Transcript talk - Centre for Astrophysics and Supercomputing

X-ray to Radio Mapping of
the Virtual Cosmos by GCD+
Daisuke Kawata, Chris B. Brook, Tim W. Connors,
and Brad K. Gibson
Centre for Astrophysics and Supercomputing,
Swinburne University of Technology
1. Introduction
The Virtual Observatory
offers multi-wavelength (X-ray to radio) observational data
Direct and quantitative comparison
The Physics of Galaxy
Formation and Evolution
Synthesized multi-wavelength spectrum
including information about structure
Numerical Simulations of Galaxy Formation
can follow chemo-dynamical evolution of gas and stellar
components of galaxies
GCD+: Galactic Chemo-Dynamics Code (Kawata & Gibson 03)
3D vector/parallel tree N-body/SPH code
taking into account the complex dynamical and
chemical evolutions in forming galaxy self-consistently
DM, Gas, Star formation, SNe Feedback, and Metal Enrichment
Cosmological Simulations by GCD+  Virtual Cosmos
offers physical condition and chemical conposition of
gas and stellar components at various redshift and environments
plasma model, population synthesis, K-correction, etc.
Synthesized spectrum from gas and stars
+ absorption by IGM and ISM including dust
+ re-emission from dust
Self-consistent X-ray to radio mapping of Virtual Cosmos
2. Brief Introduction of GCD+
3D vector/parallel tree N-body/SPH code
DM and Stars ••• Tree N-body code
Gas ••• Smoothed Particle Hydrodynamics (SPH)
+ Radiative Cooling (MAPPINGSIII: Sutherland & Dopita)
depends on metallicity
+ Star Formation
SFR ∝ ρ1.5
(ρg ＞ 2 x10-25 g/cm3)
IMF: Salpeter type
+ SNe Feedback
SNeII and SNeIa
+ Metal Enrichment
SNe II, SNeIa, and
intermediate mass stars
H,He,C,N,O,Ne,Mg,Si, and Fe
3. Cosmological Simulation Model
follows the evolution of large scale structures as well as the galaxy
formation process, including gas dynamics and star formation
DM density map
I band image
standard ΛCDM (Ω0=0.3, λ0=0.7, h=0.7, Ωb=0.019h-1, σ8=0.9
Multi-Resolution Cosmological Simulation
snap shot @ z = 5.45
(grafic2: Bertschinger 01)
Highest Resolution Region:
mDM=2x105 M, εDM=0.14 kpc,
mgas=3x104 M , εgas=0.08 kpc
J band image
face-on
edge-on
5kpc = 0.83”
Mvir = 6x109 M
Vmax = 65 km/s
Comparison of
apparent size and
magnitude relation
with observations
Good agreement with HDF
and 2df galaxies
= reliable cosmological
simulation
High-z (z>5) galaxies which
should be detectable by
JWST
predicted size of these
galaxies < diffraction limit?
４. Analysis
gas
derive both X-ray/Optical properties
withminimum assumption
Synthetic R-band image + X-ray contours
stars
X-ray properties
fake X-ray Spectrum using XSPEC
vmekal plasma model
+ XMM EPN response function
Distribution of gas particles
(ρ,T,ZO,Mg,Si,Fe…)
Synthetic X-ray Spectrum
with XMM response function
Fit the spectrum using XSPEC vmekal model
 Lx,Tx,(Fe/H)x,(O/H)x…
Optical properties
Population Synthesis
SSPs: Kodama & Arimoto97
Synthetic Optical/NIR
Spectrum
Distribution of star particles
X-ray Spectrum
with XMM response function
(age,ZO,Mg,Si,Fe…)
 Luminosities and colours (MB, VK)
Current Status
Properties of high-z galaxies
Kawata, Gibson w/Windhorst (ASU)
Wavelength
Telescope
optical
HST, JWST
Previous Slides
Dynamics of high-z galaxies
Radio
Kawata, Gibson
(redshifted 21cm)
Tomorrow
Formation of elliptical galaxies
Kawata, Gibson
Sec. 5
Formation of Milky Way
Brook, Kawata, Gibson
w/Flynn (Tuorla)
Sec. 6
SMC and Magellanics Stream
Connors, Kawata, Gibson
Near future…
X-ray/optical
optical
(astrometry)
radio,optical
SKA, LOFAR
XMM, Chandra
Grand+Space
optical telescopes
Hipparcos,
(RAVE), GAIA
Parkes(HIPPASS),
ATCA, Southern
optical telescopes
Elliptical Galaxies
optical: stellar properties
X-ray: gas properties
B-R
5. An X-ray/Optical Study of Elliptical Galaxy Formation
in CDM Universe
5.1. Introduction
Coma
R
Any successful galaxy formation
scenario must explain both
observed X-ray and optical
properties.
Using self-consistent numerical
simulations, we are attempting
to construct such models for
elliptical galaxies.
