N.G. Bochkarev ( SAI MSU, Moscow), L. Č. Popović ( Astr on. Obs . , Belgrade ),

1. Characteristics of the long-term spectral variability of the AGNs with broad lines in the optical spectral band N.G. Bochkarev (SAI MSU, Moscow), L.Č. Popović(Astron. Obs., Belgrade), A.I. Shapovalova(SAO RAS, N.Arkhyz) + many other contributors X_SBAC Belgrade 31 May 2016

2. Broad and narrow emission lines 10000 км/s

3. Spectral variability NGC 5548

4. X-ray emission variability Кривая блеска 0 50 Время, ксек 150 200 Спектр мощности NGC 4051 -5 -4 Lg ν , Гц -3 -2

5. Unified model of the “central machine” Optical continuum 3--300pc 10^19—10^21cm X-RAY continuum BROAD LINES

6. Pioneering work by Lyuty and Cherepashchuk on BLR size determination out of line variation time lag in respect to the continuum

7. Characteristic time of spectral line variability • Defined by the largest of 3 times (Bahcall, Kozlovsky, Salpeter 1972): • Time of: • 1) ionizing continuum variations; • 2) plasma response to the variations; • 3) light travel between the X-ray source and the line formation region (broad line region (BLR)).

8. Characteristic time of plasma response to X-ray variations Relaxation time of gas with number density n=109 cm-3 is ~1 hour

9. Isochrones of response to ionizing emission variations

10. Computation of spectral line profile response to ionizing radiation fast change Bochkarev & Antokhin,1982; Antokhin & Bochkarev, 1983)

11. One telescope (or a few ones) allows to study slow variations An example is broad Нβ line disappearance in NGC 4151

13. time delay of line flux  size of the BLR  mass of BH

14. Our program of AGN monitoring NGC 5548: Our Hβ and continuum fluxes coincide within the uncertainties with the AGNW data. The correction for different apertures: F(cnt)pet.=F(cnt)our -2.5 ( in units 10-15 erg cm-2s-1A-1 )

15. Aims and outline of our program of AGN spectral and photometric monitoring investigation of the broad line region (BLR) using spectral monitoring to constrain the BLR physics kinematics and geometry needs for accurate BH mass estimates our long-term optical monitoring (10-20 yrs):NGC 5548, NGC 4151, 3C 390.3, Ark564, Arp102b, E1821+643, NGC 7469, Mrk 6, NGC 3516, NGC 7603

16. Observations Long-term monitoring of 3c390.3 20 6m and 1m telescopes - SAO RAS (Russia) 2.1 m telescope - Guillermo Haro Observatory, Cananea, Sonora,Mexico 2.1 mtelescope - Observatorio Astronómico Nacional, San PedroMartir, Baja California, Mexico 3.5 m and 2.2 m telescopes- Calar Alto observatory (Spain)

17. Long-term monitoring (published) • NGC 5548 – 9 years (Shapovalova+ 2004, Ilić 2007+, Popović+2008) • NGC 4151 – 11 years (Shapovalova+ 2008, 2009, 2010a) • 3C390.3 – 13 years (Shapovalova+ 2010b, Popović+ 2011, Jovanović+ 2010) • Ark 564 – 11 years (Shapovalova+ 2012, ApJS) • Arp 102B – 20years (Shapovalova + 2013, A&A; Popović+ 2014) -E1821+643 – 22 years (Shapovalova+ 2016) Search for variability: continuum flux, line shapes, line fluxes …

18. AGN: variable optical spectra continuum Hβ Hα maximum activity minimum activity 3C390.3 (Shapovalova+ 2010) • Everything varies! • line fluxes and continuum fluxes • line profiles (e.g. bumps/asymmetries appears, moving details, outflow) • AGN type changes: type 1 <--> type 2

19. Year-averaged normalized profiles of Hβ in 2000-2003. The vertical dashed line corresponds to a radial velocity +2600 km/s NGC 5548: The radial velocity of the peak decreases: in 2000-2001 it corresponds to ~+(2600-2700) km/s, and in 2002-2003 to ~+ 2000 km/s, within the uncertainties. (Shapovalova+ 2004, A&A)

20. AGN monitoring program outcome – 1 • The size of the "central machine" (or, rather, BLR -- broad line region) is determined. • It proved about two times smaller (its volume almost one order of magnitude smaller), and the average density three time higher than thought before.

