Transcript feedback - University of Massachusetts Amherst

Stellar Feedback and Evolution
of Milky Way-like Galaxies
Q. Daniel Wang
IRAC 8 micro
K-band
ACIS diffuse 0.5-2 keV
University of Massachusetts, Amherst
Galaxy formation and evolution context
Toft et al. (2002)
The “overcooling” problem:
Too much condensation
to be consistent with
observations (e.g., White &
Rees 1978; Navarro & Steinmetz
1997)
The “missing” baryon problem
• Stars and the ISM accounts for 1/31/2 of the baryon expected from the
gravitational mass of a galaxy.
• Where is the remaining baryon
matter?
– In a hot gaseous galactic halo?
– Or having been pushed away?
• Both are related to the galactic energy
feedback!
Origins of the galactic feedback
• AGNs (jets)
• Nuclear starbursts:
– Radiation pressure
– Superwinds
• Gradual energy inputs
– Galactic disks: massive star formation
– Galactic bulges: Type Ia SNe.
Outline
• X-ray absorption line
spectroscopy of hot gas
in and around the MW.
• X-ray imaging of nearby
“normal” galaxies
• Hydro-simulations of
stellar feedback and
galaxy evolution.
Pre-Chandra View of the hot gas
ROSAT
X-ray¾-keV
binary Diffuse Background Map:
~50% of the background is thermal and local (z < 0.01)
The rest is mostly from faint AGNs (McCammon et al. 2002)
X-ray absorption line spectroscopy:
adding depth into the map
4U1957+11
X-ray binary
AGN
X-ray binary
Wang et al. 05, Yao & Wang 05/06,
Yao et al. 06/07
ROSAT all-sky survey
in the ¾-keV band
X-ray absorption line spectroscopy
is powerful !
• Tracing all K transitions
of metals  all three
phases of the ISM.
• Not affected by photoelectric absorption
unbiased measurements
of the global ISM.
LETG+HETG spectrum of LMXB X1820-303
Yao & Wang 2006, Yao et al. 2006, Futamoto et al 2004
Fe XVII K
Spectroscopic diagnostics
OVII
OVIII
Ne IX
Ne VIII
OVI
Ne X
Assuming solar abundances and CIE
One line (e.g., OVII K) 
velocity centroid and EW
 constraints on the
column density, depending
on assumed b and T
Two lines of different
ionization states (OVII
and OVIII K)  T
Two lines of the same state
(K and K)  b
Lines from different species
 abundance fa
Yao & Wang 2005
Mrk 421
(Yao & Wang 2006)
•Joint-fit with the absorption lines
with the OVII and OVIII line
emission (McCammon et al. 2002)
•Model: n=n0e-z/hn; T=T0e-z/hT
 n=n0(T/T0), =hT/hn, L=hn/sin b
OVI 1032 A
LMC X-3 as a distance marker
• BH X-ray binary,
typically in a
high/soft state
• 50 kpc away
• Vs = +310 km/s
• Away from the
LMC main body
Wang, Yao, Tripp, et al.
2005
H image
LMC X-3: absorption lines
OVII
Ne IX
The EWs are about the same
as those seen in AGN spectra!
Galactic global hot gas properties
• Structure:
– A thick Galactic disk with a scale height 1-2 kpc,
~ the values of OVI absorbers and free electrons
– Enhanced hot gas around the Galactic bulge
– No evidence for a large-scale (r ~ 102 kpc) X-rayemitting/absorbing halo with an upper limit of NH~1
x1019 cm-2
• Thermal property:
– mean T ~ 106.3 K toward the inner region
–
~ 106.1 K at solar neighborhood
• Velocity dispersion from ~200 km/s to 80 km/s
• Abundance ratios consistent with solar
But a large-scale hot gaseous halo is
required to explain HVCs!
Feedback from disk-wide star
formation
Diffuse X-ray emission
compared with
HST/ACS images:
Red – H
Green – Optical R-band
Blue – 0.3-1.5 keV
• Scale height ~ 2 kpc +
more distant blubs.
• Lx(diffuse) ~ 4x1039 erg/s
• T1 ~ 106.3 K, T2 > 107.1 K
Li, J. et al. (2008)
NGC 5775
M83
Soria & Wu (2002)
Much work is yet to be done
to determine the differential
properties of the hot gas.
M31
Li & Wang 2007
0.5-1 keV
1-2 keV
2-8 keV
IRAC 8 micro
K-band
0.5-2 keV
Lx~2x1038 erg/s, only
~1% of the Type Ia SN
energy input
Mass-loading from the nuclear spiral?
• SMBH is not active
• No SF in the bulge
• Photo-ionization
probably due to
PAGB stars
• Correlation of
diffuse X-ray with
the H emission
 evaporation via
thermal conduction?
