Outer regions of GCs

1
Outer regions of GCs
People involved Borgani, Brighenti, Brunetti,
Colafrancesco, De Grandi, Ettori, Ghizzardi,
Mazzotta, Molendi, Moscardini, Rasia, Roncarelli,
Tozzi SRON (J. Kaastra) OBSERVATIONAL GOAL
To constrain surface brightness, temperature,
metallicity and bulk motion of the gas at Rvir.
2

• SCIENTIFIC JUSTIFICATION
• characterize the thermodynamic of the X-ray
  emitting
• plasma at the virial radius
• The above expanded in the following points
• How reliable are the total mass measurements
  through the
• hydrostatic equilibrium equation ?
• Is any residual energy deposited in kinetic
  energy of the gas ?
• How much of the gas mass fraction is measured
  within R200 ?
• Implication calibrate the masses (gas and dark
  matter)
• in local galaxy clusters to use them as
  cosmological
• probes
• 2) How is the metal enrichment process taking
  place in the cluster
• outskirts where is (primordial?) cool gas
  accumulated ?

3
SCIENTIFIC JUSTIFICATION (cont.) 3) How can we
resolve smooth/lumpy accretion shocks at Rvir
? About that, note the following items A. for
strong shock condition, one expect the surface
brightness to be gt10 times (i.e.
ngaspost/ngaspre4, Sbngas2) fainter in the
accreting region than in shocked gas. B.
non-equilibrium effects should be present in
shocked regions (see Sarazin 2001) collisionless
shocks are generally not as effective in heating
electrons as ions and, thus, equipartitions
between electrons and ions through Coulomb
collisions happens on timescale of teq few1e8
yrs. Postshock gas has a relative velocity w.r.t.
the shock front of vsound/4 and, thus, one'd
expect that Te reaches equipartition at a
distance vsound/4 teq (see section 3.2 in
Sarazin 2001).
4
SCIENTIFIC JUSTIFICATION (cont.) 4) If we are in
condition to observe the outskirts of a single
galaxy clusters, we are definitely in condition
to map cluster merger even in their early stage.
An implication of this is that it becomes
justifiable the attempt to resolve the excess in
the hard X-ray due to no-thermal emission in the
cluster regions where the interaction is taking
place and the thermal component is expected to be
cool and soft.
5

• WORK DONE SO FAR
• ( PEOPLE THAT HAVE DONE THE WORK)
• material at http//pico.bo.astro.it/ettori/outer/
  
• (1) Physical properties of ICM at Rvir
• Ettori, Molendi, Borgani
• (2) Cosmological simulations
• Borgani, Ameglio, Roncarelli, Tornatore
• (3) Spectral simulations (3.a) background
• Molendi, Ettori, Mazzotta
• (4) No-thermal component
• Brunetti, Vazza

6
(1) Physical properties of galaxy clusters at Rvir

• Which gas density, temperature and surface
  brightness values do we expect at Rvir ?
• Are simulated radial profiles consistent with
  extrapolation from observed data ?

