There and Back Again: Cycles of Activity in Radio Galaxies

1
There and Back Again Cycles of Activity in Radio
Galaxies

• Stefi Baum and Chris ODeaRochester Institute of
  Technology

2
The Life Cycle Paradigm

• Birth Infancy (GPS/CSO) Youth (CSS)
  Adulthood (FR1/2) Death Rebirth
• 0 ltNLR (10-1000pc) ltGal (1-20Kpc) IGM
  (20-1000Kpc)
• IR GHZ Peaked 100MHz Peaked
  Steep Spectrum Combo
• 0 102-104 yrs 104-106 yrs
  106-108 yrs Combo

3
Spectral Energy Distribution of AGN

uv
radio
IR
optical
X-ray
Sanders et al, 1988
4
Radio Output continuued

• Early type galaxy AGN hosts have strong radio
  emission, large scale jets

• Late type galaxy AGN hosts have weak radio
  emission, small jets

5

• RGs constrain the total energy output of central
  engine (kinetic plus radiative)
• Lifetime stability of ejection axis constrain
  accretion disk BH physics
• RGs may be an important source of magnetic field
  and cosmic rays in the ICM/IGM
• The interaction of jets, outflow and energy with
  the ambient medium probes the gaseous environment
  of and affects evolution of the host galaxy ands
  it surrounding cluster.
• RGs at high z are obscuration-independent
  signposts to the early formation of massive
  bulges
• But we dont understand how RGs evolve during
  their life times

Cygnus A. VLA 6 cm 0.5 resolution image,
courtesy of Chris Carilli (NRAO)
6
Schematic of Supersonic Jet Model
Models of classical doubles assume radio sources
are supersonic flows. The jets terminate in
strong shocks and fill an over pressured cocoon
which drives a bow shock into the ambient medium.

• Concept from Scheuer 1974, Blandford Rees 1974.
  Illustration from Carvalho ODea 2001.

7
Luminosity Function

• Luminosity function depends on host galaxy
  magnitude AGN luminosity
• Field galaxies 10-1 Mpc-3
• Bright spirals 10-2 Mpc-3
• Bright ellipticals 5x10-3
• Seyferts 10-4 Mpc-3
• Radio galaxies 10-6 Mpc-3
• QSOs 10-7 Mpc-3
• Quasars 10-9 Mpc-3

The fractional luminosity function of radio-loud
AGNs in integral form. Fk(gtP1.4) denotes the
fraction of all galaxies which host a radioloud
AGN with radio power greater than P1.4. (Mauch
Sadler 2007, MNRAS, 375, 931)
8
Demographics of AGN

• Luminosity function is dependent on AGN
  luminosity and host galaxy magnitude
• Radio output is strongly host galaxy dependent
• Strong number density redshift evolution seen,
  paralleling star formation evolution

• 1 of galaxies brighter than L have AGN with
  Lbol gt 1043 erg/sec
• 1 of AGN are radio loud, 99 are radio quiet
• factor of 30 increase in number density from z0
  to z2.

9
Key AGN Questions
10
Key AGN Questions

• What is the realized parameter space of Macc,
  MBH, angular momentum of BH. How is that realized
  parameter space represented in our categorization
  of the Zoo of AGN? Do (how do) individual sources
  evolve in the Macc, MBH, angular momentum space?
• What is the total extracted energey and What
  fraction of extracted energy is emitted
  radiantly, in energetic particles, and in
  directed outflow and how does that relate to the
  basic parameters?
• How are BH demographics, hosts properties and
  galaxy and cluster evolution linked?

11
The Life Cycle Paradigm

• What are the Life Cycles of Radio Galaxies and
  what Physics is Driving them?
• Which galaxies are active and at which times and
  at what luminosities as a function of cosmic
  time?
• How does the radio luminosity function evolve
  with z?
• What is the duty cycle of activity in active
  galaxies?
• What is the Activity Path of an individual
  galaxy as it forms and evolves? How is that
  activity path tied to the evolution of the galaxy
  itself and its nuclear black hole?
• Birth -gt Death -gt Rebirth?

