Active galaxies

Manju Sudhakar
Junior Research Fellow, ISAC

Abstract : This essay presents an overview of Active Galaxies -for instance, how
they are classified, what observations tell us about them, what causes the galaxy to
emit enormously luminuos flux.It also dicusses what high energy observations reveal
about these galaxies.

1. Introduction

An active galaxy is a galaxy that has exceptionally high luminosity and which
radiates large amounts of non-stellar radiation.The spectrum of radiation emited by
an active galaxy is very different from that of an ordinary galaxy, which shines with
the combined light of its stars and nebulae.A galaxy is said to be ’active’ if it generates
substantial amounts of energy that is not produced by stellar evolution.The spectrum
of a normal galaxy’ M31 shows prominent absorption lines of sodium,magnesium
and calcium [fig 1], whereas the spectrum of the active galaxy NGC4151 is entirely
different - it is composed of strong, broad emission lines [fig 2].
fig 1

| T
fl:f.' - .

[
e s T
 e

fig 2
 This difference clearly shows that the radiation emitted by NGC4151 corre-
sponds to an energy outburst and some other mechanism is needed to explain this
emission.
In most cases, this non-stellar emission appears to come from the very center or
nucleus of the galaxy.This central region of the galaxy is called the Active galactic
Nucleus or AGN.
The following properties are observed in Active galaxies:
e An overall luminosity exceeding 10** erg/sec.
e A much higher output of X-ray, ultraviolet, infrared and radio radiation than a
normal galaxy.
e A highly luminous and compact central core, known as AGN, which varies rapidly
in brightness.
e Broad emission lines in its spectrum.
o Narrow jets of radiating matter emerging from the central core.
o Large scale clouds of radio-emitting material.
Section 2 briefly covers the classification of active galaxies according to their
appearance, luminosity and spectra.In section 3, the information obtained from ob-
servations of AGN is discussed, while section 4 discusses the variability studies of
AGN.Section 5 discusses the physics behind the central engine of AGN and the uni-
fied model.Section 6 discusses Blazars and what high energy observation of AGN can
tell us about them.Section 7 winds up with current problems and future developments
of the study of these wonderful objects.

2. Classification of Active Galaxies

According to their appearance, luminosity and spectra, active galaxies are clas-
sified into the following types.

1. Seyfert Galaxies : These have the following regions

e Broad Line Region : typically has density > 10 em~? and line widths correspond
to velocities of 1-10,000 kms~" or more.
e Narrow Line Region : typically has density < 107 ¢m™ and line widths correspond
to velocities upto 1000 kms~*.
There are two categories of Seyferts:
e Type 1 Seyferts, with broad allowed lines of HI,Hel,HeIl.. and narrow forbidden
lines, say of O[III] whose FWHM suggests v ~1000-10,000 kms~* and v ~1000 kms~!
respectively.
Particular objects :NGC 4151, NGC 5506, NGC 7469 .
e Type 2 Seyferts, with allowed and forbidden lines of equal width whose FWHM
suggest v < 1000 kms~t.
Particular objects :NGC 424, NGC 1068.
o Intermediate Seyferts, which contain both broad and narrow allowed line com-
ponents.
2. Quasars : Radio telescopes discovered a large number of strong, compact ra-
dio sources, whose optical counterparts were stellar in appearance.These had unique
spectra which consisted of broad emission lines that could not be identified with any
known element or molecule.Later this pattern was recognized as Balmer hydrogen
lines redshifted by a large amount.This corresponded to a huge recessional veloc-
ity.Figure 5 shows the spectrum of 3C273 with z=0.158 and v = 14.6% the speed of
light.
L, ~ 10*L@®.Depending upon the radio emission from these galaxies, they are
categorized as
e Radio Loud Quasars which constitute < 10% of all quasars.
e Radio Quiet Quasars whose radio emission is weak compared to emission in the
optical

3. There are certain objects which show the following properties

e powerful relativistic jets
e strong polarization
e rapid variability
o flat radio spectra
o radio loud.
BL Lac objects and Optically Violently Variable Quasars (OVVs) share these
properties but they have certain distinguishing characteristics.
Initial observations indicated a featureless continuum for these objects but now, with
better detection techniques and more targeted observations, emision lines have been
observed in the optical part of the spectra of these objects.OVVs show the presence
of very strong broad emission lines in their spectra.BL Lacs are further categorized
as
e Radio selected BL Lacs are radio loud objects
e X-ray selected BL Lacs which show strong x-ray emission and are significantly
less radio-luminous than radio selected BL Lacs.
These objects are categorized as Blazars - Bl from BL Lac and zar from quasar.It
is an object that possesses beamed emission from a relativistic jet, aligned roughly
toward the line of sight of the observer.