7
Slopes given ngasra, Tgasrb, ?DMrc
We have that surface brightness r2a10.4b
0ltb/alt2/3 Mgasgt0, a3gt0 MDMr3c
Mhydrb1
Sb from CXO
?DMr-2.5(0.2) Navarro05
Tgas Markevitch98, Vikhlinin05
Krb-2a/3r1.1
fgasr0
8
(1) ICM at Rvir Simulated clusters
Tgas
ngas
Roncarelli et al
9
(1) ICM at Rvir Simulated clusters
Simulations 4 massive objects Mvir 1.9-3.4e15,
Tvir 5.4-9.9 keV Rvir sphere that encloses a
mean density of 100 ?c Mvir 4/3 ? Rvir3 100 ?c
10
(1) ICM at Rvir Simulated clusters
Simulations 4 massive objects Mvir 1.9-3.4e15,
Tvir 5.4-9.9 keV Rvir sphere that encloses a
mean density of 100 ?c Mvir 4/3 ? Rvir3 100
?c R2500, R500, R200 ? 0.2, 0.49, 0.74 Rvir
40
20
10
5
11
(1) ICM at Rvir Simulated clusters
Tgas
ngas
Unresolved CXB w. CXO (Hickox Markevitch 05)
Sb 0.5-2 keV
Roncarelli et al
12
(1) ICM at Rvir Simulated clusters
Simulations 4 massive objects Mvir 1.9-3.4e15,
Tvir 5.4-9.9 keV Rvir sphere that encloses a
mean density of 100 ?c Mvir 4/3 ? Rvir3 100
?c R2500, R500, R200 ? 0.2, 0.5, 0.74
Rvir Quant. 0.2Rvir 0.5Rvir
0.7Rvir 1.0Rvir ngas 1
0.127 (0.004) 0.051 (0.002) 0.018
(0.002) Tgas 1 0.735 (0.044)
0.613 (0.055) 0.491 (0.085)
13
(1) ICM at Rvir observed T( r)
Data points XMM (Pratt et al. 06) Red line
Chandra (Vikhlinin et al. 05) Green line
Beppo-SAX (De Grandi Molendi 02) Gray area
ASCA (Markevitch et al. 98) Declining outer
profiles, consistent with polytropic eq. of
state
Pratt, Boehringer et al. 06
14
(1) ICM at Rvir simulated T( r)
Loken???????? 03 non-radiative and radiative
runs Reasonable profiles outside the cool-core
regions.
SB????????04 cooling SF galactic winds Too
steep profiles in the central regions.
Kay????????04 cooling targetted feedback
Better agreement with data tuning required?
15
(1) Surface brightness 0.5-2 keV at Rvir
simulations
R500, z0.035, T3 keV print,moment(
1.36199d-15/sqrt(6.5), 1.52560e-15/sqrt(9.9),
8.44269e-16/sqrt(5.8), 9.74844e-16/sqrt(5.4)s
qrt(3.)/1.0354, sdsd), sd mean (disp)
6.75e-16 (1.20e-16) erg/s/cm2/arcmin2 Rvir pri
nt,moment( 4.92236d-17/sqrt(6.5),
9.58008e-17/sqrt(9.9), 3.33753e-17/sqrt(5.8),
5.52125e-17/sqrt(5.4) sqrt(3.)/1.0354, sdsd),
sd mean (disp) 3.30e-17 (1.06e-17)
erg/s/cm2/arcmin2
16
(1) Surface brightness 0.5-2 keV at Rvir
observations
Use the values in Tab.2 and 4 in Mohr et al. ApJ
517 627 Assuming Rvir 2R500 as measured in
numerical simulations a119, Tx5.8 keV,
z0.0442 bb0.662 print,1.18d-13(1(21.635,1
.826/0.483)2)(-3.bb0.5) 3.8851550e-16
2.8154025e-16 bb0.8 print,1.18d-13(1(21.635
,1.826/0.483)2)(-3.bb0.5) 7.9029728e-17
5.2354670e-17 _at_ R500, z0.035, T3 keV
print,2.7797225e-15,6.6278778e-16/sqrt(5.8)sqr
t(3.)1.04424/1.0354 2.07120e-15 4.93849e-16
erg/s/cm2/arcmin2 _at_ Rvir, z0.035, T3 keV
print,3.8851550e-16,5.2354670e-17/sqrt(5.8)sqr
t(3.)1.04424/1.0354 2.89486e-16 3.90099e-17
erg/s/cm2/arcmin2
17
(1) Surface brightness at Rvir observations vs
simulations
Rescaling everything to T3 kev and z0.035, we
find that --gt R500 0.5 Rvir simula mean
(disp) 6.75e-16 (1.20e-16) erg/s/cm2/arcmin2 2
objs in Mohr etal 4.94e-16 - 2.07e-15,
3.58e-16 - 2.65e-15 erg/s/cm2/arcmin2
Osservations / Simulations 0.5/4 Note that the
extrapolation from Mohrs data use (i) R500 from
both R-T relation and direct estimate from
beta-model, (ii) 2 beta values, the observed one
and 0.8, more reliable for the steepening of the
surface brightness in the outskirts. --gt
Rvir simula mean (disp) 3.30e-17 (1.06e-17)
erg/s/cm2/arcmin2 2 objs in Mohr etal
3.90e-17 - 2.89e-16, 2.80e-17 - 4.62e-16
erg/s/cm2/arcmin2 Osservations / Simulations
1-14 We consider in the spectral simulations an
object at T3keV and z0.035 with brightness
8e-17 erg/s/cm2/arcmin2
18
(2) Cosmological simulations

• The "simulators" are asked to try to characterize
  the shocked
• regions, providing answers to questions like
• Accretion shocks. Can you quantify the typical
  change in surface
• brightness, gas density, temperature both in
  regions with well-defined lumpy
• accretion (probably where filaments are present)
  and in regions more smooth
• but nearby the virial radius ?
• How can we characterize lumpy and smooth
  accretion ?
• How much does a typical shock extend in angular
  scale ?
• How/what do you resolve of the accretion shocks
  in isolated and merging
• simulated clusters ?
• The condition of strong shock (compression factor
  of 4) are difficult to realize.
• Do (cosmological) simulations confirm lower Mach
  number (for a relation
• between Mach number and compression factor, see
  eq.27 in Sarazin) ?

