Документ взят из кэша поисковой машины. Адрес оригинального документа : http://lnfm1.sai.msu.ru/~rastor/Ralf_Bender/imprs_chapter_12_2007_active_galaxies.pdf Дата изменения: Mon Mar 3 12:04:06 2008 Дата индексирования: Mon Oct 1 23:07:58 2012 Кодировка: Поисковые слова: закон вина

Page 1

I*M*P*R*S on ASTROPHYSICS at LMU Munich

Astrophysics Introductory Course
Lecture given by:

Ralf Bender and Roberto Saglia in collaboration with: Chris Botzler, Andre Crusius-WДtzel, Niv Drory, Georg Feulner, Armin Gabasch, Ulrich Hopp, Claudia Maraston, Michael Matthias, Jan Snigula, Daniel Thomas
Powerpoint version with the help of Hanna Kotarba

Fall 2007
IMPRS Astrophysics Introductory Course Fall 2007


Page 2

Chapter 12 Active Galactic Nuclei (AGN) and Supermassive Black Holes

IMPRS Astrophysics Introductory Course

Fall 2007


Page 3

Typical signs of nuclear activity are (not all present always):
compact, very bright centers, Rnucl 3pc spectra with strong emission lines ultraviolet-excess X-ray emission jets and double radio sources with Rjet ~ kpc -Mpc variability over the whole spectrum on short timescales: tvar ~ minutes... ~ days AGN luminosities:

Lnuc = 1045 - 10

48

erg s

1012 - 1015 L

IMPRS Astrophysics Introductory Course

Fall 2007


Page 4

12.1 AGN types
12.1.1 Radio galaxies
Radio galaxies emit extremely high radio luminosities: Lradio 108 L E.g., Cygnus A is the second brightest radio source on the northern sky, with a luminosity Lradio ~ 1011L. Cygnus A is a typical radio galaxy and was discovered in 1946 by Hey. 1954 Baade and Minkowsky identified it optically with a giant elliptical galaxy, showing dark dust lanes and a central emission of H lines. The radio emission comes from two extended emission regions (radio lobes) outside of the galaxy. The radio lobes receive their energy from jets which originate in the nucleus and extend 0.2 Mpc. The radio radiation is produced by synchrotron emission of relativistic electrons. Radio galaxies are giant particle accelerators with Ee 1012 eV Radio surveys found radio `galaxies' up to redshifts of z = 4 - 5. Of these about 50% are (relatively nearby) E0/S0 galaxies, and 50% are quasars. The jets typically extend between 0.1 and 0.5 Mpc. Jets may appear one-sided because of `Doppler-boosting'.

IMPRS Astrophysics Introductory Course

Fall 2007


Page 5

Examples of Radio Galaxies

Cygnus A Radio jet: long term stability required

IMPRS Astrophysics Introductory Course

Fall 2007


Page 6

Relativisitic motions

The optical jet of the nearby radio galaxy M 87 (the central galaxy of the Virgo cluster in a distance of about 17Mpc). The jet is highly collimated and shocks are visible within the jet. The emission is synchrotron radiation from relativistic electrons.
IMPRS Astrophysics Introductory Course Fall 2007


Page 7

1963
Maarten Schmidt: 3C273, a star-like object with large redshift, Nature (1963): The stellar object is the nuclear region of a galaxy with a cosmological redshift of 0.158, corresponding to an apparent velocity of 47,400 km/s. The distance would be around 500 megaparsecs, and the diameter of the nuclear region would have to be less than 1 kiloparcsec. This nuclear region would be about 100 times brighter than the luminous galaxies which have been identified with radio sources so far...

3C 273
IMPRS Astrophysics Introductory Course Fall 2007


Page 8

12.1.2 Quasars
1963 M. Schmidt discovers that the radio source 3C273 can be identified with an optical point source (stellar) with a jet. The spectrum shows broad emission lines H,,..., MgII, OIII . . . which are redshifted by z = 0.158 vrad = 47400km/s . So, the object was called a QUAsi StellAr Radio source QUASAR. In 1965 A. Sandage discovers many more objects which show the typical qualities (colours, spectra, redshift, luminosity) of 3C273 but are missing the radio emissions, these are called Quasi Stellar Objects QSO. Today it is established that Quasars and QSO's are similar phenomena, but 90% of the optically found QSOs are radio quiet and 10% are radio loud. Quasars are particularly bright and compact centers of galaxies which outshine the rest of the galaxy. Quasars are mostly found in elliptical galaxies.

