The Quest for Cluster Simulations
Klaus Dolag
Max-Planck-Institut für Astrophysik
Intro
02/06/2005 – p.1
Outline / Motivation
• Do we model the ICM accurate enough to do cosmology
with galaxy clusters ?
⇒ State of the ICM vs. Cluster properties
see poster by E. Rasia on ”x-ray Mass”
⇒ ICM properties vs. model assumptions (e.g. physical
processes included)
• Many physical processes are linked together (e.g. thermal
conduction, turbulence, magnetic fields)
see poster by D. Sijacki on ”Central AGN feedback”
⇒ Can we model their individual effects ?
⇒ Can we overcome numerical issues and start to study such
processes (e.g. SPH,Gadget-II-XXL)
Outline:
• General properties of simulated galaxy clusters
• ”Turbulence” in SPH simulations
• Magnetic Fields in galaxy clusters
02/06/2005 – p.2
Hutt (High resolution Cluster set)
39 Haloes (> 0.7 × 1014 Msol ), up to 4 × 106 Particles in Rvir !
• DM-only (dm)
• none radiative gas (gas)
• cooling+starformation+winds (csf)
• no/week/strong winds (csfnw,csf,csfsw)
• thermal conduction (csfc)
• new scheme to avoid damping of turbulence (gas nv)
• numerical tests (e.g. resolution, grid vs. glass, etc.)
• Metals and chemical enrichment
⇒ poster by Luca Tornatore on ”chemical enrichment”
02/06/2005 – p.3
The revenge of the ICM
4.5
subhalo position (DM sim.)
subhalo position (GAS sim.)
main halo position
4.0
h−1 Mpc
3.5
3.0
0.02
2.5
0.018
DM simulation
GAS simulation
CSFC simulation
0.016
2.0
sigma7.5 [(Mpc/h)2]
0.014
0.012
1.5
0.01
2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 3.3 3.4 3.5
0.008
−
0.006
0.004
0.002
0
0.1
0.2
0.3
0.4
0.5
redshift z
0.6
0.7
0.8
The presence of gas changes dynamics and profiles !
• Strong lensing cross section decreases for none radiative
gas with strong turbulence (e.g. none thermal pressure)
• Increases strongly for cooling and starformation.
Puchwein, Bartelmann, Dolag & Meneghetti 2005, A&A in press, astro-ph/0504206
02/06/2005 – p.4
State of the ICM
Emission is complex, mixture of dynamic & physical processes !
Shape crucial for interpretation of global quantities !
Dolag et al. (in prep.)
02/06/2005 – p.5
State of the ICM
g1.csf.a.z
g1.csf.b.z
g1.csf.c.z
1000
1000
1000
800
800
800
600
600
600
400
400
400
200
200
200
200
2
400
4
600
6
T [keV]
800
8
1000
10
200
12
400
1
600
2
g1.csfc.a.z
800
3
1000
200
4
1
T [keV]
2
T [keV]
g1.csfc.c.z
1000
1000
800
800
800
600
600
600
400
400
400
200
200
200
2
400
4
600
6
T [keV]
800
8
1000
10
12
200
1
600
g1.csfc.b.z
1000
200
400
400
600
2
800
3
1000
4
T [keV]
200
1
400
1000
3
600
2
T [keV]
800
800
1000
3
First cosmological simulations including thermal conduction.
⇒ but no solution for the catastrophic cooling !
⇒ flat temperature profile for T > (8-10)keV !
Jubelgas, Springel & Dolag 2004, MNRAS, 351, 423;
Dolag, Jubelgas, Springel, Borgani & Rasia 2004, ApJ 606L, 97
02/06/2005 – p.5
State of the ICM
Deviation from gaussianity contain information about dynamics !
02/06/2005 – p.5
State of the ICM
Mass weighted temperature vs. with of the gaussian fit !
⇒ Correlated, but physics, e.g. thermal conduction !
02/06/2005 – p.5
Turbulence in the ICM
Old viscosity scheme
New viscosity scheme
Artificial viscosity completely switched of outside of shocks !
• Instabilities less damped (e.g. Kelvin-Helmholtz).
⇒ Inset of turbulence
⇒ Enlarged energy-fraction in gas velocity
Dolag, Vazza, Brunetti, Tormen & Springel (in prep.)
02/06/2005 – p.6
Turbulence in the ICM
lvisc
svisc
ovisc
Turbulence can leave to significant pressure support !
02/06/2005 – p.6
Turbulence in the ICM
Unsharpened masked: image - smoothed(image,200kpc)
2Mpc x 2Mpc x-ray emission of g1 comparing the two viscosity
schemes.