19

B. Merging. Given a merging between objects with
(let's say) equal masses, how the surface
brightness and temperature evolve in the region
of the interaction e.g Ricker Sarazin 2001 ?
C. Metal enrichment. How the shocked regions
can be characterized by the observed metal
abundance ? How much "primordial" is the
accreting gas ? Are the alpha-elements like
silicon and sulfur relatively more abundant that
Iron in those regions due to a larger number of
SN CC than Ia ?
20
(2) Cosmological simulations

• The simulations that are available for analysis
  are the following
• 4 clusters with Mvir(1-2.3)x1015 Msun/h (ICS
  from K. Dolag LCDM cosmology with ?80.9)
• Code GADGET-2 (Springel 2005)
• Physics Z-dep. Cooling, SF, feedback, chemical
  enrichment from SNIa, SNII and AGB (Tornatore et
  al. 04, 06)
• Maps of Compton-y, gas density, SBX (0.5-2
  keV), Tmw, Tew, Tsl, Si, O, Fe (S. Ameglio)

21
?gas
?X
Y
4 Rvir
Tew
Tsl
Tmw
O
Si
Fe
22
Following the chemical enrichment of the ICM
Data De Grandi et al. 03
? Profiles of ZFe steeper than observed! ? Lack
of metal diffusion (by turbulence?) ? Higher
resolution to better treat the stripping of
metal-enriched gas in galactic halos?
Salpeter IMF
23
(2) Cosmological simulations
Available products - Maps of the described
quantities - Fits files (or idl procedure to
generate them) for all the maps, along three
projection directions Upon request -
Metallicity with other weights all
mass-weighted - Spectral maps (E. Rasia) - Maps
of velocity and velocity dispersion - Maps of the
Mach number (see Brunetti)
24
(2) Cosmological simulations

• Possible items to investigate at R200
• - Temperature and entropy structure
• Metals (0.1 Zsun or higher ?), their
  distribution and correlation with the above
  point.
• Line broadening and centroid shifts correction
  of hydrostatic masses?

25
(3) Spectral simulations with Micro-calorimeter
and WFXRT

• Files with the modeled background (local CXB
• instrumental) in XSPEC and responses of the
  WFXRT
• Micro-calorimeter were prepared.
• We use a bash script (simback.sh) that calls
  XSPEC11 and
• does
• Faked spectra with random flactutations with
  assigned
• amplitude in bkg
• b. Faked spectra of given cluster model bkg
• c. The joint-fit src bkg files to constrain T
  and EI.
• Simulations with Micro-calorimeter already
  circulated
• (molendi_micro.ps Molendis talk).

26
(3) Spectral simulations RESULTS with
Micro-calorimeter
(See Molendis talk)

• Up to 0.6 Rvir, velocity measurements in the
  order of 100-200 km/s can be performed surface
  brightness can be measured if systematics error
  on fluxes can be kept lt few per cent. EW of O
  VIII can be derived. Abundances can be measured
  if T is known. Estimates of T from Fe-L shell
  forest only for cool systems.
• Constraints on the physical properties at Rvir
  cannot be reached.

27
(3) Spectral simulations RESULTS with
Micro-calorimeter
To reach Rvir it is required to improve
significantly the instrumental properties, either
in spatial or spectral resolution. Reduction of
extragalactic bkg by a factor of 3 by resolving
70 of CXB with 5 PSF allows to resolve line at
Rvir. Otherwise, improvement by a factor of 5 in
spectral resolution, i.e. FWHM 0.4 eV. Expected
line broadening does not permit further
improvements. Improvements in effective area are
not critical.
28
(3) Spectral simulations RESULTS with
Micro-calorimeter
NOTE all these results are based on realistic
assumptions but must be taken with some caution.
To make line measurements the major obstacle is
the fore/back-ground emission. Our simulations
consider average properties of the galactic
foreground, about which (1) very little is known
on the spectrum (see McCammon et al. 2003), (2)
these estimates refer to 1 sr and we know from
RASS that strong spatial variations are expected.
29
(3) Spectral simulations RESULTS with WFXRT
At Rvir, for typical exposure time of 1e6 sec, we
can put constraints with relative error 1? of
30 on T and of 15 on EI by a joint-fit of the
source bkg spectra. To do this, we require to
take under control the fluctuations on the
background component at 1 level. With spatial
fluctuations of 5, the relative error on T and
EI measurements rise to gt45 and 30,
respectively. No much improvement is obtained by
increasing the effective area up to 6 keV.
30
(3) Spectral simulations RESULTS with WFXRT
The greatest concern is about the capacity to
control the modelling of the background
components affected from spatial variations up to
the requested level of 1. The joint-fit
extended to background spectra observed with
microcalorimeter can help in the modelling.
31
(3.a) Background The components