Cluster & group 1
Xue & Wu (00)
10
5.2. Cosmological Simulation Model
High Resolution Region:
mDM=4x108M, εDM=4.3kpc,
mg=5.9x107 M , εDM=2.3kpc
Target galaxy
Largest galaxy in the simulation volume
Mvir=2x1013M  NGC4472 (Virgo elliptical)
3 Different Models
model A: adiabatic model
model B: cooling + weak feedback
model C: cooling + strong feedback
5.3. Results
model A: adiabatic model (no cooling = no star formation)
model B: with cooling and minimum SNe feedback
model C: with cooling and 100 times stronger feedback
5.3.1 LxTx relation
Adiabatic model (model A)
incompatible with data
 higher Lx and lower Tx
Inclusion of cooling leads to lower
Lx and higher Tx
 consistent with observed
Lx and Tx for NGC4472
(models B & C)
adiabatic simulation of clusters
(Muanwong et al. 01)
extrapolation of
cluster relation
(Edge et al. 91)
model A
model C
model B
ellipticals
(Matsushita et al. 00)
consistent with simulations of Pearce
et al. (00), Muanwong et al. (01)
Semi-cosmological galaxy formation model
advantage: less computational costs = can achieve higher resolution
disadvantage: not exactly follow the cosmological evolution,
e.g., might underestimate later accretion of the gas and
satellite dwarf galaxies
 update to full cosmological simulation in near future
5.3.3. Optical properties
ColourMagnitute relation
Problem!: An excessive
popuation of young stars
result due to cooling flow.
Colours are too blue,
regardless of feedback.
Coma ellipticals (Bower et al. 1992)
model C
model B
Double check in both X-ray and optical properties
gives stronger constraints on the theoretical models
6. Self-consistent modeling of Milky Way formation
Brook, Kawata, Gibson, Flynn
GAIA (also RAVE by UK Schmidt)
Astrometry, radial velocities, and
chemical composition for more
than 1 billion stars within 10 kpc
Chemo-dynamical modeling of formation and evolution of Milky Way
templates of Milky Way like galaxies with different formation histories,
such as major and minor merger history, to extract useful information
from such huge data set.
 what observational signatures tell what formation history.
The detailed formation history of Milky Way
Galactic Halo Stars in Phase Space: A Hint of
Satellite Accretion? Brook, Kawata, Gibson, & Flynn (2003, ApJL in press)
disrupted satellite which is identified at z=0.5
gas particles
Solar neighbourhood stars
Chiba & Beers (00)
eccentricity
Traditional interpretation: sign of rapid collapse (Eggen et al. 62)
Phase Space properties
Simulation
disrupted satellite
field stars
Observation
stars with low [Fe/H] and high e
Identical phase space distribution
Observed low [Fe/H]/high-e stars concentration can be explained by
the recent accretion of high-e orbit satellite.
= alternative explanation from “rapid collapse” scenario
7. Conclusion
GCD+ can provide observable values from numerical simulations.
= equivalent data to what the Virtual Observatory offers.
Ultimate Goal
The Virtual Observatory for Virtual Cosmos
Quantitative comparison between GCD+ VO for VC
and VO in multi-wavelength regime
should be exciting for
studies of galaxy formation and evolution
The Virtual Observatory is great for our science!
Contribution to the Theory Virtual Observatory (plan)
Public GCD+ VO for VC, using the same interface as VC
black box (= reducing process in observation)
store: the raw data
physical and chemical data for DM, gas, star particles

analysis code
synthesized image and spectrum
similar interface to VO
Image, spectrum
requests
luminosity function
user
looks great and all cosmological simulators can follow this
with minimum amount of effort (probably), however…
Problem: There is no perfect theoretical model.
i.e. we can create lots of different virtual cosmos
Therefore, the VO for VC should be provided with the description
of modeling.
 unified format for such description and classification of modeling
would be also important.
Interface allow to chose whose which model
e.g., GCD+ no feedback model or with strong feedback model
If all (cosmological) simulators follow this sort of idea, what is the benefit?
for simulator who knows differences between the codes
easy to compare with the results from other code  reduce the bugs
for observer or other theoretician
helpful to understand their observation and/or analytic model
confused by lots of different model?
 show the idea how to chose the model (whose one is the best,
in which case?) or enquiry
to prepare this, regular meeting and comparisons among
the simulator are necessarily…
5.3.3. Optical properties
ColourMagnitute relation
Problem!: An excessive
popuation of young stars
result due to cooling flow.
Colours are too blue,
regardless of feedback.
If the contribution of these young
stars is ignored, the observed
colour is recovered.
Coma ellipticals (Bower et al. 1992)
ignore young stars
(age<8 Gyr)
model C
model B
Young stars formed in later cooling might have a bottom-heavy IMF?
(Fabian et al. 1987; Mathews & Brighenti 1999)
and/or
Extra heating source (AGN?) to suppress star formation, but then the
LxTx relation and Lx-(Fe/H)x must be checked again.