IMPRS Astrophysics Introductory Course

Fall 2007


Page 9

The Quasar QSO 1229+204:

IMPRS Astrophysics Introductory Course

Fall 2007


Page 10

A typical quasar spectrum in the optical and UV range. (see also diagnostic plots in the ISM Chapter)
IMPRS Astrophysics Introductory Course Fall 2007


Page 11

Quasar Statistics:
luminosities: Lqua
sar

1045

-48

erg/s

variability in the complete electromagnetic spectrum synchrotron jets extending between 0.1pc and 1Mpc. about 104 quasars are known, of these 10% are radio loud but many surveys continue to discover quasars (e.g. SLOAN Digital Sky Survey) redshifts: z = 0.1...5.8, maximum space 10-5 Quasars/galaxy, and at z = 2: 10-2 one out of 100 galaxies forms a quasar quasars are short-lived: tQSO 107yrs, all contain a black hole. density around z = 2...3 (roughly we have today: Quasars/galaxy). If quasars live long, then only ( 1% of all galaxies contains a black hole). If then all luminous galaxies were once active and

IMPRS Astrophysics Introductory Course

Fall 2007


Page 12

When Active Galactic Nuclei were most active...

Schmidt, Schneider & Gunn 1991, in The Space Distribution of Quasars (ASP),109 Quasars were ~1000 times more numerous per comoving volume at redshifts of 2...3 than today. Where are the local quasar remnants?
IMPRS Astrophysics Introductory Course Fall 2007


Page 13

12.1.3 BL Lac Objects
BL Lac Objects are quasars with enhanced continuum emission and (almost) no emission lines. They are: highly variable extremely luminous highly polarized Presumably the jet is pointing to us and we directly look into the central machine.

IMPRS Astrophysics Introductory Course

Fall 2007


Page 14

12.1.4 Seyfert Galaxies
First discovered in 1943 by Seyfert and Slipher, these are spiral galaxies showing: very bright unresolved nuclei, with luminosities L 1042 Quasars)
-45

erg/s (less luminous than

line emission of highly ionized atoms which cannot be produced by stars sometimes very broad lines of the permitted hydrogen lines (Seyfert 1, otherwise Seyfert 2). a wide range of variability in the broad lines and the continuum (even including their disappearance) in a time range from hours to days. This implies that the Broad-LineRegion (BLR) has a size of 1/100 pc RBLR 1pc

IMPRS Astrophysics Introductory Course

Fall 2007


Page 15

12.2 Structure and physics of AGNs
12.2.1 Sizes
The variability of AGN can be used to gather information about the size of the emission region. Assuming that the state of the emission region is changed by a physical process, two timescales are important: process: time scales for synchrotron radiation, heating, cooling, acceleration . . . t: crossing time needed to cross the emission region

t



process

If the state change spreads with c, then the size of the emission region will be:

lemis = c t
For the observed timescale of the variability tobs applies:

t
In any case applies:

obs

=

process

+ t

c t

obs

lemis

IMPRS Astrophysics Introductory Course

Fall 2007


Page 16

Characteristic time and length scales

radio/optical tobs 10d lemis 0.01 pc radio/optical tobs 1d lemis 10-3 pc TeV tobs 1h lemis 10-5 pc
In comparison: Schwarzschild radius RS = 2GMBH/c
2

M /M 106 108 109

RS 10-7 pc 10-5 pc 10-4 pc

variability in the vicinity of super massive black holes?

IMPRS Astrophysics Introductory Course

Fall 2007


Page 17

12.2.2 Luminosity source
Stars Assuming O stars with a luminosity L* 105.5L and a typical mass M* 50M, then to reproduce the luminosity of the AGN LAGNtotal = N* · L* a total number of N* = 3 · 108 O stars (with a mass M = N* ·M* = 1010M) would be needed. This would result in a stellar density

n =
and a mean distance

N

3

4 l 3 2
1/ 3

2 1014 pc

-3

1 l = n

750 R

25 (2 R )

Such a high stellar density would lead to collisions, dynamical instabilities and presumably to a partial collapse of the system. Observations of AGNs do not show stellar spectra. This scenario is impossible. (It even gets worse with smaller stars)
IMPRS Astrophysics Introductory Course Fall 2007