1. Instrumental
2. Cosmic

• Extragalactic
• Local component

32
Relevance of Bkg

local gal
CXB
cluster
instr
Background components are dominant
33
Ratio of source to total spectrum

34
Bkg Subtraction

Measure Total spectrum (sou bkg) Bkg
Spectrum Comparison of spectra
(subtraction/modeling)
35
(3) Spectral simulations Subtracting the
background

• Input values
• exposure time 3e5, 5e5, 1e6 sec
• CXB 70 resolved, ? photon_idx 0.1, 0.02
• with/without instrumental bkg

• ARF exponential cut at 5 6 keV

36
(3) Spectral simulations substracting the bkg
Dashed line input values dotted lines 30
unc. at 90 c.l.
37
(3) Spectral simulations RESULTS with WFXRT
Joint-fit src bkg spectra INPUT T 3 keV,
Sb8e-17, texp1e6, ARFpn_off_axis_exp.arf Resul
ts after 100 MC simulations (16-50-84 percentile
relative err ) err0
T -35 46 K -9 11
err1 T -28 72 K -13
16 err1, arfexp6 T -26 47
K -14 10 err1, texpbkg10x T
-28 40 K -13 10 err5
T -43 gt100 K -26 37
38
(3) Spectral simulations RESULTS with
WFXRTcalorimeter
Joint-fit src bkg spectra Unresolved CXB 0.3
wfxrt 0.8 calorim
Results after 50 MC sim err1 T
(-25, 39) K ( -7, 12) err5
T (-67, gt100) K (-67, 24)
39
(4) No-thermal component
See Brunettis talk
On the mergers, we are interested to have some
expectations on the hard X-ray emission
(brightness and spectral shape) in the different
stage of the mergers. Remember If we are in
condition to observe the outskirts of a single
galaxy clusters, we are definitely in condition
to map cluster merger even in their early
stage. An implication of this is that it becomes
justifiable the attempt to resolve the excess in
the hard X-ray due to no-thermal emission in the
cluster regions where the interaction is taking
place and the thermal component is expected to
be cool and soft.
40
Conclusions

• We know what wed observe at Rvir (Tgas, Sb)
• WFXRT At Rvir, for typical exposure time of
  1e6 sec, we can put constraints with relative
  error 1? of 30 on T and of 15 on EI by a
  joint-fit of the source bkg spectra. We require
  to take under control the fluctuations on the
  background component at 1 level.
• micro-cal Up to 0.6 Rvir, velocity
  measurements in the order of 100-200 km/s can be
  performed surface brightness can be measured if
  systematics error on fluxes can be kept lt few per
  cent. EW of O VIII can be derived. Abundances can
  be measured if T is known. Estimates of T from
  Fe-L shell forest only for cool systems.
• Constraints on the physical properties at Rvir
  cannot be reached.

41
Conclusions

• NOTE all these results are based on realistic
  assumptions but must be taken with some caution.
• The major obstacle for spectral measuraments is
  the subtraction of the fore/back-ground emission.
  Our simulations consider average properties of
  the galactic foreground CXB. In particular, the
  former is expected to vary spatially in a
  significant way.
• We need to control the modelling of the
  background components affected from spatial
  variations up to the requested level of 1.
• Which are the best targets ?
• Low nH, high T, possibly with QSOs in the bkg,

42
(3) Spectral simulations degeneracies
Z vs T
Case 8e-16 cgs, 1e6 s, with Inst Bkg
norm vs T
43
(3) Spectral simulations degeneracies
Phot_idx CXB vs T
Norm CXB vs T
Norm local thermal vs T
44
(3) Spectral simulations
local gal
CXB
cluster
instr
45
(3) Spectral simulations

• ARF exponential cut at 5 6 keV