02/06/2005 – p.6
Turbulence in the ICM
Unsharpened masked: image - smoothed(image,200kpc)
2Mpc x 2Mpc pressure map (e.g. SZ) of g1 comparing the two
viscosity schemes.
02/06/2005 – p.6
Turbulence in the ICM
Due to large contribution of bulk motions and beam smearing,
the imprint of “true“ turbulence will be hard to detect, even
02/06/2005 – p.6
resolution like Astro-E2 !
Turbulence in the ICM
Turbulent energy content in galaxy clusters as function of mass.
Vazza, Tormen, Brunetti & Dolag (in prep.)
02/06/2005 – p.6
Coruscant
Constrained Local Universe including Magnetic Fields
Coma
Coma
Virgo
Virgo
Hydra
Centaurus
Perseus
Hydra
Centaurus
A3627
Perseus
A3627
104
log(Y)
−9.5
−8.5
−7.5
−6.5
Cl l (l+1) / 2π (µK) 2
−10.5
asinh(w)
102
−5.5
Coma
−4.5
−5e−7
5e−7
Coma
100
Virgo
Virgo
10-2
Hydra
Centaurus
Perseus
−9.5
−8.5
−7.5
100
Multipole l
1000
Centaurus
Perseus
A3627
+ Hubble volume
log(Y)
−10.5
10
−6.5
−5.5
−4.5
Hydra
A3627
asinh(w)
−5e−7
5e−7
2 × 50.000.000 particles, mgas = 4.8 × 108 MSol /h
Mathis et al 2002 (DM-Only), Dolag et al 2004 (Gas + MHD)
SZ Maps: Dolag et al. 2005, submitted, astro-ph/0505258
02/06/2005 – p.7
Coruscant with MHD
Saturation
Shear + Turbulence
+ Major Merger
Shear
Magnetic fields powered by compression and anisotropic collaps
(see also Bruni et al. 2003), sheer flows (see also Birk et al.
1999) and merger events (see also Roettiger et al. 1999).
Full ideal MHD (Phillips & Monaghan 1985, Dolag et al. 1999,2002), Brove et al.
02/06/2005 – p.8
2001/2004), Price & Monaghan 2004) assuming a seed field at ”high” z.
Coruscant with MHD
3
5
Magnetic Field - Temperature relation
7
9
02/06/2005 – p.8
Coruscant with MHD
RM ∝
Z
ne Bk dx
RM map of central radio galaxy in A400 (2.3keV) on the left
side and 3C449 in a 1.2keV cluster on the right.
~ with T , ρ, r ... ?
02/06/2005 – p.8
⇒ scaling of RM (e.g. B)
Coruscant with MHD
+ 8 rad/m
Comparison of radial RM profile.
2
02/06/2005 – p.8
Magnetic power spectrum
Coma
Slope of the (3D) magnetic field power spectra (k2 B(k)2 ) !
02/06/2005 – p.9
Magnetic power spectrum
Example A400, Slope observed to be high !
⇒ Signature of Merger or Turbulence ?
02/06/2005 – p.9
Conclusions
• Many physical processes need to be included and
understood in detail to better predict the state of the ICM.
This includes turbulence, conduction, viscosity, magnetic
field, relativistic component(s), feedback and more.
• Thermal conduction even with κ = 1/3 does not suppress
the catastrophic cooling, does not change global quantities
(e.g. mass weighted temperature) but can give rise to
dramatic changes in the overall thermal structure.
• Cluster assembly can provide large amount of turbulence
which could strongly affect cluster properties.
• Cluster formation leaves its imprint in magnetic field.
• High precision cosmology with galaxy clusters is still ´´Far
Far Away´´ !
02/06/2005 – p.10
Outlook
z = 1.7
Ultra high resolution (12keV) cluster (Luke)
8
m gas = 2.8 x 10 Msol/h !
~ 10 Mpc/h
g8_Fe.a.z
50 Million gas particles !
3 Mpc
1400
1200
~ 160 Mpc/h
1000
~ 25 Million particles will
end inside cluster halo !
800
Runs using ’metal extension’
− follows C,N,O,Mg,Si,Fe
Fe distribution
− metal cooling
− following SN I and SN II
− distribute metals into ICM
− explicit follows stellar population (no IRA !)
− includes mass transfer back into ICM by stellar winds
− allows various IMFs, even time evolving
600
600
0
800
0.0001
1000
0.0002
(counts)
1200
1400
0.0003
0.0004
Stars (= Galaxies) in the Local Universe
Following evolution of stelar population and chem. enrichment.
(Gadget2 extension by Tornatore et al. 2004/2005)
02/06/2005 – p.11