12
The Game Get Clues from Ages Duty Cycles

• The fractional representation of an AGN
  population gives its life time (e.g., Schmidt
  1966).
• Combination of Radio jet modelling, size
  luminosity functions, spectral aging studies,
  stellar age studies, and proper motion/jet
  propagation measurements all used to constrain
  lifetimes and on/off cycles.
• Complications arise because of the many
  manifestations of activity (are they linked
  evolutionarily or are they distinct beasts) and
  the ties to uncertain physical models.

13
Fueling the AGN
kpc - pc scale
gt kpc scale
Harris 1988
lt pc scale
Miracle Inflow? Stellar bars?
Interactions? Feedback?
Reviews Phinney 1994 Rees 1984 Heckman
Balick 1982 Gunn 1979
14
Fueling structures and origins?

• Tens of parsec to 100s of parsecs to MPC

15
Starburst-AGN Connection

• Emerging paradigms?
• A starburst (not necessarily circumnuclear)
  always accompanies the onset of AGN activity
  (e.g., Kaufman et al).
• Early stages of activity, particularly at the
  high luminosity end, will be highly obscured by
  dust and gas. Starburst dating can give
  indication of onset of triggering event.
• In later stages, AGN activity, SNR, and stellar
  winds blow gas out of the central regions,
  exposing the AGN, but reducing available fuel,
  leading to strong frequency and luminosity
  evolution in individual sources. (Sanders et al.
  1988 ulirgs)
• High z and low z triggers, high power and low
  power triggers may be different (accretion vs
  interactions)

16
Baby Sources

• Feeding and
• Growing

17
GHz Peaked Spectrum and Compact Symmetric Objects

• Sources have spectra which peak at GHz
  frequencies.
• Sizes 10-100 pc
• Measured Hot spot separation velocity - 0.42c in
  1943
• Inferred age 1000 yr in 1943

1943546. Contours are global VLBI 4 cm image
from epoch 1997.73. Grey scale is the difference
between this image and one from 1993.17. Black is
positive and white is negative (Polatidis etal.
1999). Ho100
18
Turnover Freq. Vs. Linear Size

• There is continuity in the properties of the GPS
  and CSS sources
• There is a relationship between turnover
  frequency and linear size ?m L-0.65
• Implies the mechanism for the turnover depends
  strongly on source size (likely to be SSA, but
  FFA not ruled out in GPS)

ODea Baum 1997 Fanti etal 1990
19
Why Evolution?

• RADIO MORPHOLOGY AND LUMINOSITY of the GPS and
  CSS galaxies are similar to those of the large
  scale radio sources.
• GPS and CSS sources live in the SAME HOST
  GALAXIES as the large radio galaxies similar
  absolute magnitudes, colors, profiles, evolution
  on Hubble diagram (e.g. ODea etal 1996 Snellen
  etal 1996, 1998 De Vries etal 1997, 1998a,b,
  1999).
• GPS and CSS sources have the SAME AGN BOLOMETRIC
  LUMINOSITY as the large radio galaxies similar
  mid-far IR properties (Heckman etal 1994 Fanti
  etal 1999).
• Confinement by the ISM seems unlikely the HOT
  ISM IS NOT SUFFICIENT TO CONFINE sources (ODea
  etal 1996) and there does not seem to be
  sufficient cold gas to achieve confinement (e.g.,
  Conway 1996 Vermeulen etal 2003 ODea etal
  2003).
• Proper motions 0.1-0.2c detected in dozen CSOs
  (e.g., Owsianik Conway 1998). (THE SMOKING GUN)

