Transcript Titolo - INAF - OA

SEARCHING FOR
COOLING FLOWS…
Silvia Caffi
IASF/CNR Sez. Milano
IMAGING:
central regions feature ~ constant
surface brightness
in outer regions S.B. falls off as a
power-law with index ~ 3
emission is traced out to 1-2 Mpc
from the core
No absorption
X-RAY SPECTRA:
Exponential
cut-off
emission from optically thin thermal
plasma polluted by heavy elements
typical values: ne ~ 10-4 – 10-2 cm-3,
Tg ~ 107 – 108 K (heavily ionized gas),
R ~ 1 Mpc.
Properties of the Intra Cluster
Medium:
hot, heavely ionized and tenouos gas at rest in the
potential well of the cluster
1/ 2
p
1/ 2
1
1/ 2
v sound   10 
 Tg n p
  Tg 
tcool  8.5 x
10 yr 
  8 

3
3
 10
cm

 10 K 
1 / 2
considered as a fluid in hydrostatic
equilibrium
T




D
g
8
   8 
tsound  6.6x 10 yr 
 Mpc   10 K 
tsound  tcool  tHubble
dissipates energy at a very slow rate by X rays
tcool ~ Tg1/2 np-1
for large radii np is small
in the core np is large
tcool << tHubble
tcool ~ tHubble
COOLING FLOW CLUSTERS
surface brightness strongly
peaked at the center
low ionization lines in soft
X-ray spectra
temperature gradients
toward the center
DeCanizares
Grandi & et
Molendi
al. (1984)
(2002)
In-homogenous model for CF (Nulsen 1984)
multiphaseness of the gas: different phases (T and ρ)
coexist, in pressure equilibrium (tsound < tcool) at every r
the phases comove, under the pressure of the gas
immediately on top, with <v> << vsound.
at T~106 K tcool ~ tsound : the cold blob decouples from
the flow while the others continue to flow inward.
the mass deposition rate scales as rα, implyng that
deposition occurs everywhere in the cooling flow region.
Typical value for mass dep. rate: dM/dt=100Msun/yr
cooling flows are seen in a large number of clusters
(~ 60% - 70%), so they can be resonably considered
as persistent phenomena
dM / dr
 100Msun / yr