4. Radio Galaxies : These are powerful sources of radio emission with extended
lobes of synchrotron emission fed by a jet of relativistic electrons originating from the
central region.There are two categories of radio galaxies
e Lobe Dominated galaxies are dominated by radio emission from two extended
lobes symmetrically distributed about a central galaxy.Many of thes show one sided
jets.
e Core Dominated galaxies are dominated by strong radio emission from a compact
core.These too show prominent one sided jets.They exhibit superluminal motion
and often consist of the Blazar family of objects.
Radio emission is a powerful tool in showing that non-thermal emission is tak-
ing place.Where the jet is clearly revealed we see evidence for particle acceleration,
which is a clue to the presence of a central engine.Even when the jets are invis-

4
ible or very weak, the extended radio emission points to this central engine of activity.

3. Observations of Active Galaxies

AGN are the most luminous long-lived sources in the universe, emitting strong
radiation over the entire observable wavelength range, from x-rays and ~-rays through
long-wavelength radio.A complete picture of AGN emission can be obtained by simul-
taneous multiwavelength observations over the entire spectral range.
[fig 3] shows the spectral energy distribution (SED) for radio loud and radio
quiet AGN.SED is the amount of energy emitted in various wavebands.The absence
of high energy observations is due to technical limitations - high energy detectors are
not sensitive enough to detect most AGN - those that have been so far detected are
Core Dominated (CD) Radio Loud (RL) AGN.There is a EUV gap, between Far UV
and soft X-ray due to the large opacity of the ISM of our own galaxy to Hydrogen
ionizing photons.These photons are important because they drive the strong emission
lines seen in the UV and optical spectra of AGN.
fig 3
 The most prominent feature in AGN SED is the strong peak in the UV part
- the Big Blue Bump - which starts in near IR and may extend all the way to soft
x-ray energies and often contains the largest portion of total AGN luminosi
emission in Radio Quiet (RQ) AGN is thought to be dominated by thermal emission
from dust in the host galaxy and that from RL AGN is probably a mix of isotropic
thermal and beamed non-thermal emission, which dominates in CD RL AGN as it is
boosted by relativistic outflow along our line of sight.This is shown in the SED by
the IR bump.
The UV-optical-IR emission from AGN appears to be predominantly thermal in
origin while both the high and low ends are thought to be produced by non-thermal
processes.The hard x-ray spectra of RQ AGN is a power law continuum over the
range ~ 1-100 keV and it turns over at energies >100 keV.RL AGN show a harder x-
ray spectra than RQ AGN and the relative strength of this hard component increases
with the core dominance of the radio source.This implies Compton up-scattering
of radio synchrotron photons.
CD RL AGN are the only AGN detected in high energy ~-rays, implying that
beaming is a key factor.A few very high energy observations - TeV range have been
made using ground based Cerenkov shower detectors.So far all the AGN detected
by this method are all BL Lacs in which the observed emission is dominated by a
relativistic jet.
Therefore, multiwavelength studies of AGN serve two purposes
® observations across the entire electromagnetic spectrum helps to identify the major
emission processes that take place.
e multiwavelength monitoring searches for links between variability at different wave-
lengths.
Strong broad emission lines are defining characteristics of AGN.In the UV-optical-
IR part of the spectrum, the most prominent emission lines are Lya A1215, C IV
A1549, C IIT A1909, Mg II A\2798, HB A4861, Ha A6563 qmong the allowed lines and
O[III] among the prominent forbidden lines.
The Narrow Line Region (NLR) can be spatially resolved in nearer AGN with
HST and even ground based observations.The long axis of the NLR coincides with
the radio axis in those sources in Lobe Dominated (LD) RL AGN.The optical and
UV spectra of AGN contain absorption features due to material along the line of
sight between us and the source.These spectra are dominated by Lya A1215.This is
known as the Lyman alpha forest and provides a lot of information concerning the
distribution of cold material in the universe and is very important for cosmological
studies.
One of the most important observed properties of AGN is the rapid variability
of their continuum and line spectra.Many AGN vary in brightness by substantial
amounts over timescales as short as years, months, days or even hours.The very fact
that a significant fraction of luminosity from an AGN can change on timescales of
100 seconds or less for x-ray emission to less than a year for optical emission
enables conclusions to be drawn on the size of the emitting region.Rapid variability
implies a compact source size, because no stationary source of radiation can
vary in brightness faster than the time it takes for light to cross it.