Page 18

Supernovae The brightest supernovae reach in the maximum 1010L. So 104 supernovae in the maximum would be permanently needed, or, because of ESN 1052 erg up to 1010 supernovae within l 10-3 pc in 107 years. This would need the formation of 1010 stars that are permanently producing supernovae, resulting in the same problems as the last scenario A successive formation of stars while the supernovae explode is impossible No supernova spectra were observed

IMPRS Astrophysics Introductory Course

Fall 2007


Page 19

The luminosity of active nuclei is due to accretion onto black holes. (Zel'dovich 1963)
The total energy output from a quasar is at least the energy stored in its radio halo 1054 J, via E = mc2, this corresponds to 107 Msun. Nuclear reactions have at best an efficiency of 0.6 % (H burning). So the waste mass left behind in powering a quasar is 109 Msun Rapid brightness variations show that a typical quasar is no bigger than our Solar System. But the gravitational energy of 109 Msun compressed inside the Solar System 1055 J, i.e. 10 times larger than the fusion energy. "Evidently, although our aim was to produce a model based on nuclear fuel, we have ended up with a model which has produced more than enough energy by gravitational contraction. The nuclear fuel has ended as an irrelevance." Donald Lynden-Bell (1969) This argument convinced many people that quasar engines are supermassive black holes that swallow surrounding gas and stars.
IMPRS Astrophysics Introductory Course Fall 2007


Page 20

The standard model of AGNs: Accretion onto massive black holes
The AGN contains a black hole with a mass M 106 . . . 109.5M that accretes 10-4 . . . 10 M/yr gas from a surrounding disk. The jets and the nonthermal radiation are created by the rotating magnetosphere of the accretion disk. (Transformation of gravitational energy into thermal energy and radiation). We have discussed the basics physics of accretion disks in the chapter above and in the Stellar physics chapter. There we showed that accretion onto a black hole can provide the following luminosity:

LAcc

1 mc 16 LH

2

1g 106 kWh

Again, for comparison, the efficiency of hydrogen burning is:
- burn

0.007mc

2

An estimate of the maximal possible luminosity of an AGN is given by the EddingtonLuminosity LEdd (see Chapter 4). It is reached, when the radiation pressure is higher than the gravitational acceleration per area.

LEdd =

4 GM

BH
-

m

p



Te

IMPRS Astrophysics Introductory Course

Fall 2007


Page 21

or:

LEdd = 1.3 10

38

M BH erg M s

This implies for Seyfert galaxies, and

M

BH

107 M

M

BH

109 M

for quasars. Using
L
Acc

1 mc 2 = LEd 16

d

yields the corresponding maximum accretion rate:
mEdd 5 10
-10

M BH M M yr

s

The typical temperatures of accretion disks have also been derived in chapter 4 where we obtained:
M T = 1.3 107 K BH M
-1/ 2

(dm / dt ) -10 10 M / yr

1/ 4

R 3kmM BH / M

-3 / 4

IMPRS Astrophysics Introductory Course

Fall 2007


Page 22

If we assume R RS, then the last bracket is 1; inserting the Eddington accretion rate, we then get:

T
i.e. for a typical quasar:
qu TAccasar

M 2 107 K BH M

-1/ 4

105 K Radiation in the UV

and the accretion disks of smaller black holes are hotter

IMPRS Astrophysics Introductory Course

Fall 2007


Page 23

IMPRS Astrophysics Introductory Course

Fall 2007


Page 24

12.3 The Unified Model of the Active Galactic Nuclei
Black Hole in the center: MBH ~ 106 . . . 1010M. Accretion disk extending to ~ 100 - 1000RS, that is emitting radiation in the X-ray, EUV, UV, . . . optical and TeV. Broad line region: Clouds of thick gas (ne 109 -1010cm-3) that are moving with vBL 104 km/s around the black hole and extend to ~ 0.1 . . . 1pc. Emission of broad allowed lines.
R

Narrow line region: Clouds of thin gas (ne 105cm-3) that are moving with vNLR 102 - 103 km/s around the black hole and extend to some pc. Emission of narrow allowed and forbidden lines. Dust/molecular torus with inner radius: ~ 1pc and outer radius: ~ 50 - 100pc produces IR - mm emission. Jets: Synchrotron radiation over the whole spectrum on scales from 0.1 - 106pc.

