Cooling flow

1
Cooling flow

• Adriana Gargiulo
• Seminario
• Corso di astrofisica delle alte energie

2
the fundamental parameter
ICM energy
Cooling time
Energy loss due to radiation in X rays

• nH proton density
• L(T) value of cooling function at temperature T

tcool lt tHubble cooling happens
tcool nH-1 !!!
NOTE
3
Observational evidences for cooling flows -
Imaging
Telescopes
Copernicus

• Surface brightness (SB) strongly peacked at the
  center.
• Since SB depends upon the square of gas density
• very short tcool

4
Observational evidences for cooling flows

• Rcool radius at which tcool tH.
• P(r rcool) weight of overlying gas
  where cooling is not important.
• P(r lt rcool) cooling reduces the gas
  temperature to maintain the pressure the gas
  density must rise

Rcool

The gas must FLOW inward
5
Observational evidences for cooling flows

• In absence of a suitable fine tuned heating
  source, the cooling and condensation of the gas
  in the central region is a straight-forward
  consequence of the basic energy equation of the
  hot gas.
• Silk (1976)
• Fabian Nulsen (1977)
• Cowie Binney (1977)
• Mathews Bregman (1978)
• Fabian et al. (1984)
• Fabian (1994)

70 - 80 of clusters have a cooling flow
common and long - living
6
Observational evidences for cooling flows -
Spectra
Independently Strong
support from spectroscopic observations
The spectra show the existence of low temperature
phases in addition to the hotter temperature gas.
Einstein Observatory Solid State
Spectrometer Focal Plane Cristal Spectrometer
White et al. 1994 Allen et al. 1994
7
Observational evidences for cooling flows -
Spectra

• Strong cooling flow found from images where not
  confirmed by spectraBUTA478 SSS data show
  strong X rays absorption.

Allen et al. 1993
8
Imaging vs spectra

• Two different kind of observation lead to the
  same result

.
Ms mass deposition rate computed
from spectra analysis MI mass
deposition rate computed from image
analysis
.
White et al. 1991
9
The cooling flow problem

• X vs Optical
• X large cooling rates of the KeV gas in
  the centers of clusters (tens to hundreds
  of solar masses per year).
• Optical small star formation rates observed in
  central cluster galaxies (few to
  several tens of solar masses per year).

Only a small fraction of the cooled gas can form
stars with a normal IMF most must remain dark

Fabian, 1994.
10
Observational evidences against

• The surface brightness is not as peaked as would
  be expected if all the cooling gas were to reach
  the center
• Mass Dropout a fraction of the gas cools out of
  the flow, at large radii, before reaching the
  center and some continues to flow inward
  most of the cooling gas never makes it to the
  center M(r) proportional to r.
• The gas is heated in someway

.
The gas must be inhomogeneous
11
Inhomogeneous model (Nulsen 86)

• Each radial zone in the cooling flow region
  comprises different plasma phases covering a wide
  range of T,r.
• The gas comprising different temperature phases
  features an inflow in which all phases move with
  the same flow speed ltvgt ltlt vs, forming a comoving
  flow
• There is no energy exchange between the different
  phases, between material at different radii, and
  no heating.

12
XMM Newton Chandra

• To further test the cooling flow picture
  most detailed X-ray spectroscopic observations
• ASTRO E ? launch accident

13
XMM Newton (2001)
Unpreceded detailed spectroscopic diagnostics of
the central regions of clusters
Spectral signatures of different temperature
phases range from the virial temperature Tvir
to a limiting temperature Tlow (Tvir/3) , which
is still above the drop out temperature where
the gas would cease to emit significant X-ray
radiation
Evidence of failure of inhomogeneous standard
cooling flow model
14
Fe L series

• The spectroscopic signatures sensitive in the
  temperature range of cooling flow are the
  emission lines from the complex of iron L series
  ions that have their
  ionization potentials in the temperature range
  near and below cluster virial temperature.

Fractional abundance of a given ion plotted
against temperature in KeV (Arnaud Raymond,
1992) . The fractional abundance is multiplied by
the abundance of that element relative to
hydrogen in the solar neighborhood.
15
Fe-L series

• The Fe-L line complex in X ray spectra as a
  function of the plasma temperature for a
  metallicity value of 0.7 solar.
• The energy change is caused by the fact that with
  decreasing temperature the degree of ionization
  of the Fe ions also decrease.

H. Bohringer et al., 2002
16
Spectra model

• Spectral prediction for an inhomogeneous flow
  based on

Peterson et al. 2003
17
Spectra model
Comparison between the model and the spectrum of
Abell 1835. Notably absent in the data are the
Fe XVII lines. The plasma appears to match the
cooling flow model between 3 KeV and the maximum
cluster temperature of 8 KeV but not below 3
KeV.
18
Spectra model
Model where the emission below 3 KeV is
suppressed.
19

• other examples

20
The cooling paradox

• Does the gas cool?
• The gas is radiatively cooling, but for some
  reason it evades detection.
• The gas is being heated in some way so that very
  little gas cools.

