GH2005Gas Dynamics in Clusters II
Craig SarazinDept. of Astronomy
University of Virginia
A85 Chandra (X-ray)Cluster Merger
Simulation
De-Projected Gas Profiles• De-project X-ray surface brightness profile
→ gas density vs. radius, (r)
• De-project X-ray spectra in annuli → T(r) • Pressure P = kT/(mp)
)()]([)(1
)(
)()(
2
22
2
2
22
2
2
2
rnrTrb
dbbIdrd
r
br
drrbI
er
b
Gas and Total Masses• Gas masses → integrate
• Total masses → hydrostatic equilibrium• Dark matter Mdm = M – Mgal - Mgas
drdP
Gr
rM
drrrrMr
gas
2
0
2
)(
)(4)(
Total Masses ProfilesXMM/Newton (Pointecouteau et
al)
(curves are NFW fits)
Gas Fraction ProfilesChandra (Allen et al)
(r2500≈0.25 rvir)
• Cluster total masses 1014 – 1015 M
• 3% stars and galaxies• 15% hot gas• 82% dark matterClusters are dominated by dark matterEarliest and strongest evidence that
Universe was dominated by dark matter
Masses of Clusters
• Mass gas ~ 5 x mass of stars & galaxies!Hot plasma is dominant form of observed
matter in clusters Most of baryonic matter in Universe today
is in hot intergalactic gas• Compare baryon fraction in clusters to
average value in Universe from Big Bang nucleosynthesisΩM = 0.3, early evidence that we live in a
low density Universe• Compare fraction baryons high vs. low
redshift, assume constantEvidence for accelerating Universe (dark
energy)
Masses of Clusters (cont.)
• Central peaks in X-ray surface brightness
Cooling Cores in Clusters
cooling core non-cooling core (Coma)
IX
Cooling Cores in Clusters• Central peaks in X-ray surface brightness• Temperature gradient, cool gas at center
Cooling Cores in Clusters• Central peaks in X-ray surface brightness• Temperature gradient, cool gas at center• Radiative cooling time tcool < Hubble time
tcool ~ 2 x 108 yr
GyrK10cm10
692/1
8
1
33
Tnt ecool
Cooling Cores in Clusters• Central peaks in X-ray surface brightness• Temperature gradient, cool gas at center• Radiative cooling time tcool < Hubble time• Always cD galaxy at center• Central galaxies generally have cool gas
(optical emission lines, HI, CO), and are radio sources
Cooling Cores in Clusters (cont.)
Theory:• X-rays we see remove thermal energy from
gas• If not disturbed, gas cools & slowly flows into
center• Gas cools from ~108 K → ~107 K at ~100
M/yr
Cooling Cores in Clusters (cont.)
Steady-state cooling of homogeneous gas
isobaric constant,
subsonic very
constant 4
entropyor energy 25
21
onconservati momentum 0
onconservati mass ,01
22
2
22
2
22
P
cv
vrM
LPv
vrdrd
r
drd
drdP
drdvv
vrdrd
r
s
Cooling Cores in Clusters (cont.)
Bremsstrahlung cooling
Reasonable fit to X-ray surfacebrightness
Ṁ ~100 M/yr
rL
rI
rr
rrT
TL
X
X
~
/1~
/1~
~5/6
5/6
2/12
ln IX
r-1
ln r
r-3
The Cooling Flow “Problem”
• Where does the cooling gas go?• Central cD galaxies in cooling flows have
cooler gas and star formation, but rates are ~1-10% of X-ray cooling rates from images
• Both XMM-Newton and Chandra spectra → lack of lines from gas below ~107 K
High-Res. Spectrum (XMM-Newton)
Peterson et al. (2001)
Brown line = data, red line = isothermal 8.2 keV model, blue line = cooling flow model,
green line = cooling flow model with a low-T cutoff of 2.7 keV
How Much Gas Cools to Low Temperature?
• Gas cools down to ~1/2-1/3 of temperature of outer gas (~2 keV)
• Amount of gas cooling to very low temperatures through X-ray emission ≲ 10% of gas cooling at higher temperature
Cooling gas now consistent with star formation rates and amount of cold
gas
Heat source to prevent most of cooling gas from continuing to low temperatures:• Heat conduction, could work well in
outer parts of cool cores if unsuppressed•Works best for hottest gas Q ∝ T7/2,
how to heat mainly cooler gas?• Supernovae?• AGN = Radio sources
Heat Source to Balance or Reheat Cooling Gas?