• Size 105 pc
• Hot spot separation velocity 0.42c
• Inferred age 1000 yr
• 1943546. Contours are global VLBI 4 cm image
  from epoch 1997.73. Grey scale is the difference
  between this image and one from 1993.17. Black is
  positive and white is negative (Polatidis etal.
  1999). Ho100

20
The Teen Age Years
Still living at home
Calvin becomes A teenager
21
Compact Steep Spectrum Radio Sources

• On host galaxy scales
• Well matched to provide feedback to host galaxy
• Interact with both hot and cold components of ISM

VLA 8.4 GHz image, 0.25 arcsec resolution (Akujor
Garrington 1995, AAS, 112, 235).
22
The Alignment Effect in CSS Sources

• CSS radio sources at all redshifts show emission
  line gas strongly aligned with the radio source
  (De Vries etal. 1997, 1999Axon etal 2000Privon
  et al).
• The OIII clouds have densities in the range
  100-1000 cm-3. Path lengths of 1 pc imply N(H)
  1020 1021 cm-2.
• There is a population of dense clouds throughout
  the ISM which accounts for both the emission line
  nebulae and the atomic hydrogen.
• Jet/cloud interactions may produce observed
  asymmetries in radio structure.
• The bow shock compresses and heats the clouds, to
  sufficiently high temperatures so they need to
  cool before emit OIII.
• R(gap) vhs tcool

(Top) HST/WFPC2 F702W imaging. (De Vries et al.
1997). (Bottom) 3C303.1 z0.270. HST/WFPC2 F702W
and LRF imaging. (De Vries etal 1999).
23
Significant Impact on Host Galaxy

• Cooling time arguments suggest lobe expansion
  speed is 6000 km/s (De Vries et al. 1999). Sound
  speed in 0.8 keV ISM is cs 460 km/s
• Implies M 13, Strong shock Rankine-Hugoniot
  conditions imply T 43 keV.
• Numerical simulations suggest volume of shocked
  gas is 3 times radio luminous cocoon,
• so Vsh 14 kpc3, consistent with XMM spectra

3C 303.1. (Top) Spectral fits to two temperature
model (ODea et al., 2006, ApJ, 653, 1115).
(Bottom) Cartoon of evolution of outline of jet,
cocoon and bow shock (based on Carvalho and
ODea 2002, ApJS, 141, 337)
24
The Adult Years
25
Radio jets and lobes, tens to 1000 KPC
Two main flavors of radio structureFR1 and FR2
Fanaroff and Riley Laing and Bridle 1987 Bridle
et al 1994
26
Fanaroff and Riley Class I Paradigm

• lower power radio galaxies
• jet initially relativistic on parsec scale
• source would be seen as a BL Lac object at small
  inclination angle
• jet decelerates on tens to hundreds of pc scale
  becoming mildly or sub relativistic
• on large scales both jets are seen due to lack of
  strong Doppler boosting
• jet is transonic - no strong shock at end of jet
• bendable by ISM/ICM weather

27
Fanaroff and Riley Class II Paradigm

• more powerful radio galaxies
• jet relativistic on all scales up to hot spot
• quasar or OVV/blazar at small inclination angles
• Doppler boosting on kpc
• scales is important
• typically jets are either both not observed
  (Doppler de-boosting) or only one jet is seen
  (favoritism)
• jet is super sonic - strong shocks at end of jet
• orientation extremely stable
• jet back-flow from hot spot inflates
  over-pressured cocoon (lobes).

28
Two Accretion Modes?

• At a given radio power, FRIs have fainter nuclear
  X-ray and UV emission, suggesting they have
  fainter accretion disks.
• Does the correspond to a different accretion
  mode? (e.g., Fabian Rees 1995, BZO)
• Do RGs change from FR1 to FR2
• or visa versa?