21. AGN monitoring program outcome – 2 Stratification of broad line formation region is confirmed: • The higher a ion ionization potential the less the volume it is present in and emits from, • regions of HeII ion line emission are smaller than those of hydrogen emission, • thrice ionized carbon CIV lines are formed inside an even smaller region.

22. BLR stratification and change in size

23. 3C 390.3: (top) – Maximum response of the correlation functions of the Hb line segment light curves with the continuum light curves; (bottom) – Time delay of the individual line segments of Hb for the period 1995-2002 (filled squares) and for 2003-2007 (filled triangles) with respect to the continuum light curve in velocity space. The outer red and blue wings respond much faster to continuum variations then the central part(Popovich+ 2011, A&A)

24. AGN monitoring program outcome – 3 • - A detailed study of iron line variability in active galactic nuclei was accomplished for the first time: • - The behavior of the broad components of FeII optical spectral lines is analogous to that of Нβ lines: the lines of both types are formed practically in the same region.

25. Ark 564 - Nearby narrow- line Sy 1 galaxy with narrow permitted lines (FWHM~2000 km/s) • registered five flare-like events lasting ~1-3 days • almost lack of correlation between the Hα and Hβ line fluxes -> beside the photoionization some additional physical processes may be present * lag of 2–6 days, but with large errorbars • the Fe II emission is probably coming from the intermediate line region with velocities around 1500 km/s • Shapovalova+ 2012, ApJS

26. AGN monitoring program outcome – 4 • Red and blue wing of broad emission lines vary practically simultaneously. • It indicates the prevalence in broad line region (BLR) of circular (and/or chaotic) but not radial (inflow or outflow) movements. • Evidence of virial theorem applicability to AGN

27. Arp 102b:Left - the mean normalized H-alpha profile fitted with disk model;Right – bump in H-alpha profile at ~-2000 km/s in years when it is most intensity (Popović+ 2014, A&A)

28. 5 10-12 4 3 2 1 10-12 NGC 4151 outflow Hα Hβ 2005 – 2003 – 2001 – 1999 – 1997 – Flux in erg cm-2 s-1 Flux in erg cm-2 s-1 years years 2005 – 2003 – 2001 – 1999 – 1997 – 1.2 10-12 1.0 0.8 0.6 0.4 0.2 10-12 Vr, km/s: -5000 -3000 -1000 1000 3000 5000

29. NGC 4151: the Balmer decrement BD(Vr,t)

30. Variations of BLR NGC5548 gas temperature in 1996-2003 Te, K годы

31. AGN monitoring program outcome – 5 • The "central machine" parameters (accretion disc and surrounding gas) undergo substantial changes on a characteristic time-scale 3 – 10 years. • E.g., NGC 5548 broad Нβ line formation region radius varied slowly inside 6 – 26 light day interval over the 15 years of intensive monitoring of the AGN.

32. K-band light curve of NGC 1068 between 1970 and 2006 (Taranova et al., 2006)

33. Long-term variability of NGC 4151 BLR NGC 4151 nucleus activity varied considerably over the last 40 years. • One gets the impression that during that period the "old" accretion disc passed away and another one formed in its place. • The new disc probably lies in a plane not coinciding with the plane of symmetry of the collapsed accretion structure.

34. AGN monitoring program outcome – 6 • 2nd BLR: • broad emission lines are emitted by two kinematically and physically different regions: • BLR1 surrounding the "central machine" (ionized by the emission from the accretion disk and its hot corona); • BLR2 in a subrelativistic flow (ionized by emission from the relativistic plasma in the jet).