IMPRS Astrophysics Introductory Course

Fall 2007


Page 25

IMPRS Astrophysics Introductory Course

Fall 2007


Page 26

The diversity of AGN types can be explained by the aspect angle under which we observe the AGN and by the evolution AGNs.
AGN type BL Lac radio loud quasar radio galaxy radio quiet quasar Seyfert AGN type broad strong radio galaxy strong/weak weak radio galaxy BL Lac radio quiet quasar Seyfert weak none strong none, strong/weak line of sight directly into the jet 20° - 70° 20° - 90° 20° - 70° 20° - 70° emission lines narrow strong/weak weak none/weak strong/weak strong/weak Sa- Sbc
Fall 2007

evolution M strong, jet M maximum, jet with vjet ~ c M mean, jet with vjet < c M maximum, no jet M weaker, no jet host galaxies type E E E luminosity high high high high

IMPRS Astrophysics Introductory Course


Page 27

Formation paths for supermassive black holes (by M. Rees).

IMPRS Astrophysics Introductory Course

Fall 2007


Page 28

12.4 Supermassive black holes in nearby galaxies
Some key questions:
Where are the dead quasars in the local universe? (in all galaxies, in some only?) How are black hole masses related to galaxy properties? Do all galaxy centers contain massive black holes? Do all massive black holes live in galaxy centers? Have we really discovered black holes? (or only clusters of compact objects?) What spin do black holes have and what is happening in the immediate vicinity of black holes? Can we find binary black holes? Are black holes sources of gravitational radiation? When and how are the first massive black holes formed? How do they grow? Did seed black holes of 105-6 Msun exist? How do black holes influence the early formation (and evolution) of galaxies (and vice versa)?
IMPRS Astrophysics Introductory Course Fall 2007


The closest confirmed supermassive black hole... ...is found in the Galactic center

Page 29

Mellinger 2000

Genzel et al. 1992-2003, Ghez et al. 1998-2003 Star S2: Movie Vmax Dmin Black hole: MBH RS ~ ~ ~ ~ 5000 km/s, 18 Billion km 3 Million Msun 9 Million km
Fall 2007

IMPRS Astrophysics Introductory Course


Page 30

no viable alternatives to a black hole in the Galactic center
size (Rs(3x10 M )) 10
6

6

1

10

3

10

5

3x10 M BH

1024

boson star SgrA* size/motion relative S2 >105 yrs S2 orbit

10

2

density (M pc )

10

19

10-3 10-8
gas motions mini-spiral

10

14

SgrA* cluster proper motions fermion ball (17 keV) >10 yrs >1010 yrs
9

109

density stellar cusp at 0.5" density nuclear star cluster at 0.3 pc

10-13

10

-7

10

-5

10

-3

10

-1

size (pc)
IMPRS Astrophysics Introductory Course

(Genzel et al. 2003)

density (g cm )
Fall 2007

-3

-3


Page 31

Water Masers disks: NGC 4258

0.14 pc 0.28 pc

Keplerian rotation

mas

Warped disk of gas. Water maser emitting at 2 GHz observable with VLBI reaching 0.5 mas precision. Inner orbits just 40000 Schwarzschild radii.

M

BH

= 3.9 в 107 M

> 4 в 109 M pc

-3

IMPRS Astrophysics Introductory Course

Fall 2007


Page 32

The Mass of the M 31 black hole
(3-7)в107 M (Dressler & Richstone 1988), (0.05-1)в108 M (Kormendy 1988), (4-5)в107 M (Richstone et al. 1990), ~7в107 M (Bacon et al. 1994), (0.7-1)в108 M (Emsellem & Combes 1997) (1.5-4.5)в107 M black hole location (Kormendy & Bender 1999) ~1в108 M , eccentric disk model (Peireis and Tremaine 2003) ~1.2в108 M , blue cluster dynamics (Bender et al. 2005)

IMPRS Astrophysics Introductory Course

Fall 2007


Page 33

Finding black holes in galaxy centers ...
Passive black holes can only be detected if they noticeably influence the motion of stars and gas at radii which we can resolve observationally:

we need very good spatial resolution HST or adaptive optics required.
IMPRS Astrophysics Introductory Course Fall 2007


Page 34

Hubble Space Telescope

IMPRS Astrophysics Introductory Course

Fall 2007


Page 35

Beyond the Galaxy and M31 ...
Rotating gas disks provide a comparatively easy way to find black holes: Gas is collisional and dissipative and prefers circular orbits. Keplerian velocity profiles are good indicators for a central point mass. Caveat: strong non-circular motions can be present if the potential is nonaxisymmetric or if non-gravitational forces (radiation) are important.

