Compton, Hot corona + reflection Continuum Spectral energy distributions:

Quasar and Seyfert 1 Spectra: Broad permitted lines (widths > 10,000 km/sec) Narrow forbidden lines (widths only few hundred km/sec), e.g. [OIII] 4959, 5007 CIII] 1909 is broad Forbidden Lines: [O III] Semi-Forbidden: C III] Recombination: Ly alpha, Halpha

Aside of forbidden v. permitted lines, recombination lines : In photoionized gas, the lines of hydrogen and helium are “recombination” lines, and to first order do not depend on physical conditions like density or temperature. e-e- Photon of Ly alpha, Halpha, etc 1 2

Most metals are collisionally excited (the first excited level is low enough that for nebular temperatures, collisions can populate the upper levels) Whether or not the electron de-excites via emitting a photon or by a collision with a free electron depends on (1) density of particles, and (2) the Einstein A for the radiative de-excitation transition. Permitted lines: Radiative de-excitation is fast (Einstein A is high), so atom radiatively de-excites before a collision Forbidden lines: Radiative de-excitation is not likely. If the density is low enough, a collision will not happen, and the atom will de-excite radiatively  we see the line. If the density is greater than the “Critical density” then the atom collisionally de-excites before it emits a photon  no line

Ionization Parameter

Emission Lines arise in 2 separate “regions” 1.Narrow Line Region (NLR) Extended spatially (~kpc) in nearby Seyferts Low density n~ cm -3 since you see forbidden lines like [OIII] FWHM < 1000 km/s Line diagnostics  photoionized gas 2. Broad Line Region (BLR) Permitted Lines only T~10,000 K from CIII 977 / CIII]1909 ratios Line widths up to ~40,000 km/sec Recall thermal gas at T=10 4 K has width of ~10 km/sec  Must have Doppler broadening due to bulk motion of emitting gas * Densities n~ cm -3 Since you don’t see forbidden lines, n>n(critical) they must be collisionally de-excited, although you do see broad CIII] so n<n(critical) for CIII] NLRG, Seyfert 2’s: NLR only, BLR obscured or absent Seyfert 1’s, Quasars, BLRG: NLR and BLR visible

Unified Models for Sy1’s and Sy2’s : Antonucci & Miller 1985 ApJ 297, 621 Antonucci 1993 ARAA Vol. 31, p. 473 Spectropolarimetry of the Seyfert 2 galaxy NGC 1068: Polarized Flux shows broad permitted lines: Looks like a Sy 1 in polarized light HβHβ[OIII]

Unified Model: Sy 1’s and Sy 2’s are the same object, seen at different aspect angles. Polarized light is Thomson scattered BLR Obscuring Torus BLR light scattered by electrons to us BLR Scattering electrons

Although local Seyferts, imaged with HST, often show disk-like, dusty structures in their cores on relatively large scales, direct evidence for an obscuring torus, required by the Unified Models, doesn’t really exist. Theoretically, it’s not clear how you’d “make” a torus and why it would be where it is WFPC images of Seyfert Nuclei

Another part of the puzzle : Broad Absorption Line Quasars (BAL QSOs) Warm (i.e. ionized) UV and Xray absorbers Associated Absorbers All these objects have outflowing, radiatively driven winds In a few cases, the absorption varies with time  must be very close to the central engine Typically see very high ionization states, not typical of ISM clouds, e.g. NV, OVI, OVII etc

Although BAL winds are thought to be radiatively driven, it’s hard to have enough radiation force without totally ionizing the material and making it Impossible to radiatively drive out

Perhaps the winds and the BLR are one and the same

Double-peaked Balmer-line profiles are seen in a few AGN Shape and variability of line profile suggest an origin for Hbeta in a rotating disk NGC 1097 Storchi-Bergmann et al. (2003)

Reverberation Mapping Brad Peterson and many collaborators Powerful probe of BLR structure and kinematics Measure the emission-line response to continuum variations Need good enough time sampling to derive time lags between continuum and emission line variations unambiguously Emission lines respond to changes in the continuum flux with a “lag” corresponding to the light travel time from the ionizing source to the BLR First Campaign: NGC 5548 Obs. With IUE every 4 days in season Continuum Emission line NGC 5578

 = r/c “Isodelay Surfaces ” All points on an “isodelay surface” have the same extra light-travel time to the observer, relative to photons from the continuum source.  = r/c Courtesy B. Peterson

Key Assumptions 1Continuum originates in a single central source. –Continuum source (10 13–14 cm) is much smaller than BLR (~10 16 cm) –Continuum source not necessarily isotropic 2Light-travel time is most important time scale. Cloud response instantaneous  rec = ( n e  B )  1  0.1 n 10  1 hr BLR structure stable  dyn = (R/V FWHM )  3 – 5 yrs 3There is a simple, though not necessarily linear, relationship between the observed continuum and the ionizing continuum.

Reverberation Mapping Results Reverberation lags have been measured for 36 AGNs, mostly for H , but in some cases for multiple lines. AGNs with lags for multiple lines show that highest ionization emission lines respond most rapidly  ionization stratification

No significant lag is seen between the optical and UV continuum The delays for the different emission lines are smaller than models predicted The lag increases with decreasing ionization  one zone models are out, must have radial stratification of BLR Revised BLR models since the BLR gas is closer to the ionizing photon source than previously thought, so density must be higher to keep the ionization parameter constant  n~ cm -3 Really optically thick

Time-Variable Lags in NGC years of observing the H  response in NGC 5548 shows that lags increase with the mean continuum flux. Measured lags range from 6 to 26 days Best fit is lag  L opt 0.9 Structural changes in the BLR   L opt 0.9 Optical luminosity Hbeta lag

Locally Optimally Emitting Clouds Baldwin, Ferland, Korista & Verner 1995 ApJ 455, L119 Ferguson ApJ 487, 122 All quasar emission line spectra “look alike” i.e. A single-zone model is pretty successful with a narrow range of density and ionization parameter  implausible “fine-tuning” of BLR parameters Use CLOUDY to make a huge grid of BLR models as a function of density and ionization parameter (i.e. distance from the central engine) For each emission line, there is a narrow range of density & ionization parameter where the line formation is “maximally efficient” and most of the line emission is formed So you can have a completely chaotic BLR with no preferred density, etc. The observed spectrum is some average of the “full family” of models

Locally optimally-emitting cloud (LOC) model The flux variations in each line are responsivity-weighted. –Determined by where physical conditions (mainly flux and particle density) give the largest response for given continuum increase. Emission in a particular line comes predominantly from clouds with optimal conditions for that line. Ionizing flux Particle density