H image with 0.5-2 keV contours
NGC 4594 (Sa)
•
Average T ~ 6 x 106 K
•
Lx ~ 4 x 1039 erg/s, ~ 2% of Type Ia SN energy
• Not much cool gas to hide/convert the SN energy
• Mass and metals are also missing!
– Mass input rate of evolved stars
~ 1.3 Msun/yr
– Each Type Ia SN  0.7 Msun Fe
Li et al. 2007
Summary for hot gas seen in
nearby galaxies
• At least two components of diffuse hot
gas:
– Disk – driven by massive star formation
– Bulge – heated primarily by Type-Ia SNe
• Characteristic extent and temperature
similar to the Galactic values, although a
possible very hot component (T > 107 K) is
poorly constrained.
Observations vs. simulations
Little evidence for X-ray
emission or absorption
from IGM accretion.
No “overcooling”
problem?
Galaxy
Vc
NGC 4565 250
NGC 2613 304
NGC 5746 307
NGC 2841 317
NGC 4594 370
Simulations by Toft et al. (2003)
The missing energy and large LX/LK
dispersion problems of low-mass ellipticals
David et al (2006)
SNe
AGN
• Observed LX is < 10%
of the energy inputs.
• Galactic wind model
fails:
– Predict too small LX
with little dispersion
– Too high T, fixed by
the specific energy
input
– Too steep radial
surface intensity
profile.
Problems with the existing models
• Model of the stellar feedback
– Typically with no consideration of the evolving
galactic environment
– Fixed outer boundary  unstable subsonic solution
– No additional mass-loading
• Model of galaxy evolution via IGM accretion
– Typically with no stellar feedback (e.g., BirnBoim,
Dekel, & Neistein 07)
– Predicted massive cooling hot gaseous halos are
not observed.
– But a large-scale, low-density hot gaseous halo is
required to explain HVCs around the MW.
1-D Simulations of galaxy formation
with the stellar feedback
• Evolution of both dark and baryon matters
(with the final mass 1012 Msun)
• Initial bulge formation (5x1010 Msun) 
starburst  shock-heating and expanding
of gas
• Later Type Ia SNe  bulge wind/outflow,
maintaining a low-density high-T halo,
preventing a cooling flow
Tang, Wang, Lu, & Mo 2008
1-D Simulations of galaxy formation
with the stellar feedback
z=1.4
z=0.5
z=0
• A blastwave, initiated by
the SB, is maintained by
the Type Ia SN
feedback.
• The IGM is heated
beyond the virial radius.
• The accretion can be
stopped.
• The bulge wind can be
shocked at a large
radius.
Dependence on the specific energy
of the feedback
z=1.4
z=0.5
z=0
• If the mass-loading is
important, the wind may
then have evolved into a
subsonic outflow.
• This outflow can be stable
and long-lasting  higher
Lx, lower T, and more
extended profile.
• Consistent with
observations!
Galaxies such as the MW evolve in hot
bubbles of baryon deficit!
• Explains the lack
of large-scale Xray halos.
• Bulge wind drives
away the present
stellar feedback.
Total baryon
before the SB
Cosmological
baryon fraction
Total baryon
at present
Hot gas
2-D simulations of the feedback in M31
• Qualitatively
consistent with the 1-D
results
• Instabilities at the
contact discontinuities
 formation of HVCs?
An ellipsoid bulge (q=0.6),
a disk, and an NFW halo
Tang & Wang (2008)
3-D simulations of a galactic
bulge wind
• Adaptive mesh
refinement, down
to 6 pc
• Stellar mass
injection and
sporadic SNe,
following the
stellar light.
10x10x10 kpc3 box
density distribution
• Emission primarily from
shells and filaments.
• Fe-rich ejecta dominate
the high-T emission
• Not well-mixed with the
ambient medium
3-D Effects
1-D
• Large dispersion 
Low
Res.
Log(T(K))
1-D
– enhanced emission at both
low and high temperatures
– Overall luminosity increase
by a factor of ~ 3.
– Low metallicity if modeled
with a 1-T plasma.
• Consistent with the 1-D
radial density and
temperature
distributions, except for
the center region.
Conclusions
• Diffuse hot gas is strongly concentrated toward
galactic disks/bulges (< 20 kpc) due to the feedback.
• But the bulk of the feedback is not detected and is
probably propagated into very hot (~107 K) halos.
• The feedback from a galactic bulge likely plays a key
role in galaxy evolution:
– Initial burst led to the heating and expansion of gas beyond
the virial radius
– Ongoing feedback keeps the gas from forming a cooling flow
and starves SMBHs
– Mass-loaded outflows account for diffuse X-ray emission
from galactic bulges.
Galaxies like ours reside in hot bubbles!
No overcooling or missing energy problem!