6
 So we see that, in AGN the combination of the enormous amount of energy emit-
ted and the very small size of the emission region set by the short variability timescale
is not compatible with the energy being produced by normal stellar proce €S,

4. The Central Engine and The Unified Model

The main astrophysical problem is to understand how AGN can generate such
vast amounts of energy from very compact volumes - we need some clue on the central
engine’ of AGN.
The key pieces of observational data we have are
o the extreme luminosities.
o the rapid variability of emission.
Consider [fig 4].Let R be the radius of the emitting region, considered to be
a spherical object.Suppose this region varies in radiation output.Imagine that this
happens at the center and then propagates outward to the limbs.The radiation has
to travel a distance /; from point A, which is closest to the observer while that from
point B has to travel distance lo.Radiation from B reaches the observer after time At
which the timescale of the variabilty, which can be measured.Therfore, At = (extra
distance travelled by radiation)/c.
fig 4
 If the radiation from the point closest to the observer and theat from the limb
subtend an angle 6 at the observer, from the geometry of the figure,

cosf = (i + R)/ly or

ly= (1 + R)/cosf

For cosf~ 1

lp=01+R

Therefore, radiation from B to observer travels extra distance Iy — l; = R, which is

the size of the emitting region.

At ~R/c

For At = 1 hr, R turns out to be 1.08 x 10" ¢cm ~ 7.2 AU.Thus the rapidity of
a luminosity change can be used to set the size of the emitting region.
Typical AGN luminosity is of the order of 10%6 ergs~!.For any sperically sym-
metric object in equilibrium there is an upper limit to the luminosioty it can have -
it should be < Legg = 1.5 x'9%8 M/M ©.Using this and the typical value of AGN
luminosity, the mass of the object should be > 3.3 x 108 M ©.
Therefore, finding such a large mass in such a small space of 7.2 AU can only
mean one thing - that the central engine is a supermassive black hole.This can be
proved as follows.The Schwarzschild Radius of an object is given by

Ry, =2GM/¢? or
M = R,2/2G
Using R, = 7.2 AU = 1.08 x 10" cm in the above eqn, M ~ 3.6 x 10 M @, which
is the same order of magnitude as obtained above.
The above argument supports the widely held view that each and every AGN
contains a supermassive black hole (SMBH), with mass between ten million and sev-
eral billion solar masses that is accreting matter from its surroundings.These SMBH
possess deep gravitational potential wells.The gravitational energy released by the
infalling matter is converted to kinetic energy and it forms a rapidly spinning accre-
tion disk around the SMBH.Kinetic energy provided by infalling matter as it ploughs
through the disk and frictional effects within the disk, raise the temperature of the
disk to very high values and this is observed as UV and x-ray emission. Therefore, the
gravitational potential well of the SMBH is responsible for the enormous luminosity
of AGN.
The observational evidencefor presence of SMBH in AGN are as follows
o X-ray variabiblity : The variability of active galaxies shows the highest amplitude
and shortest timescales in x-rays indicating that this radiation arises closest top the
central source.This is not a very good argument for obtaining the size of the emitting
region because the observed emission may be anisotropic, yielding an underestimate

8
of the emitting region.
® Velocity field of matter close to the nucleus : HST images of the acyive galaxy M81
show the presence of disk like structure in the innermost few arcseconds.Spectroscopy
provided a measure of the velocity of the gas at an angular distance of 0.25 arcsec
from the nucleus, showing it recede from us on one side and approach on the other
with a velocity difference ~ 920kms~1.
o Profile of Fe K emission line : Most convincing evidence comes from the measyre-
ments of the shape of Fe K« line arising from the accretion disk.Inner part if the disk
is illuminated by x-rays.The strongest discrete spectral feature predicted from this
disk is the 6.4 keV Fe K« flourescent line with an equivalent width of 150 eV.Since
it arises from matter in motion, its profile is a tracer of the velocity field of accreting
matter. The [fig 5] shows the line to have the characteristic two pronged shape - mat-
ter approaching us is responsible for the blue wing while that receding produces the
red wing.
fig 5
13010~
]
u
L
aria*
L
 Now that we know the central engine of the AGN, the [fig 6] can be used
to explain the unified model for AGN.The SMBH is the compact central object
surrounded by the BLR and NLR - both of which have a clumpy structure composed
of clouds.A thick dusty accretion torus lies between the two regions.When the line of
sight is such that it passes right through the BLR and NLR towards the innermost
regions of the AGN, we see a Seyfert 1 galaxy with broad allowed lines and narrow
forbidden lines in its spectrum.When our line of sight passes through the NLR but is
blocked by the accretion torus, we see a Seyfert 2 galaxy with narrow allowed and
forbidden lines in its spectrum.From a view point close to the plane of the torus, the
central engine is hidden from view and and we se only the jets and lobes of a Radio
galaxy.The jet radiates most stongly along the axis of the torus and if the line of
sight lies along or very close to this axis, the view of the observer will be dominated
by a head-on jet and a violently variable source with no spectral lines will be seen,
a BL Lac or Blazar.Looking at a modest angle to the jet, we see an unobscured
compact source, a Quasar inside the torus.
fig 6

