Radio-loud AGN energetics with LOFAR

1
Radio-loud AGN energetics with LOFAR
Judith Croston
LOFAR Surveys Meeting 17/6/09
2
Understanding radio-galaxy physics is important
for galaxy feedback models!

• X-ray cavity measurements show energy is
  available to balance cooling in cluster cores,
  but timescales uncertain various detection
  biases.
• When central AGN switches off, up to ¾ of
  available energy still contained within radio
  lobes subsequent evolution of lobe contents
  impact on the cluster depend on cavity particle
  B content.
• FRIs (typical cluster centre sources) and
  powerful FRIIs have different energetics and
  particle/field content (e.g. JC et al. 2004,
  2005, 2008 Dunn et al. 2004, 2005 Kataoka
  Stawarz 2005) understanding the origins of this
  difference is crucial for relationship between
  accretion mode, jet production and feedback.

JC et al. 2003
Wise et al. 2007
3
Radio-galaxy energetics, particle field
content

• k is unknown, and in general B and N0 cant be
  disentangled common to assume minimum
  energy/equipartition.
• The main exception is when inverse-Compton
  emission from the same electron population can be
  detected typically true for FRII radio galaxies
  and quasars.
• Measurements of external pressure/X-ray cavity
  detections can also constrain ETOT (rule out
  equipartition in FRIs).
• Emin and shape of N(E) below observable radio
  region are important low-energy electrons
  dominate relativistic particle population.

4
The low-energy electron population

• Most of the energy density in extragalactic radio
  sources is at energies below currently observable
  radio region.
• Radio-source properties depend strongly on
  assumed spectrum below 300 MHz alow and gmin.
• See discussion in Harris (2004, astro-ph/0410485)

Figures from Harris (2004)
5
Inverse-Compton emission from FRII radio lobes

• IC X-ray emission breaks the ne/B degeneracy of
  radio synchrotron gt direct probe of low-energy
  electron spectrum and of lobe energetics.
• IC useful in jets hotspots too, but for lobes
  beaming other X-ray emission processes
  unimportant.
• In most cases CMB photon field dominates over
  nuclear photons (e.g. Brunetti et al. 1997)
  SSC .
• Can now routinely detect IC emission from the
  lobes of FRII radio galaxies and RL quasars 30
  X-ray detections spanning redshifts of 0.006 2.

Colour XMM IC Contours radio
JC et al. 2004
6
Comastri et al. 2003, Hardcastle et al. 2002,
Brunetti et al. 2002, Isobe et al. 2002,
Hardcastle JC 2005, JC et al. 2004
7
IC/CMB from FRII lobes results for large samples
X-ray detected lobes

• X-ray detection in at least one lobe in 70 of
  X-ray observed 3C FRIIs
• Consistent with IC/CMB with B (0.3 1.5) Beq
• gt 75 of sources at equipartition or slightly
  electron dominated gt magnetic domination must
  occur rarely, if at all.
• Unlikely that relativistic protons dominate
  source energetics.
• Total internal energy in FRII radio sources is
  typically within a factor of 2 of minimum energy
  (see also Kataoka Stawarz 2005)
• But assumptions about the low-energy electron
  population introduce significant uncertainty in
  these results...

Lower limits for non-detected lobes
JC et al. 2005 ApJ 626 733
8
Low-energy electron distribution

• Assume cut-off frequency, gmin 10
• in hotspots, gmin 100 1000 required (e.g.
  Carilli et al. 1991)
• adiabatic expansion gt lower energy electrons in
  lobes
• Assume spectral index, alow 0.5 (flattening)
• shock acceleration models predict d 2 2.3 (a
  0.5 0.7)
• hotspot observations (e.g. Carilli et al. 1991,
  Meisenheimer et al. 1997)

• If gmin 1000 (instead of 10)
• Utot and Bobs/Beq unchanged
• IC/nuclear --
• conclusions not affected
• If gmin 1
• increase in Utot by 25
• small decrease in Bobs/Beq
• IC/nuclear