Harms, Ford et al., HST
IMPRS Astrophysics Introductory Course Fall 2007


Page 36

Stellar motions are more complicated to model but provide more reliable black hole masses: - stars move collisionless. - stellar motion is only affected by gravity. - however: anisotropy needs to be measured.

IMPRS Astrophysics Introductory Course

Fall 2007


Page 37

A simple example of how an unknown orbital structure prevents an accurate mass determination: if only the velocity at the pericenter is known the mass is uncertain by at least a factor 2.

Tangential and radial orbits correspond to different shapes of the line-of-sight velocity distributions. The different shapes can be measured and described by Gauss-Hermite functions (h3 = asymmetric component, h4 = symmetric component).

IMPRS Astrophysics Introductory Course

Fall 2007


Page 38

Modeling of stellar systems using Schwarzschild's method (1979): (Richstone&Tremaine 1988, van der Marel et al. 1998, Gebhardt et al. 2000):
deproject observed surface brightness profile to derive 3D axisymmetric density distribution of stars (needs inclination) choose a mass-to-light ratio for the stars and derive the potential from Poisson's equation; add the potential of the BH calculate several thousand orbits with different energies, angular momenta and drop points and derive their time-averaged density distribution superimpose the orbits such that: (1) the surface brightness distribution is matched, (2) the velocity distribution (rotation, dispersion, higher moments) is matched (3) the phase space distribution is smooth (e.g. by maximizing the entropy) repeat this procedure for a range of inclinations, stellar mass-to-light ratios and black hole masses.
IMPRS Astrophysics Introductory Course Fall 2007


Page 39

Demographics of black holes in nearby galaxies

M 87: M

BH

~ 2 109 Msun

M 104: MBH ~ 5 108 Msun

M 31: M M 32: M

BH BH

~ 1 108 Msun ~ 3 106 Msun

M 33: M

BH

< 1500 Msun
Fall 2007

IMPRS Astrophysics Introductory Course


Page 40

K line emission in X-ray
Fluorescence Fe emission at 6.4 keV observed in 80% of Seyfert I galaxies thanks to spectroscopic Xray Telescopes (ASCA and XMM). The line is intrinsically narrow, but seen extremely broadened (2 keV ~ 0.3 c) and skewed. Rapidly rotating disk near SMBH: double horn signature like HI profile of spiral galaxies. Blue peak: relativistic beaming Red "peak": smeared due to Gravitational redshift

4

5

6 keV

7

8

Nandra et al. 1997, ApJ, 477, 602

IMPRS Astrophysics Introductory Course

Fall 2007


NGC 5548 Light curve CC with UV

Reverberation mapping

Page 41

In Seyfert 1 the broad absorption region is visible. Any variation in the ionizing flux continuum will cause a flux variation of the emission lines. The time delay between the variations is proportional to the size r of the region. r ct Moreover, the width of the lines gives the velocity dispersion of the clouds. The virial theorem gives:

M

BH

= fr 2 / G

The factor f factories the uncertainties due to geometry. Note that the method does not depend on distance and r can probe regions as small as 1000 Schwarzschild radii. Peterson, 2002
IMPRS Astrophysics Introductory Course Fall 2007


Page 42

Demographics of black holes in nearby galaxies
Kormendy and Nukers 2002

All spheroids contain supermassive black holes with: MBH ~ 0.002 Msphe Pure disk galaxies do not seem to contain black holes.

roid

IMPRS Astrophysics Introductory Course

Fall 2007


Page 43

Demographics of black holes in nearby galaxies
Kormendy & Gebhardt 2001 Gebhardt and Nukers 2000; & Merritt 2000

We see a very tight correlation between BH mass and spheroid velocity dispersion: MBH ~ 0.1 4 (units: solar masses, km/s) At a given mass, more compact bulges contain more massive black holes, i.e., if baryons collapsed and dissipated more than on average, then the BH is bigger. Black hole growth and spheroid formation & evolution proceeded in lockstep.
IMPRS Astrophysics Introductory Course Fall 2007