L Powarful .
Radis
\ . Beams -

Acoracicn
Tonus

T -
- .
Seyferc 2 :
4
, < L3
-~ i Y
hd A ~ - ~ ~
R [ , .
Sepfere I

10
 5. Jets and Superluminal Motion

Jets from AGN highlight the ultimate source of power - the activity at the nu-
cleus of the active galaxy.Extended radio lobes seen in RL AGN are fed by jets that
emanate from the nucleus.Jets are observational clues to the activity in the nu-
cleus.
A jet is a well defined outflow of material or radiation, evidenced by a clear
structure, separate from the surrounding medium.For a structure to defined as a jet,
its length should be atleast 4 times greater than its width.Some AGN show
twin-sided jets, while some show one-sided jets.There are a number of theories ex-
plaining one-sided jets - the material in the jet is moving relativistically and due to
relativistic boosting that part of the jet is brightened, whose velocity vector is aligned
to our line of sight.
VLBI studies show that core emission of these jet sources is very compact and
knowing distance of galaxy through redshift measurements, the brightness temper-
ature of the core can be calculated.This exceeds, in many cases the 10'2 K limit
of synchrotron radiation.These high temperatures correspond to x-ray emission and
therefore we can say this emission is due to synchrotron self-Compton.But this is
not generally observed.So, we can conclude that the emission is due to synchrotron
processes, but the source size has been underestimated due to the assumption that
the emission is anisotropic while it should actually be highly beamed.
The synchrotron emission that we see from the jets tells us that there should
be energetic electrons in the presence of a magnetic field.Also, an ensemble of syn-
chrotron radiating electrons has a lifetime given by

t = (mpc?)’ /107 B2E seconds

The electron lifetimes calculated from this are much smaller than the time re-
quired for electrons to propagate to the outermost parts of the jet.To explain this
some mechanism which supresses synchrotron radiation is required or some sort of
re-acceleration is taking place.But strong polarization is found in the jets, which is a
further argument for synchrotron radiation.
VLBI observations showed that single-sided, parsec-scale jets were composed of
an unresolved core and blobs of emission.Observations taken few years later showed
that the separation of the blobs from the core had increased.
The redshift of the galaxy
can be easily measured from the spectra and using the relation

(14 2)? = (c+v)/(c-v)

the recessional velocity of the galaxy can be determined.From Hubble’s law,

v = H,D

11
the distance to the galaxy can be calculated.From this the velocity of separation of
the blobs can be easily determined.This turned out to be many times the velocity
of light.This is called superluminal motion.It occurs when a relativistically moving
blob in the jet is viewed at a small angle with respect to the jet.
Consider the [fig 7|
fig7

-
Pt '
# a

" it 1 )sin@ é
3 ///flz
Ae= l ! 2
oL i SRR &
V(‘B_ {,‘) cosh
 The blob moves at a speed v ~ ¢ and in time (¢, — ¢,) has moved from position A
to B.The arrival times of the photons at the observer emitted when the blob is at A
is

ty=1ts + da/c
and similarly for photons emitted from position B, the arrival time is

ty=1tp + dg/c

The observer therefore sees the blob executing a transverse motion of amount
v(tg — ta)sing in an observed time

ty — t1=(tp — ta) — (da — dp)/c=(tp — ta)[l — vcosf/c]

The apparent transverse speed is therefore

Vapp = ¢(Bsinf/1 - Bcosb)

where § = v/c.The quantity in the parentheses has a maximum at cos = § with
the maximum value being equal to I'8 where I' = (1 — #2)!/2 is the Lorentz factor of
motion.For § ~ 1 this quantity could be much larger than unity and the blob would
seem to move at an apparent faster than light speed - superluminal motion.