21
Cool cores

• What happens to the gas which should be cooling
  on very short timescales?
• Two classes of solutions have been
  proposed

The gas is prevented from cooling below a
certain temperature by some form of heating.
Different classes of mechanisms have been
considered Turbolence, shock, merging Heating
from SN Conduction Heating from the central AGN
The cooler gas is there but it is somehow
hidden
22
Properties of a successful heating model
The heating source have to

• Provide sufficient heating to balance the cooling
  flow losses (1043 1044 erg s-1)
• be fine-tuned mass deposition triggers the
  heating process and the heating process reduce
  the mass deposition
• Provide a global heating effect local energy
  deposition would result in local heating while
  the mass deposition can still go on in the less
  well-heated regions.

23
Heating from AGN

• The vast majority of cooling flow clusters
  contain powerful radio sources associated with
  central cD galaxies.
• Spectacular anti-correlation between decrements
  in the X-ray emission and extended radio
  emission.

Chandra results Holes in the X ray surface
brightness are seen to coincide with some radio
lobes ? bubbles of relativistic plasma blown by
AGN
24
Heating from AGN
Radio lobes inflated by jets of central AGN
appear to be making their way pushing aside the
X ray emitting plasma.

• The first cooling flow cluster with a central
  radio source observed by Chandra was Hydra A.

Cooling time at center 6 x 108 yr Diameter of
cavities evacuated by the radio source 25 kpc
Radio source / X ray gas interaction (Mc Namara
et al. 2000)
25
X ray / Radio interaction
Radio sources have a profound effect on the X
ray emitting ICM

• Is the energy deposition into the ICM from the
  radio sources sufficient to account for the lack
  of gas seen at very low temperatures in cooling
  flow clusters?
• TOTAL ENERGY OUTPUT OF A RADIO SOURCE
• 
  Churazov et al. 2002
• Internal energy of the bubble Work done to
  expand the bubble
• V volume bubbles
• P pressure of X ray bright shell surrounding
  the bubbles

26
X ray / Radio interaction

• Compare energy input rate with luminosity of
  cooling gas
• Hidra A
• Erad 2.7 x 1044 erg s-1 Lcool 3 x 1044
  erg s-1
• In many systems the amount of energy is
  comparable to the amount required to offset
  cooling .

27
Self regulation mechanism
The most simple physical situation would be given
if simple Bondi type of accretion from the inner
cooling core region would roughly provide the
order of magnitude of power output that is
observed and required
Spherical accretion on to the black hole

• Black hole mass 3 x 109 Msol
• Mass accretion rate 0.01 Msol yr-1
• Energy output 7 x 1043 erg s-1
• Accretion radius (vkep vs) 50 pc

28
How this energy is distributed on the right
spatial scale ?

• About 40 of the energy is transferred by the PdV
  work done on ambient medium. Since, on average,
  the bubbles expand subsonically this energy will
  be converted into sound waves and in low
  amplitude shock waves.

Ripples in the gas interpreted as due to sound
waves generated by the cyclical bubbling of the
central radio source (Fabian et al. 2003)
The gas directly bounding the bubbles seems
colder ?the energy is not deposited directly in
the boundary of the bubbles, as it would be
expected for supersonic expansion.
Dissipation of sound waves, if ICM is viscous,
may produce diffuse heating.
29
some problems

• For Hydra A and Abell 2052 the radio source is
  depositing enough energy into the ICM to offset
  the cooling gas, but
• For Abell 262 the radio source power is more than
  an order of magnitude lower of that required to
  offset the cooling luminosity!!!
• Dimension of cool cores vs accretion disk???

Current efforts are concentrated on finding
plausible heating sources to balance the cooling
flow.
Grazie
30
Bibliografia

• Fabian Cooling flows in clusters of galaxies
  ARAA 32, 277-318, 1994.
• Bohringer et al. The new emerging model for the
  structure of cooling cores in clusters of
  galaxies AA 382, 804-820, 2002
• Mathews Brighenti Hot gas in and around
  Elliptical galaxies ARAA 41, 191-239, 2003.

31
Deprojections analysis of X ray imaging

• Starting point surface brightness.
• Goal deriving a temperature T appropriate to the
  count rate per unit volume (C) produced in the
  Einstein detector at the local pressure.
• Method counts rate are accumulated in concentric
  annuli. Counts from the outer
• annulus were used as
• background. Counts
• from inner annuli are
• assumed to originate
• from spherical shells.

Fabian et al. 1981
32
Deprojections analysis of X ray imaging

• Q(E) effective area of High Resolution Image
  (HRI)
• e(T,E) dE emissivity of the gas in the band E
  E dE
• NH s(E) optical depth
• D distance to the cluster
• P outer pression

33
Estimate of mass deposition ratefrom imaging

• From the deprojection analysis ? temperature
  profile T(r)
• estimate of mass deposition rate

Lcool
T temperature at rcool
PdV work
Radiation of thermal energy
34
Estimates of mass deposition rate from X spectra

• A volume V of gas at density n cooling at
  constant pressure from T to T dT emits a
  luminosity
• m mean molecular weight of the gas.
• The luminosity of the spectrum at each frequency
  is
• en is the emissivity at frequency n.

35
Estimates of mass from X spectra

• Integreting
• where
• en emissivity of cooling gas in a single
  spectral line (Canizares et al. 1988).
• Fit of the spectrum ? mass deposition rate

BACK