Radio Sources in Cooling Flows
• ≳ 70% of cooling flow clusters contain central cD galaxies with radio sources, as compared to 20% of non-cooling flow clusters
• Could heating from radio source balance cooling?
A2052 (Chandra)
Blanton et al.
Radio Contours (Burns)
Other Radio BubblesHydra A
McNamara et al.
Abell 262 Abell 133
Blanton et al. Fujita et al.Abell 2029
Clarke et al.
Abell 85
Kempner et al.
Morphology – Radio Bubbles
• Two X-ray holes surrounded by bright X-ray shells
• From deprojection, surface brightness in holes is consistent with all emission projected (holes are empty)
• Mass of shell consistent with mass expected in hole
X-ray emitting gas pushed out of holes by the radio source and compressed into shells
Buoyant “Ghost” Bubbles
Fabian et al. McNamara et al.
Perseus Abell 2597
• Holes in X-rays at larger distances from center
• No radio, except at very low frequencies (Clarke et al.)
Buoyant “Ghost” Bubbles (Cont.)
Abell 2597 – 327 MHz Radio in Green (Clarke et al.)
Ghost bubbles have low frequency radio
Buoyant “Ghost” Bubbles
Fabian et al. McNamara et al.
Perseus Abell 2597
• Holes in X-rays at larger distances from center
• No radio, except at very low frequencies (Clarke et al.)
Old radio bubbles which have risen buoyantly
Entrainment of Cool Gas
M87/Virgo Young et al.
• Columns of cool X-ray gas from cD center to radio lobe
• Gas entrained & lifted by buoyant radio lobe?
A133 --- X-ray red, Radio green Fujita et al.
Temperatures & Pressures
• Gas in shells is cool• Pressure in shells ≈ outside• No large pressure jumps
(shocks)
Temperatures & Pressures
• Gas in shells is cool• Pressure in shells ≈ outside• No large pressure jumps (shocks) Bubbles expand ≲ sound speed Pressure in radio bubbles ≈ pressure in X-ray shells• Equipartition radio pressures are
~10 times smaller than X-ray pressures in shells!?
Additional Pressure Sources
• Magnetic field larger than equipartition value?
• Lots of low-energy relativistic electrons?• Lots of relativistic ions?• Very hot, diffuse thermal gas?
– Jet kinetic energy thermalized by “friction” or shocks?
– Hard to detect hot gas in bubbles because of hot cluster gas in fore/background (but, may have been seen in MKW3s (Mazzotta et al.) In most clusters, just lower limits on kT ≳ 10 keV
Limits from Faraday Depolarization
Epol
Epol
B||
• Radio bubbles have large Faraday rotation, but strong polarization
• Faraday rotation ∝ neB∥
• External thermal gas → strong Faraday rotation and polarization
• Internal thermal gas → Faraday depolarization
• Gives upper limit on ne
• Given pressure, gives lower limit on T
kT ≳ 20 keV in most clusters if thermal gas is pressure source
Cooling
• Isobaric cooling time in shells are tcool ≈ 3 x 108 yr ≫ ages of radio sources
• Cooler gas at 104 K located in shells
Hα + [N II] contours (Baum et al.)
X-ray Shells as Radio Calorimeters
• Energy deposition into X-ray shells from radio lobes (Churazov et al.):
• E ≈ 1059 ergs in Abell 2052• ~Thermal energy in central cooling flow, ≪ total thermal energy of intracluster gas• Repetition rate of radio sources ~ 108 yr (from
buoyancy rise time of ghost cavities)
1
( -1)PV PdV
( 1)
PV
Internal bubble energy
Work to expand bubble
Can Radio Sources Offset Cooling?
• Compare– Total energy in radio bubbles, over– Repetition rate of radio source based on
buoyancy rise time of bubbles– Cooling rate due to X-ray radiation
kTμm
M
2
5Lcool
Examples• A2052: E = 1059 erg E/t = 3 x 1043 erg/s
kT = 3 keV, Ṁ = 42 M/yr
Lcool = 3 x 1043 erg/s ☑
• Hydra A: E = 8 x 1059 erg E/t = 2.7 x 1044 erg/s
kT = 3.4 keV, Ṁ = 300 M/yr
Lcool = 3 x 1044 erg/s ☑
• A262: E = 1.3 x 1057 erg E/t = 4.1 x 1041 erg/skT = 2.1 keV, Ṁ = 10 M/yr
Lcool = 5.3 x 1042 erg/s ☒ (but, much less powerful radio
source)
Blanton et al.