6. Gamma Ray Blazars

High energy v-rays are one of the most direct ways of studying the non-thermal
universe.Sources of very high energy radiation enable us to explore the highest energy
accelerators, with extreme gravitational and magnetic fields.
The most significant thrust to high energy astronomy above 30 MeV came with
the launch of the Compton Gamma Ray Observatory (CGRO) in 1991, which
carried on board the Energetic Gamma-Ray Experiment Telescope (EGRET).
EGRET produced a wealth of astrophysical results and was responsible for the de-
tection of more than 270 point sources of y-rays.
Roughly 60 blazars have been identified with > 100 MeV EGRET y-ray sources.They
show inferred isotropic luminosities ~ 3 x 10%° ergs—!.
The EGRET observations show that
e the vy-ray luminosity often dominates the bolometric power.
o variability on timescales of one day or less is seen from several sources.
Gamma-ray blazars have been detected with redshifts ranging from ~ 0.03 to
~ 2.3.The average blazar spectrum has a photon power law index of about -2.16.It
was assumed that the emitting region was stationary and isotropic.The combined
large inferred y-ray luminosity and short y-ray flux variability data from EGRET ob-
servations implies that the blazar emitting region is very compact.Assuming isotropic

13
emission led to over estimation of the luminosity in radio and optical which inferred
small spatial dimensions.If this emission was due to synchrotron radiation then the
clectrons would cool by Compton scattering with its own synchrotron radiation.This
is the Compton Catastrophe.This is a paradox, because we do see high energy
radiation with large luminosities. This can be resolved if instead of isotropic emission,
the radiating particles are highly beamed .
The radio through optical continua from blazars can be explained by the orien-
tation of our line of sight very nearly along the axis of a relativistically outflowing
plasma jet emited from an accreting supermassive black hole.There are two classes
of models to explain blazar v radiation, where either leptons or hadrons are the
primary accelerated particles.
e Leptonic Model : Here, it is argued that the y-radiation emitted by the blazars
is produced when non-thermal synchrotron-emitting electrons Compton scatter am-
bient soft photons, which could be synchrotron photons produced in the jet, photons
radiated by an external accretion disk which are intercepted by the jet or disk ra-
diation scattered by surrounding clouds of gas and dust before traversing the jet.It
has been argued that due to the differnt energy densities of the external radiation
field, the Compton scattering of external soft photons might produce the bulk of the
blazar jet radiation in flat-spectrum radio quasars (OVVs), whereas the synchrotron
self-Compton process might dominate in BL Lac sources.
e Hadronic Model : Here, accelerated hadronic particles carry most of the en-
ergy. They interact either through secondary production of particles or with photons
to produce electron-positron pairs.These cascade through Compton and synchrotron
processes to produce a power-law spectrum.This will be accompanied bty energetic
neutrino production.If this neutrino flux could be detected from blazar jets then it
would certainly identify hadrons as the primary radiating particles.

7. Current Problems and Future Developments

The variability phenomenon has been best, observed and best understood in the
case of medium luminosity AGN, L ~ 10* — 10% ergs~! but variability data on
powerful quasars up to 10*ergs~! and on very low luminosity AGN L ~ 10¥ergs—*
are still very few.These will make it possible to investigate black hole mass and the
accretion rate.
The directly observable variability of AGN occur on time scales of hours to years,
but other variability phenomena occur on longer timescales, say 10%, 10° or 10° years,
but long term observational programs are difficult to maintain. These are always un-
dersampled and tend to have substantial gaps in coverage.Observations over such
large timescales help in determining the evolution of AGN.
BLR is too small to be spatially resolved, as its angular extent is ~ 10parcsecs
even in nearby Seyfert galaxies.However, it was found that the broad emission line
fluxes vary in response to continuum changes, but the exact time scales for response
remained controversial due to relatively , poorly spaced observations.Line profile
variation has been clearly detected in a few cases, and it was suggested to be attributed
to ”excitation inhomogeneities” resulting from non-uniform illumination of the

14
BLR due to light travel time effects.So, by carefully following continuum variations
and their subsequent effect on the emission line fluxes and profiles, one can constrain
the phase-space distribution of the light-emitting gas.This gives an opportunity
to map out the spatial and kinematic distribution of light-emitting clouds.The math-
ematical formalism for this process was done by Blandford and Mckee in 1982, and
it was called Reverberation Mapping.Reverberation mapping gives a more clear
picture of the physical conditions in the BLR - its density, temperature, column den-
sity and ionization structure.

References
1. Active Galactic Nuclei - Ian Robson.
2. Quasars and Active Galactic Nuclei - Narlikar and Kembhavi.

15