(Top) X-ray luminosity of accretion component vs
178 MHz radio luminosity. Open circles are LERG,
filled circles are NLRG, open stars are BLRG, and
filled stars are quasars. Large circles indicate
an FRI. (Hardcastle etal 2006, MNRAS, 370, 1893.
(Bottom) Optical line luminosity vs. radio core
luminosity. (Baum, Zirbel, ODea 1995, ApJ,
451, 88 BZO)
29
Do Low Power Radio Galaxies Live Longer ?

• Low power sources have spectral and model ages t
  10 7-8 yr
• High power sources have ages t 10 6-7 yr
• This suggests low power sources may live longer
  than high power sources.

Estimated synchrotron ages vs radio power at 1.4
GHz for a sample of radio galaxies. Filled
circles and triangles B2 sources with type 1 and
type 2 spectra, respectively crosses 3C
galaxies with z lt 0.2 open circles 3C galaxies
with z gt 0.2 asterisks 3C quasars. (Parma et
al. 1999)
30
Old Age and Death
31
Dying Radio Sources

• When the jet outflow stops the compact
  components core, jets, hot spots should fade
  first
• The lobes will then become more diffuse and their
  spectra will steepen due to radiative losses.
• Observed examples of dying sources are rare only
  about ten candidates are currently known (e.g.,
  Cordey 1987 Parma et al 2007). Suggests time
  over which a dying source is observable
  recognizable is short (10 of source active
  lifetime, Parma etal 2007).

32
Candidate Dying Radio Source

• Candidates selected from steep spectrum sources
  in the WENSS catalog.
• WNB1150.03749 shows no bright core, jets, or hot
  spots.
• The integrated spectrum is steep (implies age of
  50 Myr)
• The spectrum also steepens away from the center.
• (Parma et al. 2007).

(top left) total intensity image (contours) on
DSS2 image. (top right) integrated radio
spectrum. (bottom left) contours of radio
emission on spectral index map. (bottom right)
spectral index profile along the lobes (Parma et
al. 2007).
33
Reincarnation
34
Cluster Feedback May Suggest Repetitive Activity

• multiple episodes of radio activity
• May be suggested by the X-ray morphologies?
• are required to account for the heating of the ICM

(Top) Unsharp masked image from Chandra Image.
(Bottom) Radio image in blue superimposed on
pressure difference map in red (Fabian et al 2006)
35
Born Again Radio Galaxies?

• About 10 of GPS sources exhibit faint, diffuse
  extended emission.
• Some of these could be contamination by
  core-dominated sources with peaked core spectra.
• Others could be genuinely young sources which are
  repetitive (e.g., 0108388, Baum et al. 1990).
• These born again sources show the current radio
  source propagating outwards amidst the relic of
  the previous epoch of activity.
• In double-double sources, the previous radio
  source is still identifiable as a classical
  double source (Schoenmakers et al. 2002). (and
  dont forget the X shaped sources)

36
Double-Doubles

• 5-10 of gt 1 Mpc radio sources show
  double-double structure.
• Working hypothesis the radio galaxy turned off
  and then turned back on --creating a new double
  propagating outwards amidst the relic of previous
  activity.
• Schoenmakers etal (2000)

Schoenmakers et al. (2002)
37
Multiple Episodes of Fueling?

• Numerical simulations suggest that during the
  course of a merger there will be multiple
  episodes of accretion onto the nucleus producing
  a high luminosity phase lasting about 108 yr.

Projected gas density color coded by temperature
(box 140 kpc across). Bolometric luminosity of
central black hole(s), with diamonds marking the
times shown above. (Hopkins et al. 2005, ApJ,
625, L71
38
3C236 - 4 Mpc Radio Source

• The largest radio galaxy known.
• WSRT 92 cm image (55x96) Mack etal. 1997)
  overlayed on DSS image.