Macc 1012 Msun 
 dM 
  tHubble
 dt 
Macc  



Macc Mcluster 1014  1015 Msun
Macc  McD 1012 Msun
IONIZED
Cold gas
NEUTRAL
lines observed in optical and UV indicate
that ionized gas is present but << Macc
21 cm observations in central galaxies
give MHI < 109 Msun
MOLECULAR recent observations (Edge 2002) have
detected molecular gas for the first
time, again << Macc
XMM point of view
In RGS spectra there is a remarkable
lack of emission lines expected from
gas cooling below 1-2 keV (see for
example the sample of 14 objects in
Petersen 2003 ... )
EPIC cannot resolve individual lines
but can discriminate bet. models with
and without a minimum temperature:
Tmin=0.9 keV: the shoulder is absent
because the low ionization lines are
missing.
Tmin=0.1 keV: we see a shoulder down
to ~ 0.8 keV, due to low ionization lines
from gas colder than 0.9 keV.
both EPIC and RGS
mesurements indicate BUT
Tmin in [1,3] keV
standard cooling flow
model predicts gas with T
down to at least 0.1 keV!
gas is NOT multiphase, at least not in the sense
required by the standard multi-phase CF model
FAILURE OF THE STANDARD
CF MODEL
If CFs are not observed something must
quench them: HEATING MECHANISM
 feedback from AGN: Chandra
observation clearly show interaction
between AGN and the ICM (radio
lobes/ X-ray cavities)
efficient mechanism only in cluster cores:
too high MBH are required to completely
quench a luminous CF (Fabian 2002)
MIXED MODELS
 thermal conduction: large heat
reservoir in the outer regions of
clusters, if ΔT/T is large this
mechanism should be efficient
Perseus (Fabian et al 2002)
estimates of κeff(r) show that this
mechanism is efficient only in outermost
regions: for innermost ones κeff exceeds
κs (Ghizzardi et al. 2003)
First candidate: A2199
nearby cluster z = 0.030, D = 131.6 Mpc (H0 = 70 km/s/Mpc)
strong X-ray emission peaked on NGC6166: LX = 6.43 x 1044 erg/s
kT ~ 4.7 keV (Peres 1998), <Z> = 0.35 (De Grandi & Molendi 2001)
why A2199 is a good subject for our quest?
extremely relaxed cluster, no evidence for azimuthal gradients in
our kT and metallicity maps (XMM data)
lack of evidences for interaction bet. cD galaxy and ICM in
Chandra images (sharper eyes than XMM-Newton)
very luminous in X band ( some 1044 erg/s)
IT LOOKS LIKE A CLUSTER HARBOURING
A STRONG COOLING FLOW
spectral models applied to A2199:
single temperature (wabs*mekal in XSPEC): 4 free parameters kT,
redshift, abundance and normalization of the thermal component.
two temperature (wabs*(mekal+mekal) in XSPEC): two additional free
parameters are T and normalization of the second component.
comparison bet. 1T and 2T models:
1T has good residuals, suggesting no need for other components
we placed upper limits to the contribution of an eventual cool
component: in the innermost region temperatures below 1.0 keV are
ruled out, in the second bin (0.5-1.0 arcmin) the value of Tmin rises
up to 1.2 keV!
Let’s make the point on the first candidate!
luminous
very relaxed
a promising cluster ...
T and ZFe gradients
BUT
the ICM is well described by a 1T model. A second
component, if present, cannot be cooler than ~ 1
keV (from spectral analysis)
SO
NO CF IN A2199!!!
Second candidate: A1068
more distant than A2199 z = 0.1386, D = 645 Mpc (h0 = 0.7)
X-ray emission peaked on cluster center
kT ~ 4.0 keV, <Z> ~ 0.5 (Wise et al. 2004)
elliptical shape (ε = 0.71), complicated central morphology (r<50 kpc)
why A1068 is a good subject for our quest?
mesures of CO emission lines tell us that a large amount of molecular
gas is present in this cluster Mgas = 8.5x1010 Msun (Edge 2001), which
could be the gas cooled out from the flow during the CF
IT LOOKS LIKE A CLUSTER HARBOURING
A STRONG COOLING FLOW
Modelling the gas in A1068
single temperature (wabs*mekal in XSPEC)
two temperature (wabs*(mekal+mekal) in XSPEC)
according to the analysis in Wise et al. (2004) we tried also with
(old) CF model (wabs*(mekal+mkcflow) in XSPEC letting the Tmin be a
free parameter together with mass dep. rate.
comparison bet. 1T and other models:
as in the case of A2199 the single temperature model has good
residuals, suggesting no need for other components
an eventual cool component in a 2T model should be no cooler than
0.8 keV in the innermost bin.
fits with CF model are very instable, however a fit on an integrated
spectrum (no annular bins!) suggest a Tmin of about 1.5 keV
Let’s make the point on the second
candidate!
luminous
large amount of
molecular gas
(Edge 2001)
again ... a good
candidate ...
BUT
the ICM is well described by a 1T model and
another thermal component, if present, cannot be
cooler than ~ 0.8 keV. Moreover the Tmin of an
eventual CF is no less than 1.5 keV!
SO
NO CF IN A1068!!!
We examined two objects that seemed perfect
candidates to host a CF.
A2199 for the
“relaxed aspect” and
its high luminosity
A1068 for the large
amount of molecular
gas found by Edge
NEITHER IN A2199 NOR IN A1068 WE
HAVE FOUND EVIDENCES FOR AN
ONGOING COOLING FLOW !!!