McNamara et al.
Blanton et al.
X-ray Ripples
How does radio source heat
X-ray gas?Perseus (Fabian et al.)
X-ray ripples = sounds waves
or weak shocks
Viscous damping heats gas?
But, is Perseus unique?
Abell 2052 (Blanton et al.)
Also has ripples, ≈ 11 kpc,
P ≈ 1.4 x 107 yr
Blanton et al.
Abell 2052 Chandra Unsharp Masked
Limit Cycle?
BHaccretes
BHaccretes
HeatX-ray
gas
HeatX-ray
gas
Radiojets
Radiojets
StopX-ray
cooling
StopX-ray
cooling
Stop BHaccretionStop BHaccretion
X-raycoolingX-ray
coolingBH
inactiveBH
inactive
Clusters from hierarchically, smaller things form first, gravity pulls them together
Cluster Formation:Mergers and Accretion
Virgo Consortium
Cluster Formation fromLarge Scale StructureLambda CDM - Virgo Consortium
z=2 z=1 z=0
• Clusters form within LSS filaments, mainly at intersections of filaments
• Clusters form throughmixture of small andlarge mergersMajor mergersAccretion
• Clusters form today andin the past
Cluster Formation (cont.)
PS merger tree: Mass vs. time
Cluster Formation (cont.)
Lambda CDM - Virgo Consortium
z=2 z=1 z=0
• Self-similar solution for spherical accretion of cold gas in E-dS Universe (Bertschinger 1985; (earlier work Sunyaev &
Zeldovich)
• Cold gas → very strongshocks
• Accretion shocks at verylarge radii (≳rvir~2 Mpc)
• No direct observationsso far
Spherical Accretion Shocks
≡ r / rta (turn around radius)
Accretion Shocks (cont.)
z=2 z=1 z=0
• Growth of clusters not spherical• Accretion episodic (mergers)• IGM not cold
Accretion Shocks (cont.)
• Growth of clusters not spherical 40x40 Mpc• Accretion episodic (mergers)• IGM not cold
40x40 Mpc
(Jones et al)
Accretion Shocks (cont.)~40x40 Mpc
External (accretion)& internal (merger)
shocks(Ryu & Kang)
Accretion Shocks (cont.)• Mach numbers ℳ ≡ vs / cs ~ 30
• Y= kinetic energy, Yth = thermal energy
Accretion Shocks (cont.)• Accretion shocks at large radii in very low density gas• X-ray emission ∝ (density)2 → very faint, never seen so far• Radio relics?• Eventually, SZ images? (SZ ∝ pressure)
Growth of LSS → most IGM is now hot, most baryons in diffuse, hot IGM
• Clusters form hierarchically
• Major cluster mergers, two subclusters, ~1015 M collide at ~ 2000 km/s
• E (merger) ~ 2 x 1064 ergs• E (shocks in gas) ~ 3 x 1063 ergs
Cluster Mergers
Major cluster mergers are most energetic events in Universe since Big Bang
Abell 85 Merger
Chandra X-ray ImageKempner et al
• Heat and compress ICM• Increase entropy of gas• Boost X-ray luminosity, temperature, SZ
effect• Mix gas• Disrupt cool cores• Produce turbulence• Provide diagnostics of merger kinematics
Thermal Effects of Mergers
• Numerical N-body for collisionless dark matter, galaxies
• Numerical hydrodynamics for gas• Initial conditions
• Draw from cosmological LSS simulations, resample at higher resolution
• Set up individual binary mergers to test physics
• Cooling by radiation• Preheating, galaxy formation
Numerical Hydrodynamics of Mergers
• Additions• Magnetic fields (MHD)• Cosmic rays, particle acceleration• Transport processes• AGNs
• Issues• Spatial resolution, particularly in cores
(AMR, SPH)• Overcooling, galaxy formation, feedback
Numerical Hydrodynamics (cont.)