39
The Inner 2 Kpc Double

Inner 2 kpc double is well aligned with outer 4
Mpc double Whatever provoked second epoch did not
affect orientation of jet
Global VLBI 1.66 GHz image (Schilizzi etal 2001)
superposed on HST WFPC2 V band image At z0.1,
and Ho75, 1 arcsec 1.7 kpc
ODea, Koekemoer, Baum, Sparks, Martel, Allen,
Macchetto, Miley (2001)
40
Time Scales

• Dynamical Ages
• Large radio source t7.8x108 (v/0.01c) yr
  (comparable to the age of the oldest blue knots)
• Small radio source t3.2x105 (v/0.01c) yr
  (much younger than the youngest blue knots)
• Dynamical time scale of the disk on the few 100
  pc scale t107 yr

ODea et al 2001
41
The 3C236 Scenario

• The small and large radio sources are due to two
  different events of mass infall.
• Spectral aging estimates in the hot spots of the
  large source imply the radio source may have
  turned off for 107 yr in between the two
  events.
• The difference in the ages of the young and old
  star formation regions also implies two different
  triggers.

42
Constraints from Radio Source Models

• Radio Power vs. Size
• Number vs Size
• Hot spot diameter vs. Radio Source Size
• Velocity vs Size (soon)

(Left) Power vs Size (Blundell et al 1999)
(Middle) Number vs Size (ODea Baum 1997)
(Right) Hot spot size vs. radio source size
(Barai Wiita 2007)
43
Power Size Diagram

• The combined sample of GPS, CSS and 3CR covers 5
  orders of magnitude in linear size.
• Radio galaxies will follow evolutionary tracks on
  this diagram (Baldwin 1982).
• The number in bins of linear size constrains the
  evolution.

ODea Baum 1997
44
Number vs. Size

• Over the entire range N L 0.25
• There is a possible flattening at small size
• Models which fit the data include
• Intermittent sources (Reynolds Begelman 1997)
• A subpopulation which is disrupted on small
  scales (Alexander 2000)

(Top) data from the GPS, CSS, and LRL samples
(ODea and Baum (1997, see also Fanti 2008)
(Middle) Fit to the data which includes
intermittent radio sources (Reynolds Begelman
1997). (Bottom) Fit to the data which includes
sources which disrupt on small scales (Alexander
2000).
45
Most Models Give Similar Results on the
Power-Size Plot for Large Sources

• Models assume same jet physics and ambient medium
  profile
• Some differences are
• Spectrum of electrons accelerated in the hot spot
• Transport of relativistic electrons from hot spot
  to lobe
• Prescription for evolution of hot spot size

Comparison of Models from Barai Wiita (2007).
BRWBlundell et al 1999, MBRWmodified BRW KDA
Kaiser etal (1997) K00 modified KDA
MKManolakou Kirk (2002) MMKmodified MK
46
Compact Sources May Initially Brighten. Slow then
speed up.

• Models by Alexander (2000) and Snellen et al
  (2000) suggest that compact sources will brighten
  for the first kpc.
• Hotspot advance speed

(Top) Models from Alexander (2000). (Bottom)
Kawakatu et al 2008 (doesnt fit medium sized
source measurements to date).
47
Timescales, on off cycles
Shabala et al, 2008 Suggest More massive
galaxies have longer lived radio sources that are
triggered more frequently.
48
Summary

• GPS/CSO 100-1000 yr (proper motions, spectral
  aging)
• CSS 104-106 yr (spectral aging)
• FRII 106-107 yr (spectral aging, statistics,
  dynamical models)
• FRI 107-108 yr (spectral aging, statistics, maybe
  models)
• If there is rebirth, are sources reborn with the
  same or differing Macc, MBH, angular momentum of
  BH etc.? Once an FR2 always an FR2?

49
Questions
From the day we arrive on the planet And
blinking, step into the sun There's more to see
than can ever be seen More to do than can ever be
done There's far too much to take in here More to
find than can ever be found But the sun rolling
high Through the sapphire sky Keeps great and
small on the endless round It's the Circle of
Life (and, perhaps, of Radio Galaxies -)
(Tim Rice)
50
The End