35. Bimodal behavior of 3C390.3 cross-correlation function (CCF) in 1997-1999 (Shapovalova et al., 2001). Two other similar examples see in: Peterson+, 2004 30d 100d

36. Separation of the moving jet components(C4-C8) relative to the stationary features: D(filled diamonds) and S1 , S2 (filled circles); apparent speeds:(0.8-1.5)c; Lines-the best linear least-squares fits

37. BLR2: line formation near jets In the case of C4-C7 components the time of breakaway from the stationary S1 component is found to coincide with the maximum in the optical continuum. All the 4 ejection events occur within 0.3 years after a local maximum of the optical continuum intensity is reached.

38. BLR is likely to be generated both:1) near the disk (Peterson et al. 2002) (BLR1, ionized by the emission from a hot corona (Fabian, 2004) or the AD (Field&Rogers 1993) ; 2) in a rotating subrelativistic outflow (Murray et al.1997) surrounding the jet (BLR2, ionized by the emission from the relativistic plasma in the jet) A sketch of the nuclear region in 3C 390.3. The drawing is made not to scale and shows only the approaching jet.

39. NGC 4151 Bochkarev+ 1991, AJ; Bochkarev & Shapovalova 2007; Shapovalova+ 2008, A&A; Bochkarev & Gaskell 2009, AstL; Nazarova & Bochkarev 2011, A&AT Saturationof line fluxes for the high continuum level Fobs(Hβ)=(2.3-9.8)x10-12erg cm-2s-1 Calculations of line response to the continuum flux changes using photoionization model observed line fluxes much larger than computed ones non-photoionized region contributes to the BEL (could be associated with radio jet) Line fluxes vs. density (left) and ionizing flux (right)

40. AGN monitoring program (reverberation mapping) outcome – 7 • The masses of the cores (supermassive black holes, MBH) are determined for about 50 AGN. • Random (statistical) errors of mass determination σ = 0.105 dex = 27%. • The systematic errors are unknown, but probably <0.3dex = 2 times.

41. Methods of estimating the mass of AGN central body • Low limit of the mass from Eddington luminosity (Zeldovich & Novikov, 1964) • Estimation based on the virial theorem (implies the absence of matter outflow; BLR size must be known, -> e.g., from reverberation mapping. • On the base of VLBI observation of megamazers: ~10 AGN ring megamazers are observed on water line 1.35 cm (Moran et al. 1999, …); • Using velocity dispersion or star proper motions in nearby galaxies (e.g., M87).

42. Соотношение масса – светимость AGN

43. Comparison of the mass estimates of Dibai (1984) with reverberation mapping masses • Masses derived from reverberation mapping are in agreement with Dibai’s 1980-1984 estimates (by single-epoch (Dibai) method). • Comparison of masses of 17 AGN covering a mass range 106 – 109 Msun. • The systematic difference between the two methods is only 40% (0.14 dex) lower. • What does the agreement tell?

44. Comparison of AGN mass as derived from reverberation mapping (Mreverb)and estimated by Dibai (Mdibai). 3C273 <lg(MDibai/Mreverb)>= -0.14±0.07 <difference in masses> ~ 40%. Dispersion of the points in the Figure ±0.28 dex. If the errors of the 2 methods are the same, then Dibai's estimation dispersion <σ> = 0.20 dex (1.6 times). NGC4051 lg MDibai lgMreverb = (1.14±0.18) lgMDibai – 0.85

45. NGC 4151 – SMBBH in the center? Popović, 2012, NewAR, 56, 74 Bon et al. 2012, ApJ, 759, 118 69

46. Some conclusions • the broad line region is complex! →different components:disk , outflows, moving details, jets,flare-like events..... → contribution of other physical processes (apart from photoionization) to line formation: shocks, jets, accretion flows around black holes, or bright spots in disk ... → geometry of BLR:a disk-like shape and sometimes an additional non-spherical component; → kinematics: predominantly circular motionand an additional : outflow, inflow.....  single-epoch (Dibai) method and even reverberation method should be used with caution for MBH estimates.

47. Thank you for your attention! 73