• If alow aobs
• increase in Utot of up to factor of 20
• Bobs/Beq decreases by up to 60,
• IC/nuclear

9
Spatially resolved IC studies

• Chandra XMM allow us to investigate spatial
  variation of N(E) and B in lobes.
• Lack of correlation between radio and X-ray
  structure indicates N(E) changes alone cant
  explain radio structure changes in B alone cant
  explain relation to radio spectral structure gt
  both are required.
• Also relies heavily on assumptions about low-n
  spectrum...

Isobe et al. 2002
Hardcastle JC 2005
Goodger et al. 2008 in prep.
10
X-ray environments cluster cavities

• FRI radio lobes at equipartition are
  under-pressured relative to their environments
  (e.g. Morganti 1988, Killeen et al. 1988, Feretti
  et al. 1990, Taylor et al. 1990, Böhringer et al.
  1993, Worrall et al. 1995, Hardcastle et al.
  1998, Worrall Birkinshaw 2000, JC et al. 2003,
  Dunn Fabian 2004, JC et al. 2008, Birzan et al.
  2008)
• Either radiating particles field are NOT at
  equipartition or some other particle population
  dominates the source energetics.

Dunn Fabian 2004 MNRAS 355 862
Worrall Birkinshaw 2000 ApJ 530 719
11
Combined X-ray radio constraints favour
entrainment of ICM

• Fraction of energy in radiating particles
  decreases dramatically with distance.
• These constraints rules out relativistic proton
  domination, electron dominance and simple
  B-dominated models (e.g. Nakamura et al. 2006,
  Diehl et al. 2008)
• Consistent with entrained, heated ICM dominating
  radio-lobe energetics.
• Good constraints for models of FRI entrainment,
  but this relies on assumptions about low-energy
  electron population...

3C 31 required entrainment rates
Hydra A missing pressure as a function of
distance
1.4 keV 5 keV 10 keV 50 keV
model (1r/rc) -2.0 (1 r/rc) -1.0 r const.
Comparison with theoretical expectations
JC et al. in prep
12
Calibrating radio-loud (FRI) feedback

• X-ray cavities provide direct measurement of
  energy input to ICM Ekin gtgt Esynch
• (e.g. Bîrzan et al. 2004, Dunn Fabian 2004,
  Dunn Fabian 2008)
• Cavity detection only possible for modest sample
  sizes at low/moderate z and is subject to
  incompleteness problems depends on angle to
  l-o-s, X-ray data quality, cluster luminosity,
  etc.
• Feedback models require radio surveys of FRIs to
  high z to relate direct measurements of energy
  input to RL AGN population statistics.
• Low-n radio spectrum promising for reducing large
  scatter in cavity scaling relations (Bîrzan et
  al. 2008)

Bîrzan et al. 2008
13
What LOFAR will do

• 10-200 MHz observations of large samples of
  radio-loud AGN will determine distributions of
  low-n spectral index ( cut-off in some cases)
  for different radio-loud AGN populations.
• Low-n spectra for large samples of FRIIs with
  X-ray coverage (100 FRIIs)
• determine electron energy distribution for the
  energetically dominant population below g 105
  via X-ray IC constraints on particle
  acceleration
• remove factor 20 uncertainty in ETOT , factor
  2 uncertainty in B assuming CMB dominates IC
  photon seed field in most cases, and uncertainty
  about the role of nuclear IC scattering
• Low-n spectra for very large samples of FRIs,
  including cavity sources, will
• Remove gt order of magnitude uncertainty in
  energetics of radiating particles field in
  FRIs/cluster cavities important to determine
  entrainment and heating rates.
• Allow detailed calibration of AGN heating
  relations via low-n observations of cavity
  samples at low-z
• Apply new calibrations to comprehensive FRI
  samples for tightly constrained AGN feedback
  models